a critical cattell's model of personality - brunel university research

A CRITICAL

0f

ANALYSIS

CATTELL'S MODEL OF PERSONALITY

P SAVILI, E PhD Thesis Brunel University

March 1978.

Education Department.

CONTENTS -------- Page

ACKNOWLEDGEMENTS 9

ABSTRACT 10

SECTION 1: GENERAL INTRODUCTION 11

1. INTRODUCTION 11

TH -- PERSONALITY Q ESTIO: 1 AIR. 12

16 = TIE : ACTOR ANA YT_C APP _t. P-L7RSO`. {AL IT'1i i' OAýri 4 CATTELL' S MODEL OF PET, 1SONALITY 19

DESCRIPTION OF CATTELL'S PRIMARY SOURCE TRAITS 25 CATTELL' S MEASUREMENT OF PERSONALITY BY QUESTIONNAIRE 50 CRITICISMS OF CATTELL' S MODEL CF PERSONALITY 52 CATTELL'S METHODS OF SCALE CONSTRUCTION 57

SECTION 2: HYPOTHESES, METHODO-L, OGY, PROCEDURE, THE SAMPLE 64

2. AIMS AND HYPOTHESES 64

INTRODUCTION 65 GENERAL HYPOTHESES 65 HYPOTHESES ON GROUP DIFFERENCES IN THE PRIMARY SOURCE TRAITS 66 HYPOTHESES ON THE RELIABILITY OF THE PRIMARY SOURCE TRAITS AW

70 HYPOTHESES ON THE INTERNAL CONSISTENCY OF THE PRIMARY SOURCE TRAITS 71 ALTERNATIVE MODELS IN THE PERSONALITY DOMAIN 72 AIMS

3. METHODOLOGY 75

INTRODUCTION 76 SAMPLING 80

4. PROCEDURE 83

THE PILOT STUDIES 84 TRAINING THE MAIN FIELDFORCE 88 INTERVIEWERS' INSTRUCTIONS 89 THE BACKGROUND QUESTIONNAIRE 102 DATA PROCESSING 108

5. DESCRIPTION OF TF SAMPLE 109

DESCRIPTION OF THE SAMPLE CHARACTERISTICS 112 THE PROBLEM OF NOP, i_RESPONSE 113

I

SECTION 3: GROUP DIFFERENCES AND THE PRIMARY SOURCE TRAITS

6. CROSS-NATIONAL AND SEX DIrFE. R . NCES

INTROD , -C T1ON CROSS-^. y' TONAL, DIFFERENCES: RESULTS SEX DIF E. RENCES: RESULTS DISCUSSION OF THE CROSS_1N. ATZONAL ANL SEX DIFFERENCES

7. SCCLkL CLASS DIFFERENCES

,. ]T: - -cr: c r

SCCI LASS DIr'_ ERELCr $: ES LTS SOCIAL CLASS DIFFERENCES: DISCUSSIO°.

8. AGE DIFFERENCES

INTRODUCTION RESULTS FOR PRODUCT MOMENT AND ETA COEFFICIENTS CATTELL' S QUADRATIC AGE CORR. c'. 'TIONS RESULTS ON CATTELL'S QUADRATIC AGE CCR. RsECTIONS CATTELL'S AGE CORRECTIONS: RESULTS AGE DIFFERENCES : DISCUSSION GROUP DIFFERENCES: CONCLUSIONS

SECTION 4: RELIABILITY AND THE PRIMARY SOURCE TRAITS

9. ALTERNATE FORM RELIABILITY OF THE PRlf; lARY SOURCE TRAITS

INTRCDUCTION TO RELIABILITY THEORY INTERNAL CONSISTENCY RELIABILITY TEST-RETEST RELIABILITY ALTERNATE FORM RELIABILITY THE STANDARD ERROR OF MEASUREMENT ALTERNATE FORM RELIABILITY COEFFICIENTS: RESULTS

10. AGE AND SOCIAL CLASS AS MODERATORS ON RELIABILITY COEFFICIENTS

r

IT. TROD` TION AGE AS A MODERATOR ON RELL, %BILITY COEFFICIENTS : RESULTS SOCIAL CLASS AS A MODERATOR ON RELIABILITY COEFFICIENTS: RESULTS AGE F.,. D SOCIAL CLASS AS MODERATORS CN RELIABILITY COEFFICIENTS:

DISCUSSION

11. VARLL. 1 . CNS IN THE STANDARD ERROR OF :, IEAS, REMENT

INTRODUCTION AGE AND SOCIAL CLASS AS MODERATORS C: 7 THE SEMS: RESULTS AGE L, D SOCIAL CLASS AS MODSPJ. TORS C THE SEMS: DISCUSSION RELIABILITY AND THE PRIMARY SOURCE TRAITS: CONCLUSIONS

Page

115

115

116 118 119 119

129

132 136

139

140 142 166 167 168 184 185

187

187

188 190 191 192

193 195

202

203 204 208

210

214

21.5 223 223 232

2

Pace

SECTION 5: INTERRZAL CONSISTENCY AND TTI-E PRIMARY SOURCE TRAITS 235

12. INTRODUCTION 236

THE PROBLEM OF REPLICATING CATTELL'S FACTOR SOLUTION 238 THE

-PROBLEMS GF RESPONSE FORMAT AND CORRELATIONAL METHODS 241 THE HSPQ STUDY 249

13. ESTI: ": ATTIýNG THE DIRECT VALIDITIES OF TIHE PRIMARY SOURCE TRAITS 258

I_ ODuc_Ia. ' 259 I `, AL1y ES OF TEE SOURCE .t AITS: PROCEDURE 251

RESULTS AI: ý SCÜSSICPý 262 SUi"iYARY AND CCiNCLUSIONS 269

14. FACTOR ANALYSIS AND THE PRIMARY SOURCE TRAITS 271

INTt CDLCTION 272 FACTOR ANALYSIS: RESULTS AND DISCUSSION 278 CONCLUSIONS 280

SECTION 6: SUiO1aP`_' AND CONCLUSIONS 281

15. SUi1NL. RY AND CC CLUSIONS 281

SU"21-. RY OF FI (DINGS ON THE PRIMARY SOURCE TRAITS 282 CONCLUSIONS ON GROUP DIFFERENCES AP 322 CONCLUSIONS REGARDING THE RELIABILITY OF THE PRIMARY SOURCE TRAITS 326 CONCLUSIONS ON THE INTERNAL CONSISTENCY OF THE MODEL 327 CONCLUDING RE. -, a. RKS 329 AREAS CF FURT}ER RESEARCH 330

16. REFERENCES

SMTTICN 7: APPENDICES

332

1-142

3

LIST0FTABLES

Pa e

SECTION 1: GENERAL INTRODUCTION

1. INTRODUCTION

1.1. Example of a Hypothetical 5x5 Correlation Matrix 18

. 2. First O- er (2- of 22

1.3. Second Order Factors of the 16PF 23

SECTION 2: HYPOTHESES, i. ': ETHODOLOGY, PROCEDURE, THE SAMPLE

2. AIMS AND HYPOTHESES

2.1. Sex and Social Group Differences and Age Trends in the Primary Source Traits Hypothesised from Cattell's Personality Model 68

5. DESCRIPTION OF THE SAMPLE

5.1. The 1971 British Standardisation Sample 110


6. CROSS-NATIONAL AND SEX DIFFERENCES

6.1. Significant Personality Differences Between the American and British Adult Groups Adjusted to Age 30 Years: Females 125

6.2. Significant Personality Differences Between the American and British Adult Groups Adjusted to Age 30 Years: Males 126

6.3. Significant Personality Differences Between British Men and Women 127

6.4. Summary of Cross-National and Sex Differences on Cattell's Primary Source Traits 128

7. SOCIAL CLASS DIFFER iNCES

7.1. The Sixteen Personality Questionnaire: Correlations Between Social Class and First Crder Factors 131

4

2. Eta Values (n) Regressing the Primary Source Traits on Socio-Economic Group: Females: Forms A+B 133

7.3. Eta Values (n) Regressing the Primary Source Traits on Socio-Economic Group: Males: Forms A+B 135

7.4. Correlations Between rge, Educational Age and Social Class in the 16PF: Forms A+B Adult Sample: Both Sexes 137 (N = 2012)

8. "GE DIFFERENCES

6.1. -Ste Values o-)n the rimary Source Traits For Six Age Groups Derived from Cattell's Age Curves 143

8.2. Conversion of N-Sten Age Curve Values to Z Scores Based on Total Group Statistics 144

8.3. The Sixteen Personality Questionnaire: Correlations Between Age and First Order Factors 162

8.4. Eta Values (n) Regressing The Primary Source Traits on Age: Females: Forms A+B 163

8.5. Eta Values (n) Regressing the Primary Source Traits

on Age: Males: Forms A+B 164

8.6. 16PF Age Trends in Personality Form A+ B Males Age Corrected Scores 169

8.7.16PF Age Trends in Personality Form A+B Females Age Corrected Scores ýº 170

8.8. Eta Values (n) For Age Corrected Scores Form A+B Females 171

8.9. Eta Values (n) For Age Corrected Scores Form A+B Males 172

SECTION 4: RELIABILITY AND THE PRIMARY SOURCE TRAITS

9. ALTERNATE FORM RELIABILITY OF THE PRIMARY SOURCE TRAITS

9.1. Group Alternate Form Reliability Coefficients (Forms A&B) For Each Primary Source Trait 196

9.2. Standard. Errors of Measurement: Primary Source Traits (Stens) - Forms A/B 197

9.3. Conversion of the Alternate Form Reliability Coefficients to Fisher Z Statistics: Adult Males and Females 199

9.4. Comparison of Adult Reliability Coefficients urith Those of Undergraduates: Form A with B 200

5

Pa e

10. AGE AND SOCIAL CLASS AS MODEFATORS ON RELIABILITY COED FICIENTS

10.1. Rank Frequencies of the Alternate Form Reliability Coefficients Classified by Age Group 204

10.2. Summary Table: Reliability Classified by Age 205

10.3. Conversion of Primary Source Trait Reliability Coefficients to Fisher Z Statistics: Both Sexes by Age 206

10. _.

Rar t: Fe '1cies .. _ e Alt? rra: e ý'o m tRc2llab_ý"i f_y

Ccetficients Classified by Socio-Econor^ic Group 207

10.5. Social Class as a Moderator on the Reliability of the Primary Source Traits: Summary Table 208

10.6. Conversion of Primary Source Trait Reliability Coefficients to Fisher Z Statistics: Both Sexes by Social Class 209

11. VARIF. TIONS IN THE STANDARD ERROR OF MEASUREMENT

11.1.16PF Form A SEMs According to Age and Social Class: Both Sexes 225

11.2. 16PF Form B SEMs According to Age and Social Class: Both Sexes 226

11.3. SEM Rank Frequencies for Social Class: Both Sexes: Form A 227

11.4. SEM Rank Frequencies for Social Class: Both Sexes: Form B 227

11.5. Summary Tables: SEMs and Social Class as a Moderator Variable: Forms A and B 228

11.6. SEM Rank Frequencies For Age: Both Sexes: Forms A and B 229

11.7. Summary Tables: SEMS and Age as a Moderator Variable: Forms A and B 230

SECTION 5: INTERNAL CONSISTENCY AND THE PRIMARY SOURCE TRAITS

12. INTIRODIXTION

12.1. Correlation Between the Point-Biserial and Biserial Approaches to HSPQ Item Scale Correlations 252

12.2. High School Personality Questionnaire: Form A, Girls NL mber of Higher Extra than Intra Scale Correlations for Each Factor (Biserials) 254

12.3. High School Personality Questionnaire: Form A, Girls Number of Higher Extra than Intra Scale Correlations = for Each Factor (Product Moment) 255

6

Pace

12.4. High School Personality Questionnaire: Form A, Girls Correlation Between Product Moment and Biserial Approaches (All Items) 257

12.5. High School Personality Questionnaire: Form A, Girls Correlation Between Product Moment and Biserial Approaches (Factors) 257

13. ESTIMATING T}E DIRECT VALIDITIES OF THE PRIMARY SOURCE TRAITS

13.1. Composite Item-Scale Correlations (Direct Validities) For Cattell's Primary Source Traits 263

13.2. Correlations Between the Source Traits of the 16PF as Scale Scores: Forms A and B 266

13.3. Comparison of the Item-Composite Correlations with Scale Intercorrelations 267

14. FACTOR ANALYSIS AND THE. PRIMARY SOURCE TRAITS

14.1.16PF Factor Analysis: Males: Eigen Values and Percentage of Trace 274

14.2.16PF Factor Analysis: Females: Eigen Values and Percentage of Trace 274

14.3. Promax Factor Analysis: Males: Table of Factor Loadings 275

14.4. Promax Factor Analysis: Females: Table of Factor Loadings 276

14.5. Correlations of the Oblique Primary Factors: Males 40 277

14.6. Correlations of the Oblique Primary Factors: Females 277

Slr-TION 6: SUMMARY AND CONCLUSIONS

15. SUMMARY AND CONCLUSIONS

15.1. Consistency of Sex, Social Class and Age Differences on Cattell's Source Traits fgr the Seven Factor Solution 323

7

LIST0FFIGURES

SECTION 1: GENERAL INTF%OD'1CTION

1. INTRODUCTION

Pacge

1.1. Cattell's Factor Theory of Personality 24

1.2. A Graphical Representation of Cattell's Concept of 'Balance Seecifics' in Scale Construction 63

SECTION 2: HYPOTHESES, ý"IETHODOLOGY, PROCEDURE, THE SAMPLE


2.1. Abstracted Personality - Age Relationships for each of Cattell's Primary Source Traits 69

2.2. A Study of Reliability and Structure in Cattell's Model of Adult Personality 74

ý. PROCEDURE

4.1. Contact Sheet 85

4.2. Background Questionnaire 86

4.3. The Sampling Units and Standard Regions of the 16PF: British Standardisation Sample (N = 2227) 107


8. AGE DIFFERENCES

8.1. Analysis of Personality Age Curves 141

8.2. - 8.17. Age Trends in Personality - Factors A- Q4 146 - 161

SECTION 5: INTERNAL CONSISTENCY AND THE PRIMARY SOURCE TRAITS

12. INTRODUCTION

12.1. Item Analysis Procedure 251

I

8

ACKNOWLEDGEMENTS

I would very much like to thank Janice Field, Clive Downs and

Iris Parry of the Test Department of the National Foundation for

Educational Research for all their help in this project. I would

also like to thank B-111 .: bey. Joan MacFarlane Smith and

Sue Rodger, previously of the British Market Research Bureau

for their contribution to the fieldwork organisation.

Thanks must go to Jill Tarryer, previously of the NFER, for

her statistical advice and to her staff for dealing with the

large quantity of data generated by this study. I would also

like to acknowledge discussions with Steve Blinkhorn, who shares

a common interest in personality measurement.

Very sincere thanks are given to Irene Stevens for all her

excellent work in typing this thesis and for remaining patient

through the various drafts and amendments which this thesis

has undergone.

Finally, I would like to state my thanks to Professor Furneaux

and Dr Ormerod of the Education Department at Brunel University

for all their help and advice.

AP

it

9

ABSTRACT

The personality theory expounded by R. B. Cattell and his

collaborators is one of the most influential models of

personality in modern-day psychology. The central and most

controversial elements of this theory are the sixteen primary

source traits, which make up the model.

This thesis describes research into the primary source traits

as represented by Cattell's Sixteen Personality Factor

Questionaire (16FF). From the responses of a large sample of

general population British adults (N: - 2,000) evidence is

presented to suggest that: -

1. inconsistencies exist between the relationships of Cattell's

personality scales with variables such as age and social class

and the relationships postulated from his model.

2. the precision of measurement afforded by certain scales is

extremely low, especially in lower social class subjects.

3. the model as represented in the scales of the 16PF is

untenable in terms of its internal consistency.

4. Cattell's personality scales are not in fact the most

parsimonious description of the personality domain.

SECTION 1:

GENERALINTRODUCTION -------------------

0

1. INTRODUCTION

1ll

THE PERSONALITY QUESTIONNAIRE

It is generally believed that Francis Galton devised the first

questionnaires in the 1880s for his studies of mental imagery at his

Kensington laboratory. His statement in the 'Fortnightly Review' of

1884 that 'the character which shapes our conduct is a definite and

durable "something" and therefore .... it is reasonable to attempt to

measure it' is not unlike many contemporary views of personality.

Despite the problems associated with the questionnaire approach to

personality measurement - one is reminded of the comment of Sir

Frederick Bartlett reported by Buzzard (1971), 'I don't know ... they

may be all right ... they always seem to me to overestimate the

self-knowledge of the subject and to underestimate his sense of

humour' - the technique flourished after World War I, until, in the

1970s there are literally hundreds of published instruments.

Most self-report personality questionnaires consist of a series of

statements or questions about the individual, his interests and attitudes.

These the subject endorses according to his strength of agreement. To

make scoring easier and more dependable, the subject is usually

required to select the most appropriate answer from choices provided

in the questionnaire form. Because item formats will concern us at a

later stage we shall briefly consider how items are typically presented.

First, there is the true/false technique of, for example, the Eysenck

Personality Inventory, where the subject has to choose between two

clear-cut alternatives.

12

I enjoy going to parties.

a. b.

Yes No

This approach has the advantage of being statistically manageable (cf

the pass/fail dichotomy cn an ability test item) and helps to overcome

certain types of bias. Ho. ", ever, it can become frustrating to the

; 'ho may find -t both di ficu_: l t and unnatural to answer

every item on a yes/no basis. There is the danger that the item

format will cause the subject to challenge the sense of questions

being asked him, and so become inconsistent in his answers. Indeed,

it is not unknown for respondents to invent a middle response category.

A second approach is the use of the forced choice technique, where

one or more of a number of alternatives are chosen by the subject as

most or least characteristic of himself. For example: AP

Mark the statement which shows which you prefer doing:

a. Going to parties.

b. Reading a serious book.

Inventories based on this system are the Gordon Personal Profile, the

Gordon Personal Inventory and the Edwards Personal Preference Schedule

(EPPS). One attraction of this approach is that by matching the items,

social desirability response sets (the tendency for subjects to put

up a good front and show socially acceptable characteristics) can to

some extent be controlled. In practice it has proved difficult to match

itens exactly in terms of social desirability and it appears that

instruments like the EPPS can in fact be deliberately faked to create

13

more favourable impressions (Dickens, 1959). It is probable also that

items matched in social desirability for one occupation or set of

circumstances will not be matched for other occupations or other sets

of circumstances. In addition, because certain personality types are

more prone to presenting answers in a favourable light, it has been

argued that attempting to eradicate the effects of social desirability

is like throwing the baby out with the bath water.

Another problem concerns scaling properties. The strength of each

variable is gauged relative to the strengths of all other variables

and as a result, it is impossible to gain high or low scores across

all scales. Cattell (1944) has termed this 'ipsative' measurement

and strictly we should not compare scores of different individuals,

although this is often done and extensive norm tables produced.

The third type of item format is an extension from the true/false

item to three or more possible choices. For example:

I like going to parties.

a. I,. C.

True False ?

or I like going to parties

a. b. Co

Frequently Occasionally Never

14

or I like going to parties.

a. b. c. d. e.

Always Frequently Occasionally Rarely Never

The first example is similar to the method of item presentation in

the Minnesota rultiphasic Personality Inventory (? 4PI) and the

secor. _. the personality questionnaires developed by RB Cattell.

This approach is more acceptable to the subject, although the

dimensionality of the middle response is often a problem, especially

if it becomes a waste-bin category of 'don't knows', 'in betweens',

'occasionally' and 'I don't understand the question'. In these

multiple choice formats the effects of faking and social desirability

are not so much controlled but detected by lie, motivational distortion

and other such keys. In the case where there are three choices the

responses become statistically less manageable, whilst five or more may

cause a practical problem in the hand marking of the questionnaires.

Obviously with most questionnaire items the prior assumption has

already been made that the underlying trait represents a continuum.

It is not usually acceptable for a subject to endorse two answers even

though he may feel that both are equally appropriate and it is this lack

of flexibility which many criticise in the questionnaire approach.

Goldberg (1970), corrsnenting upon the current proliferation of

personality inventories ('inventory' and 'questionnaire' will be used

interchangeably throughout this thesis), distinguished between those

inventories developed in response to pressures from society to deal with

specific problems and those stemming from theoretical concepts about the

15

nature of personality. In the first group are questionnaires like the

California Psychological Inventory (CPI), designed to measure practical

outcomes of social behaviour, and the Minnesota Multiphasic Personality

Inventory (MMPI), which measures aspects of personal adjustment, and

those which are concerned with vocational choice, such as the Strong

Vocational Interest Blank. Instruments of this kind tend to be

constructed on an empirical basis where items are selected which

discriminate between different groups of people.

The second category of questionnaire includes the inventory measures

of introversion-extraversion, those based on Murray's (1938) system

of needs like the Edwards Personal Preference Schedule, those which

are psychoanalytic in conception like the Dynamic Personality Inventory

and the various gqestionaires which have been based on factor analytic

methods of-construction.

THE FACTOR ANALYTIC APPROACH TO PERSONALITY MEASUREMENT

Many questionnaires have been produced, each, as Cronbach (1970) has

pointed out, using items from its predecessors, adding new ones and

scoring them in new combinations of traits and scales. Cattell (1973)

has written "... let us be clear that nothing stops anyone from making

up questionnaires for any trait he cares to imagine and since there are

an indefinitely large number of conceivable traits and roughly 50,000

psychologists, questionnaires have in fact been the prolific rabbits of

the psychomobic world! ".

16

Trait names become a source of much confusion. What was 'Introversion'

to one author was not introversion to another; 'Dominance' in system X

bore a striking resemblance to the trait of aggression in system Y.

It was the construction of what appeared to be arbitrary personality

scales that the factor analysts attacked. The only way to resolve

this Tower of Babel, the factor analytic school argued, was to

'discover' clearly defined personality variables, building blocks,

which would serve the psychologist as the table of elements serves

the chemist. Rather than have a multitude of scales to predict all

sorts of social, educational and clinical outcomes (the criterion -

related or empirical approach), is it not possible to find a basic

set of constructs by which we can understand human personality?

The factor analysts believe it is.

Although there is considerable variation within the school, the basic

principles are simple. Working with a set of questionnaire scales

which has been administered to a large group of subjects, it is possible

to measure the degree of association between the variables by correlating

every scale with every other scale. The correlations between each pair

of variables are represented in a correlation matrix as in the example,

Table I. I.

'r

Looking closely at Table 1.1, we can ask whether any of

these hypothetical variables are related to any degree with one another.

For example, do subjects consistently gain a high score on one scale

and not on another? The latter is clearly the case with scales R and S;

17

TABLE: 1.1. Example of a hypothetical 5x5 correlation matrix

VARIABLE

R S T X Y

R -. 92 . 13 -. 11 . 82

pq S -. 92 . 01 . 27 -. 86 T . 13 . 01 . 89 . 08 X -. 11 . 27 . 89 . 21 Y . 82 -. 86 . 08 . 21

the negative correlation of -0.92 means that there is a strong tendency

for those getting high scores on Scale R to get low scores on Scale S

and vice versa. The direction of the relationship is of no major

importance because it merely reflects the scoring conventions of the

scales, but what is important is the presence of a strong association

between the two variables. Indeed, the relationship is so high that

we might wonder whether these scales were in any way distinct and

whether R and S are simply different names for the same underlying

variable.

Visual inspection of this correlation matrix suggests the presence of

two underlying dimensions or 'factors' - the first common to variables

R, S and Y and the second to T and X. This kind of visual examination

of correlation matrices of course illustrates only the logic of factor

analysis, However it is in fact true to say that one of the most

sophisticated factor theories of personality, that developed by R. B.

Cattell, was to some extent formulated by such methods (Cattell 1943).

18

CATTELL'S MODEL OF PERSONALITY

There can be little doubt of the importance of RB Cattell and his

collaborators to current personality theory. Cattell's research into

personality has spanned more than thirty years of large scale study

and experiment and his writings have introduced not only new ways of

investigating the domain but also new ideas on the very structure and

functioning of personality. Howarth (1976) writes 'Cattell's long

continued and widely documented research into personality factors

represents the single most monumental effort to operationalize those

sources of individual differences often referred to as 'temperament'

or 'personality' (and is comparable to that of Guilford on cognitive

factors)". According to Berg (1972), "Cattell's major contribution

will undoubtedly be seen as the discovery of the major factorial

dimensions in rating, questionnaire and objective tests".

Cattell (1950) has provided us with a deliberately general definition

of personality as "that which permits a prediction of what a person

will do in a given situation. The goal of psychological research in

personality is thus to establish laws about what different people

will do in all kinds of social and general environmental situations ...

personality is ... concerned with all the behaviour of the individual,

both overt and under the skin".

19

Thus, Cattell's theory of personality stresses both the empirical and

descrip: -ive and is concerned with establishing a set of personality

eiemen s shared by all individuals. Rather than have a multitude

of scales to predict all sorts of social, educational and clinical

outcomes, Cattell set himself the task of discovering a basic set

of constructs by which human personalty could be fully understood.

It was in fact by way of behaviour ratings of one group of people

by another ('L' or 'Life data') that Cattell's major dimensions of

personality were derived (Cattell, 1943). Using the adjectives and

adverbs descriptive of behaviour listed in the English dictionary,

Cattell first defined. the clusters upon which his traits were to be

used by eliminating synonyms and visually inspecting correlation

matrices. Cattell then extended this work to self-ratings (questionnaires

or 'Q' data) and to objective tests (laboratory experiment or 'T' data). s

At the first level of Cattell's model of personality there are the

many surface traits of personality. Surface traits are directly

observable aspects of human behaviour, the labels which are frequently

used in our everyday experience. At the next level are the first order

or primary source traits of personality which are combined to produce

the complex and many observable surface traits. It is the primary

source traits which are the universal basic variables in the total

personality structure. They are defined both by letters of the

alphabet, technical names and popularly descriptive labels.

The order of the primaries (which are labelled from A through to Q4)

are based on their descending contribution to behavioural variance

20

as estimated from Cattell's studies. Whilst Cattell refers to the

first order factors as functionally unitary, they are oblique

factors, ie. they show some correlation between themselves.

Because the primary source traits are in fact correlated, Cattell

has factor analysed his variables to some eight second crder factors.

ii list o; these and the pr m r`y' source tja "`_s is contained in Tables 1.2.

and 1.3.. In addition, Figure 1.1. shows the relationship between the

primary traits and the major second order factors.

To Cattell the problems of studying personality are extricably wrapped

up with the problems of measurement in psychology. In early studies

he used correlational techniques on personality ratings ('L' data)

but as computer technology became more developed, his ideas and

theories rested more and more on the technique of factor analysis AP

(leading at times to bitter argument with other members of the factor

analytic school, Cattell, 1972 and Eysenck, 1972). Still later work

was with self-ratings into what has become the more heavily researched

area of questionnaire measurement and it is the personality questionnaire

which has become the most developed vanguard of his model. Certainly,

Cattell's personality scales have had very wide applicability in the

western world, only rivalled by the '? IPI and the measures developed

by Eysenck.

t

I

21

TABLE: 1.2.

FIRST ORDER (PRI -ýSRY; FACTORS OF THE 16PF

LOW SCORE DESCRIP'ION Dr^ SC T HIGH SCORE

I, -z DESCRIPTION

RESERVED, DEIACrtED, CRITICAL, OUTGOING, WARMHEARTED, EASY- ALOOF A GOING, PARTICIPATING

(Sizothymio) (Affectothymia, formerly cyclothymic)

LESS INTELLIGENT, CONCRETE. MORE INTELLIGENT, ABSTRACT- THINKING

' B THINKING, BRIGHT (Lower scho csric mental capacity) (Higher scholastic mental capacity)

AFFECTED BY FEELINGS, EMOTIONAL- EMOTIONALLY STABLE, FACES LY LESS STABLE, EASILY UPSET C REALITY, CALM, MATURE

(Lower ego strength) (Higher ego strength)

HUMBLE, MILD, ACCOMMODATING, ASSERTIVE, AGGRESSIVE, STUBBORN, CONFORMING E COMPETITIVE

(Submissiveness) (Dominance)

SOBER, PRUDENT, SERIOUS TACITURN HAPPY-GO-LUCKY, IMPULSIVELY , (Desurgency) F LIVELY, GAY, ENTHUSIASTIC

(Surgency)

EXPEDIENT, DISREGARDS RULES CONSCIENTIOUS, PERSEVERING. , FEELS FEW OBLIGATIONS

' " STAID., MORALISTIC

(Weo <er superego strength) (Stronger superego strength)

SHY, RESTRAINED, TIMID, VENTURESOME, SOCIALLY BOLD, 4P- THREAT-SENSITIVE H UNINHIBITED, SPONTANEOUS

(Threctio) (Pormio)

TOUGH-MINDED, SELF-RELIANT, TENDER-MINDED, CLINGING, REALISTIC, NO-NONSENSE I OVER-PROTECTED, SENSITIVE

(Harris) (Premsic)

TRUSTING, ADAPTABLE, FREE OF SUSPICIOUS, SELF-OPINIONATED, JEALOUSY, EASY TO GET ALONG L HARD TO FOOL WITH (Ataxia) (Protension)

PRACTICAL, CAREFUL, CONVENTION- IMAGINATIVE, WRAPPED UP IN INNER AL, REGULATED BY EXTERNAL URGENCIES, CARELESS OF PRACTICAL REALITIES, PROPER (Proxernio) (Autio) MATTERS, BOHEMIAN

FORTHRIGHT, NATURAL, ARTLESS, SHREWD, CALCULATING, WORLDLY, UNPRETENTIOUS N PENETRATING

(Artlessness) (Shrewdness)

SELF-ASSURED, CONFIDENT, APPREHENSIVE, SELF-REPROACHING, SERENE Q WORRYING, TROUBLED

; J-troubled adequacy) (Guilt proneness) CONSERVATIVE, RESPECTING ESTAB- EXPERIMENTING, LIBERAL LISHED IDEAS, TOLERANT OF TRADI- T

, ANALYTICAL. FREE-THINKING IONAL DIFFICULTIES (Conservatism) (Radicalism)

GROUP-DEPENDENT, A "JOINER" AND SELF-SUFFICIENT, PREFERS OWN SOUND FOLLOWER Q2 DECISIONS, RESOURCEFUL'

(Group adherence) (Self-sufficiency)

UNDISCIPLINED SELF-CONFLICT, FOL- CONTROLLED, SOCIALLY PRECISE, LOWS OWN URGES, CARELESS OF 43 FOLLOWING SELF-IMAGE PROTOCOL (Low integration) (High self-concept control)

RELAXED, TRANQUIL, TENSE, FRUSTRATED, DRIVEN, UNFRUSTRATED Q4 OVERWROUGHT

(Low ergic tension) (High ergic tension)

22

TABLE: 1.3.

SECOND ORDER FACTORS OF THE 16PF

SECOND CHIEF FIRST LOW SCORE ORDER HIGH SCORE ORDER FACTORS

DESCRIPTION FACTORS DESCRIPTION INVOLVED

INTROVERSION EXTRAVERSION A+, E+, F+, Invia QI Exvia H+, Q2-

LOW ANXIETY HIGH ANXIETY C-, H-, L+, O+, Adjustment QII Anxiety Q3-, Q4+

SENSITIVITY TOUGH POISE A_, I-, M- Pathelria QIII Cortertia

, E+, L+.

DEPENDENCE INDEPENDENCE E+, L+, M+, Subduedness QIV Independence Q1+, Q2+

NATURALNESS QV DISCREETNESS N+, A+, M.., O-

COOL PRODIGAL REALIST QVI SUBJECTIVITY I+, M+, L-.

LOW HIGH INTELLIGENCE QVII INTELLIGENCE B+

LOW SUPEREGO HIGH SUPEREGO STRENGTH QVIII STRENGTH G+, Q3+, F-

23

aw

HW as

N Self-

Cj sufficient

H

ö

En

ii

w c7 CH++

Ü

c.., Happy-go--lucky

41 Assertive

d Outgoing

H Tender-minded

Experimenting

0 Conscientious

z Shrewd

z Imaginative

ai Intelligence

x Venturesome

Emotional Stability

Tense

o Apprehensive

A suspicious

`"ý Controlled

24

L! J1

ýa w ýý Ha

4J H

N

aj

N N

8

I

DESCRIPTION OF CATTELLtS PRIMARY SOURCE TRAITS

As is to be expected with a trait psychologist, Cattell's theory is

very much rooted in the constructs he uses in describing individual

differences in personality.

We shall see that Cattell borrows heavily from a long list of other

psychologists to describe the traits derived from his factor analytic

research. He makes reference for example to the work of James, Freud,

Jung, McDougall, Thurstone, Maslow, Eysenck and Guilford. Indeed Cattell

frequently uses nautical expressions and analogies, seeing himself and

his researchers as modern-day explorers discovering and re-discovering

concepts and ideas observed by previous less well equipped voyagers.

This is not to say that Cattell denies the possibility of traits

unique to the individual but he believes that the primary goal of

psychology is to discover and measure the core of primary

source traits common to all individuals. Once this is achieved, it

will then be possible to observe the inter-relationships of such

traits which make each individual unique and which will allow actual

predictions of their likely behaviour under different environmental

conditions.

FACTOR A- RESERVED VS. OUTGOING

(a) Low Score Description

Reserved, Detached, Critical, Cool (Sizothymia, previously Schizothymia)

The person who scores low (step of 1 to 3) on Factor A tends to be

stiff, cool, skeptical and aloof. He likes things rather than people,

25

to be precise and 'rigid' in his way of doing things and in personal

standards, and in many occupations these are desirable traits. He

may tend, at times, to be critical, obstructive, or hard.

(b) High Score Description

Outgoing, Warmhearted, Easy-going, Participating (Affnctothymia, previously Cyclothy ia)

The person who scores high (sten of 8 to 10) on Factor A tends

to be goodnatured, easy-going, emotionally expressive (hence

naturally Affectothymia), ready to cooperate, attentive to

people, softhearted, kindly, adaptable. He likes occupations

dealing with people and socially-impressive situations. He

readily forms active groups. He is generous in personal

relations, less afraid of criticism, better able to remember

names of people. Ap

(c) Origins in Psychological Theory

Cattell traces this concept back to the cyclothymia-versus-

schizothymia dichotomy of Kretschmer (1921) which in turn was

adapted from Bleuler's (1911) cyclic-versus-schizophrenic label

of the functional psychoses. In rating studies and questionnaire

studies this is the largest variance factor.

(d) Life Associations

Cattell claims that affection is associated with a warm home

background. Higher scorers on affectia get married earlier,

usually stay married longer and have an endomorphic bodybuild.

Women score significantly higher than men. A fairly substantial }

genetic component is hypothesised for this trait.

26

(e) Age Trend

Cattell states that this trait decreases until about 10 years of

age; then there is a steady and relatively rapid increase to 30

, rohere it levels off.

FACTO ? :3- LESS Ts'TELL:: GES i VS. MORE INTELLIGENT


Less Intelligent, Concrete-Thinking (Lower scholastic-mental capacity)

The person scoring low on Factor B tends to be slow to learn and

grasp, dull, given to concrete and literal interpretation. His

dullness may be simply a reflection of low intelligence, or it

may represent poor functioning due to psychopathology.

r


More Intelligent, Abstract-thinking, Bright (Higher scholastic mental capacity)

The person who scores high on Factor B tends to be quick to grasp

ideas, a fast learner, intelligent. There is some correlation

with level of culture, and some with alertness. High scores

contraindicate deterioration of mental functions in pathological

conditions.


Cattell's view of intelligence remains very similar to the

hierarchical theories of British psychologists like those of

Spearman and Vernon (Cattell 1965). He has certainly not rejected

27

the meaning of an overall processing factor ("g") although he

has spent much time in trying to isolate two main influences

within such a construct. These he has named fluid (processing

ability unaffected by cultural influences) and crystallized

(intelligence gained as a result of learning experiences)

general ability.


Cattell (1965) believes that some intelligence is important to

predicting behaviour in a great many situations such as future

academic and occupational performance.

(e) Age Trend

Following increases up to adolescence, fluid ability then starts

to decline whilst crystallized ability can improve until about

40 years of age when it declines.

FACTOR C- AFFECTED BY FEELINGS VS. EMOTIONALLY STABLE


Affected by Feelings, Emotionally Less Stable, Easily Upset (Lower ego strength)

AP

The person who scores low on Factor C tends to be low in frustration,

tolerance for unsatisfactory conditions, changeable and plastic,

evading necessary reality demands, neurotically fatigued, fretful,

easily emotional and annoyed, active in dissatisfaction, having

neurotic symptoms (phobias, sleep disturbances, psychosomatic

complaints, etc. ). Low Factor C score is common to almost all

28

I

forms of neurotic and some psychotic disorders.


Emotionally Stable, Faces Reality, Calm, Mature (Higher ego strength)

The person who scores high on Factor C tends to be emotionally

ma-tu-ne, stable, realistic about life, unruffled, possessing ego

strength, better able to maintain solid group morale. Sometimes

he may be a person making a resigned adjustment to unsolved

emotional problems.


Cattell claims that this pattern in questionnaire items is clearly

the reality behind Freud's concept of ego structure strength.

v (d) Life Associations

Low ego strength we are informed by Cattell is associated with

unemployment, truency and delinquency; high scores go with group

participating and skilled occupations. It is low in those from

family backgrounds of matriarchal dominance and high among those

who are disciplined by reasoning as opposed to punishment by the

father. Low ego strength is also shown in some chronic physical

disabilities and in marital instability.

(e) Aqe Trend

Increasing strongly with age and then levels off at about 40 years.

29

FACTOR E- HTJ N. B: ýE VS. ASSERTIVE

(a) Low Score Descriotion

Ht,; ble, Mild, Accomumodating, Conforming (Submissiveness)

The person who scores low on Factor E tends to give way to others,

to b= docile, and to conform. He is often dependent, confessing,

anxious for obsessional corr`ctnoss. This passivity is part of

many neurotic syndromes.


Assertive, Independent, Aggressive, Stubborn (Dominance)

The person who scores high on Factor E is assertive, self-assured,

and independent-minded. He tends to be austere, a law to himself,

hostile or extrapunitive, authoritarian (managing others), and W

disregards authority.


This trait has a long history (Maslowl, 1940,1954). It has a

second tributary in animal ethology from McDougall (1932).


Those high on this factor are seen as unconventional and rule

breaking in group behaviour. There are significant relations we

are told to an early authoritarian family atmosphere and to

stricter application of discipline. Among 18-22 year olds there

is a tendency to reject sexual restraints (Maslow, 1940), to have

30

a great number of social contacts, to drink immoderately and to

have less than average sleep needs. This factor is high in

psychopaths and low in neurotics, alcoholics, drug addicts, all

psychotics except manics, in stably married wives and unstably

married husbands (Cattell and Nesselroade, 1968). Males score

significantly higher than females on this factor.

(e) Aqe Trend

Increasing rapidly in early life, then levels off and declines

in the middle years.

FACTOR F- SOBER VS. HAPPY-GO-LUCKY


Sober, Prudent, Serious, Taciturn (Desurgency)

The person who scores low on Factor F tends to be restrained,

reticent, introspective. He is sometimes dour, pessimistic,

unduly deliberate, and considered smug and primly correct by

observers. He tends to be a sober, dependable person.


Happy-go--Lucky, Impulsively Lively, Gay, Enthusiastic (Surgency)

The person who scores high on this trait tends to be cheerful,

active, talkative, frank, expressive, effervescent, carefree.

s

He is frequently chosen as an elected leader. He may be impulsive

and mercurial.

31


Cattell points out that in early years this was confused with

his second order factor of exvia (extraversion). In Cattell's

model, surgency is a distinct component, similar to one of the

four components in extraversion conjectured by Jung.

(d) Life AGscx-iations

Cattell (1973) hypothesises that this trait is related to high

economic status of the family but fewer work responsibilities.

There is low restraint in sexual activities, there are many

social contacts and in small groups a high scoring person is often

voted an effective speaker and leader. There will be appreciable

cultural differences on this trait we are informed: for example,

it is high in Australians and low in the Chinese and Japanese.

0 (e) Age Trend

Rising in adolescence but decreasing in adulthood.

FACTOR G- EXPEDIENT VS. CONSCIENTIOUS


Expedient, Evades Rules, Feels Few Obligations (Weaker superego strength)

The person who scores low on Factor G tends to be unsteady in

purpose. He is often casual and lacking in effort for group

undertakings and cultural demands. His freedom from group influence

may lead to anti-social acts, but at times makes him more effective,

32

while his refusal to be bound by rules causes him to have less

somatic upset from stress.


Conscientious, Persevering, Staid, Rulebound (Stronger superego strength)

The person who scores high on Factor G tends to be exacting in

character, dominated by sense of duty, persevering, responsible,

planful, 'fills the unforgiving minute'. He is usually

conscientious and moralistic and he prefers hard-working people

to witty companions. The inner 'categorical imperative' of this

essential superego (in the psychoanalytic sense) should be

distinguished from the superficially similar 'social ideal self'

of Q3+.

" (c) Origins in Psychological Theory

The concept of conscience, Cattell points out, is dateless; it is

discussed by Hebrew prophets and Greek dramatists. Of course this

was an important construct in Freud's personality system but

Cattell claims that the factor-analytic evidence of a distinct

and measurable unitary structure in questionnaire data is only

about twenty years old.


Higher superego strength is related to such life associations

as: election to leadership in small groups, higher sexual morals,

having suffered a bereavement, working long hours and being

dedicated to some form of work. In the home background high G

33

is related to more years of education of parents and greater

warmth of the parental relationship. Women tend to score

significantly higher than men on this factor.

(e) Aqe Trend

Moderate but decelerating increase with age.

FACTOR H- SHY VS VENTURESOME


Shy, Restrained, Diffident, Timid (Threctia)

The person who scores low on this trait tends to be shy, with-

drawing, catious, retiring, a 'wallflower'. He usually has

inferiority feelings. He tends to be slow and impeded in speeh

and in expressing himself, dislikes occupations with personal

contacts, prefers one or two close friends to large groups, and

is not given to keeping in contact with all that is going on

around him.


Venturesome, Socially-bold, Uninhibited, Spontaneous (Parmia)

The person who scores high on Factor H is sociable, bold, ready

to try new things, spontaneous, and abundant in emotional response.

His 'thick-skinnedness' enables him to face wear and tear in dealing

with people and gruelling emotional situations, without fatigue.

34

However, he can be careless of detail, ignore danger signals,

and consume much time talking. He tends to be 'pushy' and actively

interested in the opposite sex.


Like surgency this factor has been mistaken for extraversion but

Cattell claims that it has been isolated in questionnaire data

as a sub-factor of exvia since the late 1940s. Parmia is an

acronym to express the theory that it represents parasympathetic

predominance at the Hi- pole and sympathetic system predominance

at the H- pole. The latter amounts to high susceptibility to

threat - hence threctia.


This factor is associated positively with the demonstration of W

affection by parents and negatively with strict discipline

(Barton, Cattell, 1972). High H is significantly related to

numerous social contacts, a rating as cooperative and friendly

as a child but lazy and unorganised, an indication of association

with endomorphic or mesomorphic, pyknic build. Parmia tends to be

high in psychopaths and moderately young delinquents but it is

low in alcoholics, neurotics and attempted suicides. Men on

average tend to score higher than women.

(e) Age Trend

The trait increases strongly through middle age for men but only

slightly for women.

35

FACTOR I- TOUGH-MINDED VS. TENDER-MINDED


Tough-minded, Self-reliant, Realistic, No-nonsense (Harries)

The person who scores low on Factor I tends to be practical,

realistic, masculine, independent, responsible, but skeptical of

subjective, cultural elaborations. He is sometimes unmoved, hard,

cynical, smug. He tends to keep a group operating on a practical

and realistic 'no-nonsense' basis.

(b) High Score Descriptioli

Tender-minded, Dependent, Overprotected, Sensitive (Premsia)

The person who scores high on Factor I tends to be tender-minded,

day-dreaming, artistic, fastidious, feminine. He is sometimes v

demanding of attention and help, impatient, dependent, impractical.

He dislikes crude people and rough occupations. He tends to slow

up group performance, and to upset group morale by unrealistic

fussiness.


Cattell points out that the concept of a dimension of this kind is

even older than extraversion and goes back to the tender-versus-tough

minded trait of William James and earlier sources traced by Roback

(1927). As a demonstrably distinct questionnaire data factor it

was replicated in the mid-1940s by Cattell and soon after as a

pattern in attitudes by Eysenck (1952). This factor, it is claimed,

is particularly important in terms of the behavioural variance it

36

accounts for in children. In some studies the premsia-versus-harria

pattern has been called masculinity-versus-femininity but Cattell

claims that this is a gross oversimplification in that only a

fraction of the male-female difference resides in I.


Pransia has significant positive relations to lax discipline

in the home background, warmth in permitting free aggression

and more control by reasoning than by punishment. It is also

associated with poor health, chronic illness, hypochondria and

low athletic interest and performance. This factor also distinguishes

those who go to higher education, take. up social responsibilities

and who enter the professions. Women score higher than men in

premsia.

(e) Age Trend

Steep decrease into adolescence followed by slight continuous

rise throughout life.

FACTOR L- TRUSTING VS. SUSPICIOUS


Trusting, Adaptable, Free of Jealousy, Easy to Get on With (Alaxia)

410.

The person who scores low on Factor L tends to be free of jealous

tendencies, adaptable, cheerful, un-competitive, concerned about

other people, a good team worker.

i

37


Suspicious, Self-opinionated, Hard to Fool (Protension)

The person who scores high on Factor L tends to be mistrusting and

doubtful. He is often involved in his on ego, is self-opinionated

and interested in internal, mental life. He is usually deliberate

in his actions, unconcerned about other people, a poor team member.


In his survey of questionnaire factors Cattell (1946. ) found several

suggestions of a dimension resembling paranoid introversion. In

the questionnaire realm he defines this factor as a general use

of projection as a defence mechanism. The low end of the pole,

alaxia, implies the quality of resigned relaxation and acceptance

of a socially low expression of individuality. AW


As a child the protensic person had a high confidence in himself,

found school unpleasant but felt superior. In adulthood he is often

a strong drinker, does not marry early. and tends to be socially

unpopular. This factor is high in criminals, drug addicts,

neurotics and attempted suicides. Males tend to score higher

than females.

(e) Age Trend

Declines entering the middle years but increases appreciably

after forty.

38

FACTOR M- PRACTICAL VS. IMAGINATIVE


Practical, Careful, Conventional, Regulated by External Realities, Proper (Praxernia)

The person who scores low on Factor M tends to be anxious to do

the riggt things, attentive to orac tical matters, and subject to

the dictation of what is obviously possible. He is concerned

over detail, able to keep his head in emergencies, but sometimes

unimaginative.


Imaginative, Wrapped up in Inner Urgencies, Careless of Practical Matters, Bohemian (Autia)

The person who scores high on Factor M tends to be unconvention, a,

unconcerned over everyday matters, Bohemian, self-motivated,

imaginatively-creative, concerned with 'essentials', and oblivious

of particular people and physical realities. His inner-directed

interests sometimes lead to unrealistic situations accompanied

by expressive outbursts. His individuality tends to cause him

to be rejected in group activities.


Cattell claims that this source trait had no early recognition

clinically but only became evident through his factor analytic

work in the 1940s. Those high in autia have been hypothesised

to produce a greater intensity of images and ideas relative to

39

sensory stimuli. In essence their behaviour is one of accepting

an intense inner subjective life relative to the reality of the

outer world and its demands.


Autious people are likely to choose further and higher education

rather t-hen a job, take part in social organisations but are

dissenters and are rejected by groups although they are called

significant members. There is some history of delicate health.

Their parents disapproved of their sex life and companions;

they lack regard for order and are untidy in dress. This factor

appears in scientists and writers. They are accident-prone and

often change jobs. Culturally, the latin countries are more

autious than, for example, Britain and Australia. In American

groups females tend to score higher than males on this factor. AW

(e) Aqe Trend

Increases in early life and then declines.

FACTOR N- FORTHRIGHT VS. SHREWD


Forthright, Natural, Artless, Sentimental (Artlessness)

The person who scores low on Factor N tends to be unsophisticated.,

sentimental, and simple. He is sometimes crude and awkward, but

easily pleased and content with what comes, and is natural and

spontaneous.

40


Shrewd, Calculating, Worldly, Penetrating (Shrewdness)

The person who scores high on Factor N tends to be polished,

experienced, worldly, shrewd. He is often hardheaded and

analytical. He has an intellectual, unsentimental approach to

situations, an approach akin to cynicism.


Again, this factor was not recognised in the clinical literature

but was identified hesitatingly Cattell informs us in the 1940s

and has been confirmed since then: it does not account for a

great deal of questionnaire variance but apparently adds to

criterion variance predictions.

e

(d) Life Associations all

Positively related to social status and freedom from pathology.

In small groups such people are positive influences in problem

solving and are rated high in keeping to the point. This trait

relates to less social integration in marriage (Barton et al 1972. )

and, we are informed, manic depressives and schizo-affectives. Males

tend to be higher on this factor.

(e) Age Trend

Tends to show a steady slight increase throughout life.

i

41

FACTOR 0- PLACID VS. APPREHENSIVE


Placid, Self-assured, Confident, Serene (Untroubled adequacy)

The person who scores low on Factor 0 tends to be placid, with

unshakable nerve. He has a mature, unanxious confidence in

hi^: saiý and 's-. is caoacity to deal with things. He is resilient

and secure, but to the point of being insensitive of when a

group is not going along with him, so that he may evoke antipathies

and distrust.


Apprehensive, Worrying, Depressive, Troubled (Guilt proneness)

The person who scores high on Factor 0 tends to be depressed, moody,

a worrier, full of foreboding, and brooding. He has a childlike

tendency to anxiety in difficulties. He does not feel accepted

in groups or free to participate. High Factor 0 score is very

common in clinical groups of all types.

(c) Origins in Psychology Theory

This trait has a long history in clinical psychology - under such

titles as melancholy agitation, depressive tendency, moodiness and

cycloid tendency. It is not a liability to particular pangs of

guilt so much as a global sense of inadequacy, loneliness and tears.

It might also be considered (Cattell 1973) as a heightened demand of

the gregarious erg which leaves an insatiable loneliness when it is

not met.

42


Those of high guilt proneness are loyal in friendship, take social

failure badly and more observed to be moody. In home background

guilt proneness is associated with discipline by physical punishment

and less overall behavioural guidance. Females tend to score

significantly higher on this factor. High guilt proneness is

associated with less than average school participation and

occupationally is high in those of need of rehabilitation.

High 0 is almost universal in clinical cases.

(e) Age Trend

Falls after early adolescence,, tending to rise slightly after

middle age.

FACTOR Q1 - CONSERVATIVE VS. EXPERIMENTING


Conservative, Respecting Established Ideas, Tolerant of Traditional Difficulties (Conservatism)

V

The person who scores low on Factor Qlis confident in what he has

been taught to believe, and accepts the 'tried and true', despite

inconsistencies, when something else might be better. He is

cautious and compromising in regard to new ideas. Thus, he tends

to oppose and postpone change, is inclined to go along with

tradition, is more conservative in religion and politics, and tends.

not to be interested in analytical 'intellectual' thought.

43


Experimenting, Critical, Liberal, Analytical, Free-Thinking (Radicalism)

The person who scores high on Factor Ql tends to be interested in

intellectual matters and has doubts on fundamental issues. He is

skeptical and inquiring regarding ideas, either old or new. He

terc's to be more well informed, less inclined to moralise, more

inclined to experiment in life generally, and more tolerant of

inconvenience and change.


This factor was discovered as a unitary trait of large variance

by Thurstone and Chave (1929) and by Guilford before Cattell's

early factor analytic work. It has consistently appeared in

attitude studies conducted by Eysenck (1944) but apparently shows

instability of pattern across subcultures such that it is difficult

to use the same scale across groups varying in terms of age and sex.

Cattell regards this factor not as a narrow trait of political

attitude but as a more general predisposition to try the new.


There is a positive relationship with affiliation with progressive

political parties and the conservative pole is related to strong

family ties and more than average male dominance in the family.

The trait is moderately associated with good school achievement

and entering higher education. Males tend to score significantly

higher than females on this trait.

44

(e) Aqe Trend

Cattell claims no clear age trend for this factor.

FACTOR Q2 - GROUP-DEPENDENT VS. SELF-SUFFICIENT


Group-dependent, A 'Joiner' and Sound Follower (Group adherence)

The person who scores low on Factor Q2 prefers to work and make

decisions with other people, likes and depends on social approval

and admiration. He tends to go along with the group and may be

lacking in individual resolution. He is not necessarily

gregarious by choice; rather he needs group support.


Self-sufficient, Prefers Own Decisions, Resourceful (Self-sufficiency)

The person who scores high on Factor Q2 is temperamentally independent,

accustomed to going his own way, making decisions and taking action

on his own. He discounts public opinion, but is not necessarily

dominant in his relations with others (See Factor E). He does not

dislike people but simply does not need their agreement or support.


This is one of Cattell's major introversion traits which he is.

claimed to have broken down into sub-factors. It seems that those

high on self-sufficiency have a lack of gregariousness and a

steadiness of emotion that favours dependence on the self and a

45

lack of rewarding response by others.


Positively related to those whose parents were maritally harmonious;

those with this trait were seclusive as children and stubborn in

school. They developed early, were considered mature and in general

showed good scholastic ach-levement. No clear pattern of differences

between males and females has developed on this primary source trait.

(e) Age Trend

No age change in males; slight increase in females.

FACTOR Q3 - UNDISCIPLINED SELF-CONFLICT VS. CONTROLLED

(a) Low Score Description all

Undisciplined Self-conflict, Careless of Protocol, Follows Own Urges (Low integration)

The person who scores low on Factor Q3 will not be bothered with

will control and regard for social demands. He is not overly

considerate, careful, or painstaking. He may feel maladjusted,

and many maladjustments (especially the affective, but not the

paranoid) show Q3-.


Controlled, Socially-precise, Following Self-image (High self-concept control)

The person who scores high on Factor Q3 tends to have strong

control of his emotions and general behaviour, is inclined to be

46

socially aware and careful, and evidences what is commonly termed

'self-respect' and regard for social reputation. He sometimes

tends, however, to be obstinate. Effective leaders, and some

paranoids, are high on Q3.


Catteil (1965) follows the notion of a dynamic structure different

from either the ego or superego strength from McDougall (1932).

Cattell also acknowledges the work of Thurstone who discovered a

reflective factor in questionnaire. This factor is supposed to

represent the clarity with which the individual has developed

a self-concept and the strength with which he adheres to it.


As a child the person of high sentiment was usually healthy and the

family harmonious. He controls his temper and especially likes

science and mathematics at school. The loading on scholastic

success is usually positive and this trait is usually higher in

those following professional occupations. Males tend to score

higher than females on this factor.

(e) Aqe Trend

After adolescence it increases positively levelling off in middle

adult years.

47

FACTOR Q4 - RELAXED VS. TENSE


Relaxed, Tranquil, Torpid, Unfrustrated (Low ergic tension)

The person who scores low on Factor Q4 tends to be sedate, relaxed,

composed, and satisfied (not frustrated). in some situations, his

oversatisfaction can lead to laziness and low performance, in the

sense that low motivation produces little trial and error.

Conversely, high tension level may disrupt school and work.

performance.

(b) High Score Description .

Tense, Frustrated, Driven, Overwrought (High ergic tension)

The person who scores high on Factor Q4 tends to be tense, excitable,

restless, fretful, impatient. He is often fatigued, but unable to

retrain inactive. In groups he takes a poor view of the degree of

unity, orderliness and leadership. His frustration represents an

excess of stimulated, but undischarged, drive.


This factor can be initially traced back to Freud's 'id', McDougall's

'total instinctual energy' and even Aristotle's 'conation'.


In home upbringing, ergic tension is associated with use of

punishment and lack of demonstration of affection. This trait is

associated with high blood pressure and poor ability to perform under

stress, loading negatively on school achievement. No clinical groups

48

are low on ergic tension. This factor is high in alcoholics,

homosexuals, exhibitionists and all neurotic syndromes.

Frustration and depression increase this factor; low rank in

school and job, unhappy love affairs, bereavements. Females

tend to score higher on this scale.

(e) Aqe Trend

Increases in early life, then declines throughout adulthood

until about 40 years of age.

40

49 1

CATTELL'S MEASUREMENT OF PERSONALITY BY QUESTIONNAIRE

As we have already noted, the Sixteen Personality Factor Questionnaire

(16PF), develop--d by Cattell and his associates in the United States,

has found wide acceptance in Britain, Ireland and the rest of the

Western World. Buros (1972) shows a yearly average of 80 references

i : ieen -19r. 5

.d 11-6`ß for ý.: e 16E': and the ýýf nd 1s uowards.

The instrument has been put to a multitude of uses, impossible to

review here but the list includes studies on academic achievement

(Warburton et al., 1963; Crotis, 1968), teacher training (Start, 1966,

1968), computer programmers (Morris and Martin, 1972), birth order

(Allman and White, 1968), racing drivers (Johnsgard and Ogilvie, 1968),

sensory deprivation (Arnhoff and Lean, 1963), angina pectoris

(Bakker and Levenson, 1967) and hypnotic susceptibility (Hartman, 1966).

Cattell has never held that the sole or even the best method of

investigating human personality is the questionnaire. Whilst it is

true, possibly as a result of the practical advantages and economies

of the self-report technique, that his work has in recent years

centred on the questionnaire, Cattel2 has strongly emphasised the

importance of ratings and laboratory tests in personality measurement.

Cattell regards the life record (IL--data), self-rating (4-data) and

objective test (T-data) as different ways of measuring the fundamental

traits which are descriptive of human personality. Much time has

been spent, not always with complete success (Becker, 1960), in

attempting to match the factors across these three approaches to

50

personality measurement.

It is perhaps a little surprising that Cattell should have so early

adopted this view. After all it is not unreasonable to suspect that

the constructs used by an individual in rating himself may vary from

those used to rate others. Such difficulties have been acknowledged

in practice by the presence, for example, of four factors in the

16PF which have not been discovered in L or T data.

The first editions of the 16PF appeared in the early 1950s with the

issue of revisions over the next 15 years or so. The latest and

most refined versions were published in 1968 and 1969 - parallel

(alternate) Forms A and B, each of 187 items and the somewhat easier

language Forms C and D each of 105 items. Forms A and B take

approximately 50 minutes each to complete and C and D approximately

35 minutes. In addition to the main primaries and secondaries,

Cattell has provided weights to apply to the first order scales

for the calculation of four 'derived criteria'. These are amorphous

measures related to external criteria.

We should also mention the questionnaires which Cattell has developed

for measuring children's personality (High School Personality

Questionnaire, Children's Personality Questionnaire and Early School

Personality Questionnaire), motivation - (Motivation Analysis Test

and School Motivation Analysis Test) and pathological dimensions

(Clinical Analysis Questionnaire).

51

CRITICISMS OF CATTELL'S MODEL OF PERSONALITY

Despite Cattell's prodigious research output, his theories have not

gone without question. It is true to say that most criticism has

centred on the primary source traits which make up his model. Here

three main areas of debate can be distinguished. First there is the

matter of the consistency of group differences postulated by Cattell

from his model. Second is the reliability of measurment afforded by

Cattell's first order personality scales and third, the internal

consistency and very structure of the model itself.

1. Group Differences

In a study of undergraduate personality using the 16PF, Saville and

Blinkhorn (1975) reported that the primary source traits showed an

inconsistent pattern of differences between arts and science and

student and non-student groups depending upon the particular form

of the questionnaires in use.

Similarly these authors were able to demonstrate that the relationship

between Cattell's source traits and academic attainment as measured

by degree class showed considerable variations between questionnaire

forms. They conclude: -

"There is very considerable inter-form variation in the direction

and size of inter-group differences, such that to specify differences

in terms of 'Cattell's Factor N', for instance, rather than '16PF

Form A, Factor N', may be to invite immediate contradiction by those

52

using another form of the test. "

and they continue: -

"This inter-form variation may be thought to account for the

vanishingly small increment in multiple correlations with degree

CiS... .. =Lned by sun: nj scale scores on a1 ternate forms. It

would seem that at least for the purposes of multiple regression,

16PF 'alternate' forms should be treated as independent. "

We have also seen that Cattell places a great deal of emphasis

upon the relationship of the primary source with variables such as

age, sex and social class. How the source traits change over age

for example, is of very great importance to the understanding of

his model. Indeed, Cattell is one of the few authors who provides 40,

graphical and mathematical expressions of age trends in personality.

Despite this fact, there has been no independent research on this

aspect of the theory nor for that matter on other important group

differences which Cattell has postulated from his model of personality.

2. The Reliability of Measurement Afforded by the Model

One of the principal areas of criticism of Cattell's primary source

traits has been the severe lack of reliability which certain scales

demonstrate. Saville and Blinkhorn (1975) in their study of

approximately 2,500 British university students found that as

judged by the alternate form method on Forms A and B of the 16PF,

sever, factors had reliability coefficients between values of . 44 to

53

. 60 (C, E, G, I, Ql, Q2 and Q3) whilst four scales (B, L, M and N)

had an estimated reliability as low as . 43 or less.

In an interesting paper Eysenck (1972) considered the question of

the unreliability of Cattell's source traits and their inter-

relationship with other scales. Taking the published table of

inter_"or el ations be' : vee the 169= SC31C scores of 423 male and

535 female American college students (Cattell et al, 1970),

Eysenck corrected the coefficients for attenuation (unreliability),

paying particular attention to the five 'anxiety' scales (C-, L, O,

Q3- and Q4) and five 'exvia' scales (A, E, F, Q2- and H).

On the five scales which contribute to Cattell's anxiety second

order factor, twelve of the twenty intercorrelations exceeded unity.

On this evidence Eysenck states that "as far as one can see, "

practically all the information contained unreliably in the primaries

is contained reliably in the second order factor; very little

information indeed is left over for contribution by the primaries".

In the case of the exvia-invia factors, the position was less clear

cut in that whilst eight of the twenty coefficients exceeded . 9,

twelve correlations were below this level. Eysenck points out

this in part is due to certain of Cattell's exvia primaries

overlapping with the anxiety and other second orders. He concludes:

"The figures given are not incompatible with a general view which

would regard the primaries advocated by Cattell as random groupings

of items either measuring extraversion or neuroticism or occasionally

both. Such a view would also be compatible with the fact that several

54

writers have found it impossible to replicate Cattell's factors

in independent analyses, using both his items and his methods of

analysis and rotation".

3. The Internal Consistency of the Model

li' ldc itiv`rl o the reliability of Cattell's source

traits as scale scores, considerable debate has centred on the

internal consistency of the model as represented by the items

in Cattell's personality questionnaires.

In an early study Levonian (1961) using Form A of the 16PF on

252 American psychology students, intercorrelated item responses

and factor scores and found low internal consistency for many of

Cattell's factors. He concluded, "while item homogeneity is not

the sine qua non of test construction, the extreme item

heterogeneity demonstrated in the study would seem to indicate

a critical appraisal of the 16PFQ".

Becker (1961) who had earlier demonstrated low correspondence

between Cattell's Questionnaire ('Q' data) and rating ('L' data)

studies, showed in relating the 16PF with some of Guilford's scales,

that with psychology students, certain of Cattell's factors were of

low reliability and that at best only eight factors could be

distinguished in the 16PF.

In two well-known papers Howarth and Browne (1971) and Howarth,

Browne and Marceau (1972) claim that despite its prominence Cattell's

55

16PF has been developed along unsatisfactory lines and as a consequence

does not measure univocal primary factors as is the "often reiterated

claim of its originator and proponent". They point out that Cattell's

preferred method of analysing clusters of inter-related items is

probably based not upon any technical superiority but rather that

total item factoring was not available in terms of computer

ceehniiüq ý hen the c3ie$ were originally deeveioned. They point

out that Comrey and Duffy . (1968) tried to group Cattell's items

into 'factored homogeneous item dimensions (FHIDS)' leading them

to conclude that Cattell's factors were complex rather than unitary.

Arguing that items chosen to represent or 'mark' a factor would

intercorrelate highly among themselves and not outside the given

factor, Howarth et al (1972) took the responses of 567 university

students and calculated the some 17,000 inter-item correlations 4P

on Form A of the 16PF. They, in fact, found that only about 30%

of the intra-factor item correlations were significant at the

. 01 level; the most homogeneous item groupings being found in

Factors F (Surgency), H (Social Boldness) and Q4 (High Tension).

In addition to the packaging of items into FHIDS, they criticise

Cattell's method of targetting his factor solution to the twelve

factors, which he observed in two early rating studies (Cattell,

1945,1947); for when the item intercorrelation matrix was factor

analysed by the method of Householder-Wilkinson only ten factors

were extracted.

The latest appraisal of Cattell's system is in a recent paper by

56

Howarth (1976). Whilst the author remains in sympathy with Cattell's

strategy regarding multivariate definition and comprehensive choice

of variables, he believes there were questionable aspects to Cattell's

tactics. Rather than a system defence, Howarth argues, what is

now needed is a movement of investigators towards "a consensus

buttressed by co-ordinating studies". As one important factor

in such work, he continues, populations should be tested from

various social levels.

CATTELL'S METHODS OF SCALE CONSTRUCTION

All the studies summarised above have in some respect or the other

revolved around the reliability of Cattell's scales. Indeed, the

issue of test reliability and more specifically item homogeneity, a

is central to the dispute on the validity of Cattell's model of

personality. The papers by Howarth et al (1971 and 1972), Eysenck

(1972), Levonian (1961) and Becker (1961) have all taken issue with

the low homogeneity of item content of Cattell's first order

primary personality traits as represented in the 16PF.

Perhaps it is not altogether surprising that the central issue

in personality measurement by questionnaire has been reliability

theory. Because we can never be absolutely sure of what we are

measuring in psychology, because our constructs are always to some

extent arbitrary in nature and because they have no real existence

in themselves, it is imperative with general. purpose instruments

that they cover the domain to be measured as completely, succinctly

57

and accurately as possible. What is more the use of factor analysis

in terms of identifying or 'discovering' the major dimensions of

personality means that any model rests to a great extent on the

homogeneity of the items concerned.

It is certainly true that a classical test and factor theory have

emphasised the importance of measuring clear cut, non over-lapping

variables and this is generally achieved by producing scales made

up of items of high intra-factor consistency and low extra-factor

correlation. Moreover, it is perhaps significant that the aim of

apparently less orthodox approaches to measurement, such as

Rasch Item Analysis and Scaling (1960), is to produce scales

of high homogeneity.

In answer to the criticisms on his first order factors, Cattell

has been clear and prolific in reply. "The primary factors give

one most information and we would advocate higher strata contributors

only as supplementary concepts ... it is a mistake, generally, to

work at the secondary level only, for one certainly loses a lot of

valuable information present initially at the primary level"

(Cattell 1970. Cattell, 1974 ).

On the connected and important matter of item structure, Cattell

has taken a radical departure from orthodox thinking and argued the

case that in many respects. the aim of ever increasing homogeneity

is unrealistic. In simple terms he reasons that if we accept a

definition of homogeneity as the average correlation between all

the items in a scale, then perfect homogeneity suggests a scale

58

which provides no more information than a single item. In personality

measurement he claims that the desire for scales of high internal

consistency has led authors to repeat almost identical items; a

practice which may be suitable for measuring narrow, specific

factors but not appropriate to the measurement of general influences

in behaviour.

Cattell (1972) has argued: "The charge that item intercorrelations

are low .... are therefore correct; but one would have thought that

well informed psychometrists today would have recognised that ....

this might actually be a virtue. Indeed, it is actually a deliberate

feature of the design of the 16PF and other tests by the present

writer which aimed to measure a broad. personality factor over so

widely diversified a set of items that no specific variance due

to situations will be spuriously included to raise the correlations". 40

Cattell continues"..... the ideal homogeneity is not the greatest

attainable but an optimum low value based on various combined

psychometric considerations".

Cattell in this paper entitled 'The Importance of Factor Trueness

and Validity Versus Homogeneity and Orthogonality in Test Scales'

(Cattell 1964) goes on to point out that classical test theory has

almost exclusively concerned itself with what he has termed 'item

metrics' (Kuder and Richardson 1937, Cronbach 1951). He points

out that it is not sufficient to examine the homogeneity of the

test as a whole but "good, adapted construction often requires

that the test, like a watch, shall have parts with different functions,

properly balanced in certain ways". The essence of Cattell's

59

argument is that the best combination of items to measure a factor

are those which have high loadings with the pure factor but in fact

have low correlation amongst themselves. He develops the concept of

factor trueness as the extent to which the scale as a whole measures

only the required factor. Contrary to orthodox reliability theory,

he makes the case that the addition of suppressor items, that is

72 1oadi r: (j on ,- ___ ý a- which

would usually be considered unsatisfactory by conventional reliability

methods, can have a definite advantage in purifying the measurement

of the underlying trait afforded by a given scale of items.

Thus, his'concept of direct validity' is concerned with the multiple

correlation of a set of items with the pure factor. Clearly, as with

any multiple regression problem, maximisation of the relationship

of the sum of item responses which represents the scale score tend

to be higher with low rather than high homogeneity. The primacy of

high homoege; ty is rejected in that it tends to bring lower total

validity for the multiple correlation of items with the pure factor.

Cattell (1964) has provided a simple graphical representation which

is reproduced in Figure 1.2. Here items j and k are loaded to the

extent of . 71 on the wanted factor w but in opposite directions on

the unwanted factor u. They are said to suppress each other on the

unwanted and re-enforce each other on the wanted factor. It is then

possible to get full validity in that j+k gives a factor true

scale score with complete heterogeneity in that the two items

correlate zero.

60

Thus, it has become clear that the validity of Cattell's theory is

very much rooted in the psychometric characteristics of the scales which

he has developed to measure the variables in this personality model.

More specifically the arguments have centred upon the internal

consistency of the 16PF and Cattell's aim to produce 'buffered'

scales of low homogeneity. Cattell (1974) also indicates that much

criticism has been based upon obsolete editions of his questionnaires

collected on university samples.

Whilst scale homogeneity has been the central issue in the controversy,

there have been other and more general matters related to reliability

theory; questions focused on reliability of measurement as a person

as much as a test characteristic. First, there is the point mentioned

above that very little is known about the reliability of questionnaire

measurement, let alone the 16PF, on groups other than university

students. This is concerned with the effect of what Ghiselli (1967)

has termed 'moderator variables' on reliability. In test theory

is is commonly assumed that error variance is uncorrelated with

true variance. In moderator theory however, error scores may be

related to other variables. Probably the more important moderators

(Cattell 1974) are likely to be the sex, social group and age of

respondent but whereas it is fair to hypothesise that the accuracy of

measurement might vary from one specific subject cluster to another,

there have been no reported studies of its effect on personality

scale measurement.

In summary then, as have been pointed out by Howarth (1972) a critical

evaluation of the item structure of the 16PF, how it is affected by

61

certain moderator variables such as age, social class and score level

and an exploration of Cattell's concepts of factor trueness and

direct validity as they relate to orthodox reliability theory,

are long overdue.

Certainly, if Cattell is correct in his assertion that item

homogeneity has aS an L: nn, -ce S3iv and n-,: ý-alistl, c goal -in

psychometrics, it will have far ranging implications for conventional

measurement theory as it is practised today. It is the aim of this

research to investigate these areas as well as the topics of

suppressor action in scale measurement, and the role of potential

moderators in personality measurement in the questionnaire realm.

AP

62

FIGURE 1.2.

A GZAPHICAL REPRESENTATION OF CATTELL'S CONCEPT OF BALANCE SPECIFICS' IN SCALE CONSTRUCTION

W

Wanted General Personality Factor

Unwar*ed General Factor

63

SE tTION 2:

HYPOTHESE S, METH0D0L0GY,

PR 0C EDURE. THE SAMPLE

#º


64

INTRODUCTION

In view of the criticisms outlined in the previous chapter, it would

seem that Cattell's theory of personality is much in need of independent

research. Like most complex theories, there is unlikely to be a

definitive once and for all test of Cattell's model. What is

required is careful programmatic research conducted on various

aspects of the model with the aim of observing how well the theories

fit well collected data bases. Accordingly, a general hypothesis

has been developed from the literature review and from this three

groups of subordinate hypotheses have been distinguished.

GENERAL HYPOTHESES

The general hypothesis has been developed that: -

Cattell's theory of personality rests on a system of primary traits

which are untenable in terms of: -

(a) the verifiability of important group differences in personality

postulated from Cattell's theory.

(b) the reliability of measurment afforded by the personality scales

which represent the model.

(c) the internal consistency and structure of the items defining the

primary source traits

and that despite the existence of some independent, reliable and

descriptively useful factors, Cattell's model of personality is in part

65

made up of redundant traits. That is, Cattell's primary source traits

cannot be described as the simplest, most reliable definition of the

personality domain.

The first of the three groups of subordinate hypotheses concerns the

strength and direction of differences between certain groups of subjects

which have beer. mrpoth`sised from Cattell's model. Of course, there are

an infinitely large number of such groups but the most important would

seem to be the effects upon the primary source traits of cross-national

variation, sex, social class and age. Second, there is the question

of the reliabilities of the primary source traits and their fluctuation

as a result of sex, social class and age acting as

moderator variables. Finally, there is the related issue of the

internal consistency, homogeneity factor trueness and construct

validity of the model.

mo

HYPOTHESES ON GROUP DIFFERENCES IN THE PRIMARY SOURCE TRAITS

We have seen (Chapter 1) that Cattell's theory of personality, at least

as regards the primary source traits, is relatively rich in potential

hypotheses. It is difficult at times to. be absolutely sure which of

his statements remain theoretical and which are backed up by scientific

evidence. Nevertheless, it is possible to take these factor descriptions

and generate meaningful hypotheses on expected cross-national sex and

social class differences and age trends in the primaries. These

hypotheses are listed in Table 2.1.

The table contains a total of sixty-six hypotheses, twenty of which are

the null hypotheses of no difference between means. It is thought

66

unnecessary to review these here, especially as they originate in the

descriptions of the primary source traits as postulated by Cattell which

were summarised above.

As for the nature of these hypotheses, it is evident that the sex

difference cells for each of the primary traits in the table

simply contain either the null hypothesis of no difference between

the sexes, or the hypothesis that males or females will be found to be

the higher scoring sex. The social group variable is hypothesised

either to have no relationship with each of the primary source traits

or to show a moderate or slight and positive or negative linear

association; this gives a total of five types of hypothesised relationships.

In addition, we shall also seek to test the hypothesis of the non-linearity

of the regression of personality on social class.

The age variable is a little more complicated. Whilst certain personality

factors have a hypothesised positive or negative straightline relationship

with age, the majority, Cattell claims, depart from the simple linear

model. Accordingly, the hypothesised age trends in personality have

been abstracted in Figure 2.1., which shows more clearly the six trend

lines which Cattell envisages for his source traits.

Cattell (1972) has recently provided age curves for the primary source

traits but it is necessary to replicate these on groups of subjects

selected by better sampling methods than those originally employed by

Cattell. As with social class, it will also be possible to test for

the non-linearity of regression of the primary source traits on age.

67

TAFLE: 2.1.

SEX AND SOCIAL CROUP DIFFERENCES AND AGE TRENDS IN THE PRIMARY SOURCE TRAITS HYPOTHESISED FROM CATTELL'S PERSONALITY MODEL

PRIMARY SEX CROSS-NATIONAL SOCIAL SOURCE DIFFERENCE DIFFERENCE AGE GROUP TRAIT (ON AVERAGE) (ON AVERAGE) TREND DIFFERENCE

A Females American Increases 17-30 yrs, No Outgoing higher higher then levels off. Differences

B No No Negative Moderate positive Intelligence difference difference relationship relationship

C Increases up to 40 E: No ycea r s, n? ev els Slight posit ive

Wince () f. re 1 tons 1p

E Males American Level up to 40, Assertive higher higher then slight No

decline. Differences

No American Rises in adolescence Slight negative I{appy_Go-Lucky difference higher then decreases relationship

G Females No Moderate positive Conscientiousness higher difference but decelerating Moderate positive

increase with age relationship

H Males American Strong positive Venturesome higher higher males, Weak No

positive females. Differences

= Females No Slight positive Tender-minded higher difference Slight positive relationship

L Males No Declines up to 40, No Suspicious higher difference then increases. Differences

Females American Declines after Slight positive imagination higher higher adolescence relationship

N Males No Slight positive Slight positive Shrewd higher difference increase relationship

O Females No Falls after adoles- Apprehension higher difference cence, rises Slight negative

slightly after 40. relationship

01 Males No Moderate negative Experimenting higher difference No clear trend relationship

C22 No No No trend in males, Self-sufficient difference difference slight increase Slight positive

in females. relationship

Q3 Males British Strong & positive, Controlled higher higher levels off at Slight positive }

40 years. relationship

(24 Females No Increases until 40, Slight negative Tense higher difference then levels off. relationship

68

FIGURE: 2.1.

P e r sT or na ai

t -i.

Y

ABSTRACTED PERSONALITY - AGE RELATIONSHIPS FOR EACH OF CATTELL'S PRIMARY SOURCE TRAITS

A: NEGATIVE: POSITIVE/ZERO/NEGATIVE RELATIONSHIPS

L (Suspicious) 70 (Apprehensive)

Zero

B (Intelligence) M (Imagination)

Age

P

e r sT or na ai 1t i t y

B: ZERO: POSITIVE/ZERO/NEGATIVE RELATIONSHIPS

-Se 5X->

4o

Zero ', Zero Ql (Experimenting) Q2 (males) (Self-sufficient)

E (Assertive)

P e r sT or na ai 1t i t V

Age

C: POSITIVE: POSITIVE/ZERO/NEGATIVE RELATIONSHIPS

G (Conscientious) H (Venturesome)

ye I (Tenderminded) N (Shrewd)

Qov Q2 (females)(Self-sufficient)

Zero A (Outgoing) - /T _t_ _. __I rr. . .-..

Age

'. ' zmouonat ýtab1llty) F (Happy-go-Lucky)

Q3 (Controlled)

Q4 (Tense) }

69

HYPOTHESES ON THE RELIABILITY OF THE PRIMARY SOURCE TRAITS

We have seen that a great deal of controversy has revolved around the

question of the reliability of Catteil's primary source traits.

Eysenck (1972), for example, does not so much deny the existence of

lower order factors, but claims that at present they are too fickle

and of such low reliability as to be of little practical value within

a no; nothe is personality system. Moreover, if such scales are of

very low measured reliability, the fractionalization of more general

order variables like extraversion and neuroticisminto sub-factors

is at best arbitrary and at worst propagating the existence of traits

which have no functional independence or substance.

An associated issue which is likely to extenuate the reliability

problem is the fact that estimates have more usually been made for

relatively sophisticated student samples, which are closely cornparab*e

with groups used in the original model development. This is the quest-ion

of the transferability of the reliabilities of the primary source traits

to other, less sophisticated and older groups of subjects. Should there

be a demonstratable loss of reliability across various sub-groups of

adults, there are likely to be serious consequences for the generalisability

of the factor structure.

Ghiselli (1967) refers to this as the role of moderator effects upon

reliability; a test which is reliable or for that matter valid in one

group of the population may not be so for another. The number of

empirical studies, however, on moderator effects have been remarkably

few and the literature appears to contain no large scale research on

moderator effects in personality data. I

7G

The issue of moderator effects is of course crucial to the general

hypothesis stated above that Cattell's model of personality has been

derived from inadequate data bases for a generalised theory. Moreover,

it is likely that the more divorced the particular sub-group under

investigation is from the student population the less transferable

the theory becomes. Thus, the hypotheses have been developed that

the reliabilities of the primary source traits will be lower in .

aiult than in student groups and in the lower socio-economic

and higher age groups of the adult population. It is hypothesised,

on the other hand, that the sex of the subject will have no appreciable

effect upon the reliabilities of the primary source traits.

HYPOTHESES ON THE INTERNAL CONSISTENCY OF THE PRIMARY SOURCE TRAITS

We have seen that the debate over the item structure of Cattell's

primary source traits has been a long and complex one. Cattell (1972) claims

that his items are deliberately heterogeneous in order to maximise

the multiple correlation of the items with the pure factor which the

scale is designed to measure. Whilst many critics (Howarth and Browne, 1971,

Levonian, l961, Eysenck, 1972) have analysed the 16PF by conventional factor

and item analysis methods, none has investigated whether Cattell's specific

claims hold up under closer empirical scrutiny.

Accordingly, the first hypothesis that we shall test is that the primary

source traits will show low intra factor homogeneity of item content.

Assuming that Cattell's factor solution can in some way be re-derived,

we shall also test the hypothesis that items will show high multiple

correlations with their own scale. That is, they will have high factor

71

validity. These investigations and whether estimates of factor purity

can be made through conventional reliability theory, will form the third

major section of this thesis.

ALTERNATIVE MODELS IN TH PERSONALITY DOMAIN

Should it be necessary to reject the hypotheses on factor validity,

`L; - '- -1 ._i L_Lr. ^c J nd p-- , aý nl ).

_ rs; oL. S , planes i ity of 'i r} r- rcni CST

of the personality domain will be investigated. Perhaps the most obvious

alternative model to that of Cattell is Eysenck's (1972) three factor

description of personality. From the information collected in the

various sections of this thesis, it will be possible to test, for example,

hypotheses on Cattell's main anxiety primary source traits to see whether

they have separate identities demonstratable in terms of group differences,

reliabilities and the internal and factor structure.

40

AIMS

The aims of this research can be summarised as follows: -

1. To administer parallel forms of Cattell's personality inventories

to a large sample (N '72000) Qf general population adults

representative in terms of age, region and socio-economic class.

2. To test specific hypothesis on Cattell's model of personality

under three main headings: -

2.1. the verifiability of important group differences in

personality hypothesised from Cattell's theory.

t

72

2.2. the reliability of Cattell's primary source traits.

2.3. the internal consistency and structure of the model.

3. To investigate alternative theoretical models in the personality

domain.

SUMMARY

In summary we are to research in Section 3 of this thesis, the

relationship between the parameters of Cattell's model of personality

and such external factors as sex, age and social class. In section

4 we shall next investigate the precision of measurement afforded

by the model as expressed by Cattell's first order 16PF personality

scales. Finally in Section 5 of this thesis we shall undertake a

number of researches into the question of the internal consistency

of the questionaire items which are defined by and are central to,

Cattell's model of personality at the first order level.

73

N

CV

u H

H

0

H

a 0

U)

H U

z H

w a H U

E/I

Cl z

H

H

w

4

0

a

U

:j -4 4J ý 10

m a) M 4J 111

43 r4 Ö Ö U)

4-3 O C -4 0 d) ß, " "d -P S- to

0 -0 v >

4-4 O

04 S:

`° a 0 u "d

, cö .ý -P P0 - C CO . ri ý },

ll -P a) ' v r-1 N ý4 dO - ý v

Rf U1 E0

d rýö

) C) '-t r-I r-1 ri E

_ý u! b

4-) ri (a u IQ a) ul r 0 ` 'dd

Q) 0 N ri Q) d r 4 -A . "r q ýr p O (n E I 0 -H z v

.ý >

4. to

E O G)

U) p. - la

tue 0

-N S-i U) co 0 14 Ä HR `( 0

O 0ý . c 0 .N U -P "

Q1 43 U -1-) U] ý., 4-) a)

.C I Ö H C ý

H O fl

rd tU +j .c r

P H H (a A . HE-

N

4 v .Q UU] . ºJ W 0 0 0)

-r a) V)

41 rf `° (0ö .0 w s fo - 4

(a ° b z

O-P-P i 0) + 3

S: ý O EO a) (a ý r-A U En u E C t r-4 U t ö E -H -q N C2 4

O to 14 Q) , r r- I to O 4-4 C O aJ

{ QJ U) UN 0) U) (0 0

. Q) 5 Q) -P 4-) U O0 ul ul fO - ?'

O 4 O r SCI

N U) 4i -1 .a4 S4 "ri

0S (tl S -I NO

U) N 4P 0 4- -1 (0 a) u t7 S! (a r. Ul (U ý4

v O. H W4 tttddd v m ", I LI Ui r-i H $4

. f. b -P ,

p. "

to ul 4J

4 S j

04 NÖ (a

.d 'A

(11 -1 C (7i H a -1 rl "9 dO "r1 H IL uO u1 "ýI S-1

to U

N "

N b

C. O O H "i A .ÄL

C

9 0) $4 N rl r. 'D N U)

4-J vO ci. -i X -P

u C U 4- r- IN

-P C ý4

0 U2 3

.CVO S1

4-1 - ,5 0 u ro - ö av .

3 4 , 0 b V to N to Q

-P $4 Ll 4i Cu .d . tJ "rf O r--A ra 0( N o -1 + -P , U c o P -4 ö . 14 . 14 (Uý 0 "

)- PP P r4 -ri ° v -i 5 (L)

a S-l ) ro 0

a F! > i7ß. U$ 44

Q

I

74

r]

3. METHODOLOGY

75

INTRODUCTION

The first methodological problem was that of selecting the sample.

An important aim of the research was to investigate Cattell's

model on adult rather than student groups. It was also clear that

in order to undertake studies on the effect of moderator variables

and of social class and age trends in personality, a relatively

large sample size would be required.

The problems of gaining large representative samples of adults,

however, should not be underestimated. Any sampling procedure of

adults contains some bias. If there are dangers in taking results

from students as if they represented firm conclusions for the general

population, then it is equally problemmatical to infer that military

personnel, subjects known to the experimenter, those employed as

clerks, apprentices, managers or indeed those employed in any Alp

occupational field will represent the archtypical personality type.

It is true that most personality inventories have been developed

on little more than ad hoc occupational, clinical, educational

and student groups (Eysenck, 1964, C6ttelJ, 1968). This may be of

less importance for standardisation purposes in that an author

can go some way to justify his procedures by using the random

error philosophy. That is, whilst all of the constituent groups

will have their own specific bias, these will in some convenient

fashion cancel out when placed in the statistical melting pot.

This is a rather dangerous philosophy if any form of multivariate

analysis is to be used however. Child (1966) and Adcock (1973) i

76

specifically warn us for example against what they have termed

'oddballs' in the sample if factor analytic techniques are to be

employed in data analysis. Should cluster analyses be used,

moreover, it is not difficult to see that a certain structure

might evolve in one sample which would be difficult to replicate

in another.

One possible method was that of requesting the return of completed 16PF

answer sheets from British test users. After a predetermined duration,

samples could then have been stratified and weighted by current

census data. However, although this procedure has proved practicable

in the States and was the principal method employed in the American

16PF standardisation, a considerably lower answer sheet usage

rendered this approach unsuitable in Britain. Added to this was the

unknown nature and degree of bias introduced into questionnaire

data by the test-taking attitudinal and motivational state of the

subject. One serious source of this distortion is direct sabotage.

Another is where the subject assumes, whether consciously or not,

a role congruent with the particular testing circumstances, as,

for example, where an inventory such as the 16PF is used in the

first instance for purposes of vocational guidance and then for

industrial job selection on the same person or population.

Clearly the best recourse would be some form of random sampling

such as that used by the Scottish Council for Educational Research (1965

when standardising the Wechsler Intelligence Scale for Children.

Such a procedure would in fact prove very difficult in practice in

i

77

that even lists such as the electoral register are inevitably

out of date, whilst the costs of tracking down each respondent

would be prohibitive.

An alternative strategy would be to draw up some random or

stratified sample of names and undertake a postal survey.

J-iow. qever, postal surveys ha'. 'e the problem that response rates

(that is the number of returns as a percentage of those contacted)

for completion of relatively large questionnaires can be as low

as 25%. In addition there are the added problems of the quality

of the data: how much can the subject be relied upon to complete

age and social class data or indeed the questionnaire itself

accurately, and can we really be sure that the individual contacted

actually was the one completing the questionnaire? Another

drawback to the method is that using well known personality

inventories in a postal survey might have contravened the

professional and ethical controls on test standards of the

British Psychological Society.

Alternatively, it may have been practicable to test groups of

a

previously selected individuals in centres (schools, church halls,

etc) throughout the British Isles. The major objection to this

method is that difficulty is often encountered in gaining a

representative sample in that females and the lower socio-

economic classes tend to predominate in an obtained response

which is frequently below 10%.

The only solution then was to use some form of quota sampling of 11

78

people in their own homes. Ideally there would be a good representation

of regions, age and social class in the sample. Two forms of Cattell's

Questionnaire 16PF Forms A and B were to be administered under-

controlled conditions and every effort would have to be made to

keep the response rate as high as possible. If any investigation

of item structure and of particular sub-groups within the sample

was to be carried cult (Lo ca L_'gories for the sex of the subject

and five for social class would give ten possible data cells for

example), then it was decided that the sample size would need to

be approximately 2000 cases.

Cattell's Sixteen Personality Factor Questionnaire Forms A and B

each consist of 187 items which on average take 45 minutes each

to complete. Collection of socio-economic demographic and other

information would add about 30 minutes to each interview giving

approximately two hours on average for each subject. A simple

calculation gives an estimate of 4000 interviewing hours for

the survey. With the addition of travelling and preparation time,

it is evident that it would take one experimenter three years

of full-time work to complete the fieldwork.

The alternative was to employ more fieldworkers; hopefully people

who had experience in administering questionnaires and interviewing

techniques. With training, it was thought that such individuals

should be capable of administering the 16PF on an individual basis,

recording the background information and completed forms for data

processing and analysis.

4 F

79

Suitable bodies which could provide such fieldworkers were considered

to be specialist teams of certain research councils, the Office of

Population Census and Surveys and certain market research organisations.

Contacts were first made with the Office of Population Statistics

and Surveys but it soon became clear that a project of this kind

c"; o_tld he outside their reit. Similarly, the SSRC Social Survey

Unit was not principally involved in conducting its own research

surveys. For these reasons, it was decided to employ the British

Market Research Bureau (BMRB) for the fieldwork. This organisation

had the advantage of having a large fieldforce which was experienced

in social surveys of this kind.

The next problem was one of funding the research. The National

Foundation for Educational Research agreed to finance part of the Ar

survey, providing materials, card punching and initial computer

resources. In addition BMRB itself agreed to fund a portion of

the fieldwork costs on the condition that a self-completion

questionnaire on consumer habits could be left with the respondents

after the main survey. The remaining costs were covered by a

number of small donations from commercial organisations which were

willing to make a contribution to work of this kind.

SAMPLING

The sampling frame finally adopted was based on a method developed

by the British Market Research Bureau over a nw-nber of years and t

80

known as Random Location Sampling.

This technique involves the drawing up of a representative sample

of 200 Parliamentary constituencies in England and Wales and those

parts of Scotland south of the Caledonian Canal, selected with

probability proportional to their population size. These

constituencies were selected at rasdon using the currula ive

sum and fixed sampling interval technique, after stratification

by the distribution of votes cast at the 1970 General Election

(as an indicator of economic status) within conurbation or urban

or rural area, within standard region. For each of the 200

constituencies, the polling districts in it were placed in

random order within ward, the wards in random order within

administrative areas and the administrative areas in random order

within constituency. Groups of streets were then selected with

probability proportionate to the electorate in a way which was 40

consistent with polling district. Interviews were then spread

uniformly over these streets.

In random location sampling one is adopting probability techniques

down to and inclusive of the selection of street. This has an

important advantage over traditional quota sampling in that the

freedom of the interviewer to choose her subject is minimised.

In quota sampling, certain controls (sex, age, social class etc)

are to be satisfied, allowing the interviewer freedom to select

subjects in accordance with these constraints. Clearly this was to

be avoided and was one of the principal considerations when

adopting the described sampling technique. I

81

In summary, the methodology was to ensure the collection of data

on approximately 2000 adults in the general population. Ideally

the respondents would be tested in their own homes on the Cattell

inventories, Forms A and B, and the sample would be representative

in terms of geographic region, marital and educational status,

age and socio-economic class.

AP

rr

82

0

4. PROCEDURE

83

THE PILOT STUDIES

Two pilot investigations were undertaken in the lower socio-economic

areas of Ealing and Acton in London, for it was from such areas that

most difficulties were anticipated. Seven subjects were chosen in

each of those districts and the author observed as a female fieldworker

': n 't -hrou:, h the inter`. i ew, ring procedure. ( It should be pointed out

at this stage that female fieldworkers were to be preferred in work

of this kind because many housewives are understandably reluctant

to admit a lone male into their homes. )

The materials were prepared as follows: -

1. A contact sheet containing the initial introduction to the purpose

of the study. (Figure 4.1. )

2. A number of ball point pens.

3. One 16PF Form A and Form B questionnaire printed so that their order

was reversed in alternate booklets.

4. A number of 16PF answer sheets.

5. One serial-numbered background questionnaire (Figure 4.2. ).

A number of problems came to light as a result of these pilot surveys.

The most serious was one of using answer sheets with older subjects or

those with poor eyesight. Accordingly, it was decided to employ the

84

FIGURE : 4.1. CONTACT SHEET

I am from the British Market Research Bureau and we are carrying out a National Research Study of people's interests and attitudes for a Research

Body. We would appreciate it if you would help us in our study.

The research is on people's interests and views and includes a couple of questionnaires. Amongst other things this study will be used to find out the views of various groups of people. Over 2000 men and women all over the country are helping us in this work.

P1 }l0'.: _research

t"( th this questionnaire has been done. 'ten of ner parts of th

.... r . Is the tine jh Eis U en carried Oi in i; 1C1. an

. 'JCO'-lvnd/

Wales (use appropriate country). it is hoped amongst other things that the

results will be of great value in building up a picture of the attitudes of

people in various countries.

ESTIMATED Result (put down CONTACT SOCIAL Questionnaire number Full reasons for

NO. AGE ý SEX LADE if contact successful) any refusal I

r

s

V.

t

85

FIGURE. 4.2. BACKGROUND QUESTIONNAIRE

n ONNA; RE No. (Cols. 2/3/4/5)

BLOCK CAPITALS. PEEASE: -

N. m',, e Coss ti tuency

tPlr. icc t. ite Mr.. Mrs. or Mis. and yive initials)

`ulI t. , taI . ddre, s

Iýterviewer Code no.

Supervisor Time of interview:

OJ-e of inrrrview _1970

Frofl

FAMILY UNIT (related by blood, rarria9e or adoption)

Jab No.

A AEA CODE Tv

To RESPONDENT IS:

- --- ---- -------

Man I I. I:

NORMAL occupational status Marital Has paid job No paid job Status

SEX Aue last

F"II Part-time Ret. from Wid/

Line Re lat ion sF. i 0 birth- t ime

Under F/T Div/

no. to re+;: o. j"ir., t m F ddvv 31 + 8-29 8 job Others S M sep.

1. RESPONDENT

2.

3"

4

i" 8.

11 OTHER MEMBERS OF HOUSEHOLD (not related to respondent)

9.

10.

11.

12.

13.

14.

15.

16.

-------------------- - - OFFICE USE

S. G. A 14.12 g 11

cl 0 c2 I D2 E3

CHILDREN Under I year (14. ) 4

BY AGE I-4 years 5 5-9 years 6 10 - 15 years 7

None 8

RESPONDENT Yes (14. ) 9

fiEAD 0P No A

H-)'JSEHOLD

HOUSEHOLD One 15.12 SIZE Two 11

Three 0 Four I Five or more 2

WEEK, OF One (15. ) 3 INTERVIEW Two 4

Three 5 Four 6 Five 7

VI At what age did you finish

your full time education?

14 or under 16.12 15 11 16 0

17 1 18 2 19 3

20 4 21 - 23 5 24 or more 6

Still studying 7 Refused 9

86

Working 30+ hrs 12.12 Working ß-29 hrs 11 Not working 0

IV RESPONDENT'S AGE:

15 - !9 (12. ) 1 20 - 24 2 75 - 34 3 35 - 44 4 45 - 54 5 55 - 64 6 65 or over 7

V RESPONDENT'S MARITAL STATUS:

Single 13.1 2 Married 1 1 Widowed/divorced!

separated 0

IF 'Sin le' Are you engaged or planning to get married'

Yes No A

IF 'flurried' How long is it

since you were first married? Less than 1 year (13. ) 2 1-4 years 3 5-9 years 4 10 - 19 years 5 20 years or more 6 Refused 7

Ii11 (I3. ) ----------- ---- ------------------------------

VII Do you (or some other member of your household whom you have mentioned) own this house/flat, or rent it, or do you live here rent-free?

Owned 17,12 Rented 1) Rent free 0

IF 'Owned' Is it owned outright or is it being bought with a mortgage or loan?

Owned outright 1 Mortgage/loan 2 Don't know A

IF 'Rented' Is it rented from the council or from someone else?

Council 3 Someone else 4 Don't know A

Viii How long have you (or your family) been living in this house/flat?

Less than I year (17. ) 5 1-4 years 6 5-9 years 7 10 - 19 years $ 20 years or more 9

FIGURE: 4.2. (Cont'd)

jx" uhlch mcn'her of your hot,. ehold Iiviny here: Actually owns it? STATE LINE NUMBER

(FROM HOUSEHOLD

Is responsible for the rent? Q

COMPOSITION BOXES) (i. e. has his/her name on the rent book)

OF THE PERSON Is responsible for this CONCERNED household having it rent free?

x

IF LINE NUM3ER ENTERED IN iX a IS ANY NUMBER FROM 9-16 ASK STATE LINE NUMBER

IX b Which member of your family (i. e. respondent and relatives) living here FROMI is actually responsible for you having this accommodation?

If member of (o. uily unit

TICK 5.! 41": 11 yr detJt. -

iYa 1Xb APPLE

he/she is now occupied (even if temporarily out of his/her present or usual work) occupation

Man or he/she is not now occupied but has means other than single only state pension. etc. his/her former occupation woman he/she is not now occupied and has no means other occupation of senior

than state Dessins. etc* breadwinner` her husband is now occupied (even if temporarily his present or usual occu-

out of work) pation even if he is not a

Married her husband is not now occupied but he has means his former occupation woman other than only state pension, etc.

her husband is not now occupied but he has only occupation of senior state pension, etc. breadwinner'

she has no means other than state pension etc. occupation of senior

, breadwinner' Widowed,

she has means other than (or in addition to) state her husband's present divorced,

pension. etc., and part is obtained from her former occupation if he is alive separated, husband (if he is alive) or as a pension from her or his previous occupation etc., former husband's revious employers (if he is dead) if he is dead woman she has means other than (or in addition to) state her i

pension, etc. but no part of this is obtained present or prev ous

occupation from former husband

The "senior breadwinner" is the oldest full-time occupied male member of the family unit or, if there is no such man, the oldest full-time occupied female member of the family unit. If there are no full-time workers, take the oldest part-time occupied male member of the family unit, or if there is

no such ran, the oldest female part-time worker.

IF "NONE", WRITE "Dependent on state pension" BELOW

xi

Enter in this column occupational details

OCCUPATIONAL DETAILS REQUIRED: of the person indicated above in X

What type-of firm or organisation does/did (........... ) work for?

(STATE TYPE OF FIRM, WHAT FIRM MAKES/DOES, etc. )

What job does/did (...........

actually do?

IF IN CIVIL SERVICE, FORCES, POLICE, etc.

What is his/her rank or grade?

IF "OTHER" Does/did he/she hold any particular position in the organisation? (e. g. foreman, typing supervisor,

office manager, company secretary, etc. )

IF PROPRIETOR CF BUSINESS CR A 25 o. more MA NIAG EA

10 - 24 Roughly how many people work at the place where (......... "") Less than 10

works? (INCLUDE THIS PERSON) (STATE NUMBER-

ASK FOR ALL Has ...........

) any qualifications? (Such as apprenticeships, professional qualificat+ons, university degrees, diplomas, etc. ) (STATE WHAT QUALIFICATIONS HELD)

87

Xli NOW ASSESS S. C.:

AB CI C2 D E

Xiii Have You a TV set In your home?

Yes 18. 12

ýº'ý' No A

IF 'Yes'

Can your set (any of your

sets) receive: Yes No

(i) BBC I A A

(ii) BBC 2 (18. ) 11 A

(iii) ITV 0 A IF 'ITV' On average how 19. 1 or less many hours do 2 you watch 3 independent q TV progrannes 5 each night? 6 or rare

Does your set (any of your sets) receive colour?

Yes (19. ) 9

No A

:. <'w ý.: y` '.? k'ý,

i

16PF booklets only in the main study, subjects recording their preferred

answers by ringing the appropriate choice to each item directly in the

booklet. Although this had the disadvantage of increasing the material

costs considerably,. the important benefits resulted of shortening

response time by about thirty minutes and removing one clerical stage

from the operation in which transcription errors might occur.

As further outcomes of these preliminary inquiries, it was also

established that the following controls would operate in the main

standardisation sample: (1) All testees would be in the 16 to 70

year age range. (2) All testees were to be literate and capable of

reading the booklet for themselves. If required, help was to be

given with occasional words and sentences. In the case of the subject

being unable to read either through illiteracy or severe physical

or mental infirmity, the interview was to be diplomatically terminated.

(3) All testees were to hold a six-month residence qualification in

the British Isles prior to November 1971. This was imposed in order

to exclude overseas visitors and temporary immigrants.

THE TRAINING OF THE MAIN FIELDFORCE

Armed with the information from the pilot studies, certain questions

were made more clear and materials printed and collated for the main

fieldwork.

From editors' ratings of the total B. M. R. B. fieldforce, 200 experienced

fieldworkers were chosen and briefing sessions set up during October

and November 1971 in five main centres: I

88

London: South East England

Bristol: South West and Wales

Birmingham: Midlands and East Anglia

Manchester: Northern England

Glasgow: Scotland

The author took the necessary materials which were arranged in packs

to each of the interviewer briefing sessions by road. The written

ins __tctions prepared or ±ach interviewer are given below. At

each centre the interviewers were carefully taken through these

instructions step by step.

INTERVIEWERS' INSTRtLTIONS PREPARED FOR THE BRIEFINGS

Written instructions were prepared at the briefing sessions and a

copy provided for each of the fieldworkers. These are reproduced

on the following pages and can be distinguished by a vertical line

drawn down the side of the page. º

INTERVIEWERS' INSTRUCTIONS

Together with these instructions you should receive:

Three contact sheets Ten pens One red biro One questionnaire booklet containing Form A and Form B

(serial numbered) Twelve answer sheets Form A and twelve Form B Twelve questionnaires (serial, numbered)

Purpose of the Study

Basically, the aim of this study is to explore the interests and

attitudes of various types of people. You will probably find that,

unlike most surveys, this study may well arouse people's curiosity

so it will be as well if you yourself have as full an understanding as

possible into what this is all about. You will find a long Introduction

89

for you to use when speaking to the respondents at the top of your

Contact Sheets. This explains that:

1. This is a National Research Study and not a Market Research

Survey.

2. The Research Body is interested in people's opinions,

attitudes and interests. The questionnaire is a standard

inventory used in the social sciences and the main purpose

of the project is to get British data on it.

3. Similar studies have already been done in other parts of

the world (e. g. America, Europe) but have never been done

here before.

4. The results of this Study will help the Research Body

amongst other things to build up a picture of what people AP

are like in different countries.

By explaining this to your respondents, you will probably find that

they are interested and willing to help. It is important that the

response rate is kept as high as possible.

Contacting Respondents

You are to contact ten adults above the age of 15 within the streets

with which you have been issued, namely:

5 men

5 women of whom 3 should be housewives, 1 should be a working houswife and 1 an other woman

90

A six months' residence qualification in the UK is to apply to the

sample.

We would like to know about every contact you make who is eligible:

so please use your contact sheet every time. For those contacts who

refuse to take part put down their estimated aqe and social class,

? tnei. _asors for rcý, usal. For yo_. -- successful contacts wri `c°

down their questionnaire number in the space provided.

When introducing and discussing this interview do remember always

to use the word study and never survey - this is very important.

Each interview will take about an hour and a half, so in order to

be able to complete the interview, it is important that you do not

mislead respondents as to how long it will take. Remind him/her

that all the information collected is strictly confidential. 40

Before starting the interview, do check that the booklet and the

verbal questionnaire have the same serial number on them. If this is

not the case then always alter the numbers to the number on the verbal

questionnaire.

The Form of the Interview

The interviewing procedure is really quite straightforward:

1. The interview starts with you giving the respondent the booklet,

instructing him/her to read the first page, then turn over and

start answering the questions using the pen to ring the answers.

It may be as well to remind the respondent that all their answers

91

are kept absolutely confidential.

2. When the respondent reaches the half-way point in the booklet, to

give him/her a short break, you will ask him/her the questions in

section I (only) on the verbal questionnaire.

3. Thurn, as:. the respondent to return to the booklet, first reading

the instructions and then answering the questions as before.

4. Administer section 2 and 3 of the verbal questionnaire, remembering

to use the Introduction at the top of section 2. Then collect

classification details.

N. B. 5. Please write the serial number that is on the verbal questionnaire

onto the front cover of the self completion booklet before handing 40

it to the respondent.

The Questionnaire

The questionnaire is very straightforward and only a few points need

to be made: -

1. Remember to check whether the respondent is married before asking

Q. 1,2,3.

2. Q. 2. We want to know about all children under 21. If the respondent

has no children under 21, then do not ask Q. 2 (b) or (c).

3

92

3. Only ask Q. 3. of married women under about 40. If you find the

wording of this question rather abrupt in some cases, then you

may alter it to something like 'Are you likely to be having a

baby in the near future? '.

4. Q. 6. If the respondent wears contact lenses and spectacles, then

only ring the code 32.11 for contact lenses, because this is

mainly what we are interested in.

5. Q. 7. Here we are interested in birth order i. e. whether the

respondent is the first, second child etc. If the respondent's

brothers or sisters are deceased, we are still interested in the

answer to the question.

6. Remember to check respondent's age if he is male before asking Q. 9. 0

7. Section I ends after Q. 9. At this point instruct respondent to

return to the booklet and at the end of the second half of the

booklet, continue with Q. 10.

The Booklets and Answer Sheets

1. Some of you will have the booklets stapled in the order A, B, and

the others in the order B, A. This order must not be changed in

any way.

2. As mentioned earlier, the respondents will be asked to answer these

questions in the booklet itself by ringing the most appropriate

letter, 'a', 'b' or 'c'.

93

3. Whilst the respondent is answering the questions:

(a) Give help with instructions if necessary, but do not look over

the respondent's shoulder as they answer - guess:

(b) Stress that they are to work quickly, giving the first natural

answer as it comes to them, and tell them not to spend time

pondering over questions.

(c) If the respondent wishes to change an answer, get him or her

to clearly cross out their incorrect answer with the red biro.

(d) If the respondent asks for advice, merely say 'it is up to you,

or make some other equally noncommittal remark.

4. After the respondent has answered the first questionnaire, and y . lst

he or she is completing the second one, check that the respondent

has given ONE and ONLY ONE answer to each question. On completion

of the second questionnaire check this one also. If there are any

instances where more than one answer has been given, refer the

question back to the respondent and ask them which answer they

finally wish to decide on, and remember to cross out the incorrect

replies with red biro.

5. Recoding of respondent's completed questionnaires

(a) For every Form A and Form B questionnaire there is a corresponding

answer sheet. You are to recode the information from each

questionnaire onto the PINK answer sheet for Form A and the

BLUE answer sheet for Form B.

94

(b) Follow this procedure

First of all write the serial number that is on the front of

the booklet onto both answer sheets, A and B.

Then at the bottom of Form B write whether the respondent

is male or female. The appropriate word should go at the

bottom right-hand corner beneath the large column of empty

boxes.

Now recode page by page. One page of the questionnaire

corresponds to one complete column on the answer sheet.

Transfer the answers from the questionnaire to the answer

sheet carefully, ensuring that you are transferring the

respondent's answer into the correct little answer box 'a',

'b' or 'c' and that at all times the number of the question gº

corresponds to the number on the answer sheet.

N. B. A constant check on this is that whenever you turn a page

in the questionnaire the answer to that first question should

always go in the first set of answer boxes at the top of a new

column.

When all answers have been transferred and you are quite

satisfied that there are no errors due to putting an answer

in the wrong set of boxes for that question, please GO BACK

AGAIN AND RECHECK THE ENTIRE RECODING.

When completed, questionnaires and answer sheets will be

inspected for accuracy before analysis.

95

(c) It is suggested that the recoding of answers be carried out,

whilst the cuestionnaires are being completed by the next

respondent at the following interview.

(d) If you make a mistake cross out in red biro - if the answer

sheet is a total mess (eg missing out one question, leaving all

Cl _e.:. ions out , ý7r up icate.

6. Gen egal Points

(a) If the respondent is unable to read, diplomatically terminate

the interview. Give assistance on odd words. Tolerate slow

readers.

(b) In the questionnaire, the first question refers-to 'test'.

If this is questioned, please explain that there are no rigIft

or wrong answers and that the questionnaire is not a test in

any sense. It is unfortunate wording.

(c) If the respondent completes the first questionnaire, but

refuses the second, please explain the importance of this study.

If necessary, ask if you could come back at a more convenient

time. (Under no circumstances should you agreed to leave the

booklet and return for it when completed later - you must always

be present when the questionnaires are being answered. ) If the

respondent still refuses to complete the second booklet, the

first booklet and answer sheet should still be submitted, with a

bold note across the unused answer sheet stating 'Refused to

Complete'. But collect classification information if you can.

96

(d) There is a chance although a very unlikely one, that the

respondent may recognise the questions. If he is obviously

familiar with the materials, then merely say that you are of

course involved with the standardisation of the questionnaire

in the United Kingdom.

this stage the interviewers were actually put through the 16PF

questionnaire themselves. This was to ensure that they would be

familiar with the item content of the inventories.

INSTRUCTIONS ON COMPLETING THE QUESTIONNAIRES

In order that you can become fully familiar with the materials you

are going to use, I am going to ask you to complete the questionnaires

yourselves.

0

HAND OUT THE. QUESTIONNAIRES

Please don't get the materials confused.

Now, some of you will have a white page uppermost arid others a

blue or pink page uppermost. If you have a white page uppermost,

will you please sit on the right hand side of the room; if you

have a blue or pink page uppermost, please sit on the left hand side.

ALL MOVE

Thankyou, now I will deal with this side of the room and Miss Bodger

with the other side.

97

You all have a white page uppermost, will you please answer the

questions in the first booklet by ringing the best answer for you

in the booklet itself.

O. K.

? d1 ii `tOU ri. ): L , ý5e r=: a `_? hit lY Lr,, ct1. ii1J dill, :. 'I read 1

aloud.

READ

What to do:

Inside this booklet are some questions to see what attitudes and interests J

you have. There are no 'right' and 'wrong' answers because everyone has

the right to his own views.

Vr

There are three possible answers to each question. Please ring the letter

'a' or 'b' or 'c' to indicate your answer choice. Remember that you should

have one answer for each question.

When you answer please keep these four points in mind:

1. You are asked not to spend time pondering. Give the first,

natural answer as it comes to you. Of course, the questions

are too short to give you all the particulars you would sometimes

like to have. For instance, one of the questions asks you about

'team games' and you might prefer football to cricket. But you

are to reply 'for the average game', or to strike an average in

98

situations of the kind stated. Give the best answer you can

at a rate not slower than five or six a minute.

2. Try not to fall back on the middle, 'uncertain' answers

except when the answer at either end is really impossible

for you - perhaps once every four or five questions.

3. Be sure not to miss anything out but answer every question,

somehow. Some may not apply to you very well, but give your

best guess. Some may seem personal; but remember that the

booklets are kept confidential.

4. Answer as honestly as possible what is true of you.

Do not merely mark what seems 'the right thing to say'.

Please ask the interviewer - in this case that's me - if Ap

anything is not clear.

O. K. ? Any questions?

Will you please write at the bottom of the white page the time

you are starting this first questionnaire.

PAUSE

When you have finished you should also put the time on the white

instruction page.

I

99

It you have any questions, please ask theam now. Once you have

started, please don't confer with your neighbours.

PAUSE

Fine - Would you like to begin now, remember to work quickly

g; vn Your :E _oL natura answer.

Check that all is going well. After twelve minutes say'Work'as

quickly as possible. When the interviewers have finished, remind

them to put the time they finished if they have not already done

so and check.

Fine - Now we will do the second form. This one however, you

will be answering on the answer sheet provided. Would you like i

to now read the instructions on the next page and I'll read them

aloud.

READ

(From outside cover of 16PF booklet)

Well, thank you all for your help. Now I am going to ask you

to recode the answers which you put in the booklet, which you

completed first, onto an answer sheet. We will now hand out the

relevant answer sheets.

PAUSE WHILST ANSWER SHEETS HANDED OUT

Follow the recommendations in your briefing instructions - this

will stop errors creeping in.

loo

Recode page by page. One page of the questionnaire corresponds

to one complete column on the answer sheet. Transfer the

answers from the questionnaire to the answer sheet carefully,

ensuring that you are transferring the respondent's answer

into the correct little answer box 'a', 'b' or 'c' and that

at all times the number of the question corresponds to the

on the ? nswer sheet.

A constant check on this is that whenever you turn a page

in the questionnaire the answer to that first question should

always go in the first set of answer boxes at the top of a

new column.

When all answers have been transferred and you are quite

satisfied that there are no errors due to putting an answer

in the wrong set of boxes for that question, please GO BACK

AGAIN AND RECHFCK THE ENTIRE RECODING.

TIME

Go round check on the Interviewers' method of recoding.

TIME WHEN RECODING COMPLETED

40

Fine, now you have each done one Form of the questionnaire by

marking your answers in the booklet and one Form by marking your

answers on an answer sheet. In the pilot study which we carried

out we found that too many respondents were finding difficulty in

101

using the answer sheet. For this reason all respondents will be

answering by using the booklets for ringing their answers 'a', 'b'

or 'c'. Now you will recode these answers onto an answer sheet for

each respondent. We want you to do this so it will be done

accurately, the respondents might make mistakes if using an

answer sheet.

Remember to recode as laid down in the instructions. I will go

over them.

READ

CONCLUDING REMARKS

Are there any outstanding points? As was stated in the introduction,

this is a very important study and this the first time that it has

been carried out in Britain. The usefulness of the results carer ?t

be emphasised enough and, of course, good results now depend on you

and your colleagues.

THE BACKGROUND QUESTIONNAIRE

In addition to the personality questionnaire completed by the subject,

the interviewer was required to complete a background questionnaire

for each respondent (Figure 4.2. ). Information on the major questions

in this part of the survey are as follows: -

(a) Standard Region

Apart from the East Midlands and East Anglia which were treated

as one unit, the ten geographic regions employed in the study were I

102

those of the Census General. The total adult group

was made up of testings conducted in the following forty-eight

British counties: -

Northern England: County Durham, the North Riding of Yorkshire

and Northumberland.

Yorkshire and Humberside: The East and West Ridings of Yorkshire and Lincolnshire.

East Midlands and East Anglia: Derbyshire, Leicestershire, Northamptonshire, Nottinghamshire and Cambridgeshire, Norfolk and Suffolk.

West Midlands: Shropshire, Staffordshire and Warwickshire. --

Greater London: Testings were conducted in a total of 30 Parliamentary constituencies within the

administrative boundaries of the Greater London Council.

South East England: Bedfordshire, Berkshire, Buckinghamshire, Essex, Hampshire, Hertfordshire, Kent, Oxfordshire, Surrey and Sussex.

North West England: Cheshire and Lancashire.

South West England: Cornwall, Devon, Dorset, Gloucestershire, Somerset and Wiltshire.

Scotland: Aberdeenshire, Angus, Dumfriesshire, Dunbartonshire, East Lothian, Fifeshire, Lanarkshire, Midlothian, Perthshire and Renfrewshire.

Wales: Brecknockshire, Denbighshire, Glamorgan, Monmouthshire and Radnorshire.

103

(b) Social Grade

Rather than rely on information such as economic standing or

educational level, which undoubtedly affect social class in

varying degrees but which are often difficult to assess objectively,

social grading has been based upon the occupation of the head of

tee family unit as the single factor most responsible for the

social status of the family. For this reason detailed information

was collected on the job, position, type and size of organisation,

and any professional qualifications of the senior breadwinner.

Social class was then carefully assessed by way of a detailed

guide of which the following serves as an illustration.

Social Grade A: The Upper Middle Class

Professional and Semi-Professional: Doctor, Solicitor, Senior 40

Civil Servant (Principal and above), University Professor,

Chartered Accountant, Airline Pilot.

Commerce and Industry: Proprietor, Senior Executive, Company

Director and Manager of large business (25 or more in establishment),

Senior Industrial Scientist.

Police and Armed Forces: Superintendent, Chief Constable, Lieutenant

Colonel, Wing Commander, Commander (R. N. ) and above.

Social Grade B: The Middle Class

Professional and Semi-Professional: School Teacher, University

Lecturer, Executive Grade Civil Servant, Journalist.

Commerce and Industry: Owner of medium size business or shop, r

Managerof small branch office of bank or insurance company

104

(10-24 in establishment). Manager of a department in a large

factory or business.

Police and Armed Forces: Inspector and Commissioned Officers in

the Armed Forces.

Social Grade Cl: The Lower Middle Class

Professional and Semi-Professional: Junior Civil Servant, Bank or

Local Government Clerk, Pharmacist, all Students other than

Apprentices.

Commerce and Industry: Lower Managerial and Supervisory Grades,

Clerk, Typist, Laboratory Technician.

Social Grade C2: The Skilled Working Class

All skilled manual workers: Foreman (manual), Deputy (mining), sip

Chargehand, Overlooker, Overseer (manual), Superintendent (manual),

Supervisor (manual).

Social Grade D: The Semi-Skilled and Unskilled Working Class

Most semi-skilled and unskilled workers. Most adult female manual

workers. Unskilled assistants, labourers and mates of the

occupations in grade C2, all apprentices to skilled trades.

Social Grade E: Those at the Lowest Level of Subsistence

Social grade E is assigned only to those households which are

dependent on state pensions etc. and where no member of the

household is working. i

105

(c) Marital Status

Marital status was classified by one of four categories:

(a) Married, (b) Widowed, divorced or separated, (c) Single and

engaged, (d) Single but not engaged.

^1: Working Hours

Information was gained on whether the respondent was working

full-time, part-time or not working at the time of the study.

(e) Terminal Education Age

The age at which the subject completed full-time education

was established. If a respondent had worked for a period

between uninterrupted education, the most recent age was coded.

w

(f) Head of Household

The head of household was defined as the member who owns or is

responsible for the accommodation occupied by members of the

household. Suitable criteria were laid down to ensure correct

evaluation in the event of unusual household composition.

(g) A2

Subjects were required to give their age in years at their last

birthday; if this was refused, as was the case in approximately

20 instances, age was estimated and then coded by one of seven

intervals. i jr

106

FIGURE: 4.3.

THE SAMPLING UNITS AND STANDARD REGIONS OF THE 16PF

BRITISH STANDARDISATION SAMPLE (Nm 2227)

i,

107

Once all the forms had been received, edited, scored and card punched,

the data viere initially set up on the NFER IBM 1130 computer system

which was to undertake the first stage analyses. The more sophisticated

analyses at the second stage were to be undertaken by a much larger

computer installation at the Edinburgh Regional Computer Centre on

an IBM computer.

DATA PROCESSING

As subjects were tested the completed questionnaires were returned

for editing and coding of the social class data. Checks were also

made on the coding of responses from booklets to answer sheets and

those with complete and correct data were passed for handscoring,

cardpunching and further editing by computer.

When training the interviewers, the importance of control over the

quality of data had been stressed and in fact the vast majority of

scripts received for editing were complete and accurately coded.

In twenty-one cases either only Form A or Form B of the 16PF had

been answered by the respondent but after editing, results for the

complete sections were sent for card punching.

As is always the case in fieldwork of this size, odd questions had

been omitted by subjects but rather than reject the entire record,

information was filled in where possible or simply left blank.

(Computer programs had been adapted to deal with this and the associated

problem of variable sample sizes in the data analysis. ) i

108

0

5. DESCRIPTION OF THE SAMPLE

109

w

FI

Ni

z U H

Cq H Q C 4 Q L

c-ý (J.

G; t.

f"N

W

E-+

E V C'' N0 IO

iC C' itD 'N Ic")

13 co co C) !M tom ; G? J]

'

I

-: J C , t

14 ä Cl! MN C7 lN O (O 11ý

"I`- UO t0 jq IG716 ; tD ' T. IN Ito j CO "o

-. 1 u- CL

I( I!

I

f

"m tD co CO L7

icA

M tý t7 tD Iq

ici IW iV

ISO IM CO 'l,

C

CL t° 0 CC) cri cD sm C7) !v 1- t, i

U nE

bM C) 7 III N

IM

to m6 aö :. n co ri Its ei U'ý

C

Q (j C

OO 10 to

i a, co in CO

I r,

117

tD Qi M G) "7 Ir NI CO I OO ; I. -

CI I r ýl cI co rx 11 ý ý; JvL

LTJ WU Cy C

r9Cý W lo ý-pü aNi I c '- .

f-,;

O I-Hi.. U N p'r, ý

, Ö: 6 öÖ p 'ý

1 ; I C) W , ýý

z} ,ý (7 V) Z (f) Cn

aýs _ QI 73 r

- n -0 0 C) C CO

''J V -O

n n

0; N NOC . p) C. _

U

ýCýWZfncn

1 i0

M ;ýN IN

(>J Ir L) "6

r- iN 'C'7 iN

c lci IM Nay

NC lw

r- NMN

LO In 7 -7 N C) N'CD

Ir IV IN

O) AGO'C: / IN

c N G1

ö

Q. 3

W

0 n< 4J10

a o

Ux

F- -

cD ICD iO c0

ca i. -

ij

N Co CO

Ln is cD

co jr": 1 v,

r IM N

N

Cl) , _-o ra

gý

N ; iC) ci N

ca a. Q

Vl

'D

°o c iV±

a

c9 C) a) x

ýý C

X cr.

r

I t

f

Cn !- iC,

.- .r Iq

}ý iN

4

i` N IfJ

IN

V7 iý C'

1

öio

Ö cý ýi ýN C- OO'IC

L7) C CfC

.1 .Y OCOp

CL S, o ,,

CD C) C) x cý m

W.

z 0 H H lzýc L H

'7 Q d!

r

", i V) 41 O &)

UH

H

f

Fwý ý,

C) ö. E

r- <o - o

C) ö. C to

C OO

C1 ýi

C) O

? fin.

O Q E R

_C 7

C) a 00 EL

0-i i C, 4 LO (n

MTM

C)ý11I ;5a -mai W0 N 1121 t: ýi

Ü u7 '7 `ýio'j

CL

CD c, ý Y, ýJ

ýJ ý`

ýC

ýýý M

f e-

iý iN

i

ýýl ýi

i

M

cam, l0Z

in

NN (/)

co N O IM

ý

0ý Lo

ico T

T cn

I re '0

I Cl?

N, - N

N I(14 O 04 C14

Nib m_

Ico

N CO c6

c 6

O M

CO N N

ý; 6? OCl) O O O NN (V

N CO L7 .- GO V ýL)

uý

{

N

p

y

ö Cu (1,

}}} C

} }

o N Cl) { co N

? 1

In L7 Ln L7 zr NM Ln a_

0L LN N

Q C

J - O

'o ä mm_Q l

, ýý ý u' ýJ U U co

l--__ ý

crvü, zc j .2 n<n

O

Ü CO

CO O a-O

O C N G E D

öi v c

0) U rn

i

n

i

3ý

7ý

r

111

DESCRIPTION OF THE SAMPLE CHARACTERISTICS

A total of 2,227 subjects cooperated in the study. From an analysis

of the contact sheets it was found that the response rate was

approximately 78% and it was evident from interviewers' estimates

that those refusing tended to be female, older and of lower socio-

C_C than `> it ioipcants in the research.

The first computer run involved counts and percentages on the main

demographic and socio-economic variables contained in the background

questionnaire. These are represented in Table 5. l.. The population

estimates in the table come either from the Registrar General's

Annual Estimates and the National Readership Survey or from the

Target Group Index -a national survey of some 25,000 adults.

40 Before proceeding to consider the results in more detail, it is

important to remember that this sample specifically excluded the

illiterate, more seriously infirm and older subjects (70 years

and more). This is not the case with the population estimates

which can only be regarded as the best available statistics and not

the most relevant. The sample and population estimates do not share

identical parameters and direct comparison on certain variables

could be misleading. Whilst bias due to non-response is also

involved here, the greater proportion in the sample of those of

higher social grade for example, reflects to some extent the

controls used in the procedure.

it can be seen from Table 5.1. that in terms of region the sample I

112

seems relatively representative nationally with the maximum discrepancy

between sample and population estimates of 1.1%. With regard to social

group, the sample has a somewhat higher proportion of social class A,

B and C1 subjects but this is to be expected where older subjects

were excluded from the sampling frame.

mI! -re is also a relatively good representation of marital status although

the Widowed/Divorced/Separated category is a little low. Similarly,

by reference to the table, it is apparent that working hours, terminal

education age and age statistics show effects which are probably

due to the controls adopted in sampling mentioned above.

THE PROBLEM OF NON-RESPONSE

Alk

A dilemma which besets all researches of this kind and which must surely

make the exact measurement of the modal national adult personality an

unattainable goal, is the unknown degree and nature of bias introduced

into results by those who refuse to cooperate. From the very outset the

important practical consideration was recognised that adults have to be

willing to take part in research of this nature and that without full

cooperation it is impossible to apply true random sampling. It was

for this reason that great care was exercised not only in the phrasing

of the introductory remarks to the study, but also in the method of

recording responses, the provision of suitable incentives in the form of

a premium bond prize, the stressing of the confidentiality of results,

and freedom for the subject to be interviewed by appointment. Moreover,

the experience and known proficiency of the selected interviewers for =

113

gaining cooperation and their patience and persuasiveness in ensuring

that a session which was anything between 30 minutes and 3i- hours long

was not prematurely terminated, were also of crucial importance.

Although the refusal rate of 28% is somewhat higher than might have

been expected with particular subsets of the population, which are

,. s--:? 1', v rKo. _c e ;: dL av rec en 'r ýý "a1 3b I E? d? 1ý Lv? "c U_"r =i "ý? O'"1SE'ti: ^ýCý^r s (: r.

school 'leavers, for example), it is highly probable that this figure

is to some small extent spuriously high as a result of contacts,

ineligible for the reasons of illiteracy or age, declining assistance

before their ineligibility could be established by the interviewer.

As occurred in 21 of 2,227 cases in the present study, even on having

entered the home of the subject there remains the further danger that

the session will be terminated before completion. This issue of non- 40

response is of course enduring in studies of this type. Short of

providing fieldworkers With revolvers, an action which would secure

entirety but not necessarily accuracy of response, the researcher's

recourse can only be to efficient incentives whether they be social,

monetary or idealistic.

In all it is thought that despite certain limitations on response rate,

the sample will serve as a good data base to test the hypotheses

outlined in a previous section. Indeed knowing the problems of

collecting data from adult groups with an ab initio research design,

Cattell (1972) wrote " ... not even in America itself, where research

support is apt to be more lavish, has any one of these standardisations

quite reached the closeness to the ideal model that has now been }

achieved for Britain in this undertaking".

114

SECTION 3:

GR0UPDIFFERENCESAND

THEPRIMARYS0URCETRAITS

40

6. CROSS. -NATIONAL AND SEX DIFFERENCES

115

INTRODUCTION

Once data have been punched and verified, computer analysis proceeded

ui ti-I the calculation of descriptive statistics for the primary and

secondary source traits on each 16PF form combination (A, B, A+B)

according to the particular classifying variable under consideration.

with sixteen first order, eight second order factors and four derived

criteria measures, three questionnaire form combinations, three groups

for sex (males, females and both sexes combined), six age groups

and seven socio-economic classes, it is clear that an enormous volume.

of data was generated even at this early stage of the study.

The first step was to produce counts, frequency distributions,

means and standard deviations for each of the primary source traits

(16) by form combination (3) and by sex (3) giving a total of 144 tables.

These data were then converted by hand to the standard score system'O

used by Cattell, ie. the n-sten. This system relates raw scores to

a mean of 5.5 and a standard deviation of 2. Hence, standard scores

range from 1 to 10, from which arises the name 'sten' (standard ten).

Specifically, the n-sten is determined by using a normalised area

transformation: that is to say, by imposing a normal distribution

with the required mean and standard deviation in the reference

population. This procedure is discussed by Tatsuoka (1969) in

'Selected Topics in Advanced Statistics'.

Following the hand calculation of n-stens (these data are held

separately from the main thesis), a computer program was written to

automate the calculations and further tables were computed for separate

age and social class groups. These tables are contained in Appendix 6.1

116

In the case of cross-national and sex differences, 't' tests were then

calculated for each factor and these data will be incorporated into the

main body of the text. It is important to note that these results deal

with statistical significance as hypothesised in Cattell's model; they

do not of course necessarily imply practical significance for questionnaire

users.

There is the further problem that Cattell's primary source traits

are oblique factors, that is to say, there are non-zero correlations

between the scales. For instance, we may find that the difference

between the sexes on Factor C in our sample is of about the same order

as the difference on Factor Q4. Yet it may be that the difference found

is simply reflected in both sets of scores because of their inter-

correlation. It is important to note, therefore, that when scales are to

some extent inter-dependent, the use of multiple 't' test tends to

exaggerate underlying differences. But our main concern is as much

with the characteristics of individual scales as with the characteristics

of the sample. Whilst it would be misleading to suggest that multiple

't' test results give an accurate picture of the extent of the.

differences between various groups, it is certainly the case that they

give a good account of the characteristics of the individual scale.

The 't' tests which follow in the various sections are two-tailed and

adopt the conventional . 05 probability level or beyond as the standard

of statistical significance. It should also be pointed out that about

5% of the differences between means can be expected to be statistically

significant purely by chance, although this is unlikely in any way to

seriously distort our findings.

117

CROSS-NATIONAL DIFFERENCES ON THE PRIMARY SOURCE TRAITS

Because the British groups are of somewhat older mean age than the

American data provided by Cattell (1971)(mean 30 years), it was not

possible to test for cross-national differences without first having

taken account of age discrepancies. Accordingly, to bring all mean

factor scores to a common age, the British group means were corrected

on the quedratic model provided by Cattell (ly? 1)(see beloýý).

(a) Female Cross-National Differences: Results

After the application of the age correction weights, it is evident

from Table 6. l. that the British females were significantly higher

than American females on Factors L (p -:: L-. 01) ,N

(p c. 05) ,

Q2 (p< . 001) and Q4 (p G . 01) and significantly lower on Factors

G (p0.001), H (p<. 001), I (p C. 05) and M (p<. 001).

Differences were not significant on the remaining eight factors.

Then the British female group are described. as more suspicious (L+),

shrewd (N+), self-sufficient (Q2+) and tense (Q4+) and less

conscientious (G-), venturesome (H), tender-minded (I-) and

imaginative (N-) than the American female group.

(b) Male Cross-National Differences: Results

In Table 6.2. the respective American and British male groups are

compared. Eleven of the sixteen source traits are in fact

significantly different at the 5% level or above: British males

being higher on Factors B (p ` . 001), E (p C . 05), L (p-< . 001),

Q1 (pC . 001), Q2 (p<----. 00l) and Q3 (p . 01) and lower on

118

A (p C . 001), G (p x. 001), Z (p x. 001), M (P", _ . 001) and

O (p. < . 001).

These results give a picture of group differences of the British

male being more intelligent (B-), assertive. (E+), suspicious (L+),

experimenting (Al+), self-sufficient (Q2+) and controlled (Q3+)

than the average for. American males. The British male group had

significantly lower scores in terms of outgoingness (A-),

conscientiousness (G-), tough-mindedness (I-), imagination (M-)

and apprehension (0-).

SEX DIFFERENCES ON THE PRIMARY SOURCE TRAITS

The means and standard deviations for the male and female samples

on Forms A+B of the 16PF are to be found in Table 6.3.. Fourteen of

these differences are significant at the . 01 level or beyond, only

Factors F and G do not show statistically significant differences

between the sexes. Males had significantly higher means on Factors

B, C, E, H, L, M, Q1, Q2 and Q3 and females were significantly higher on

A, I, N, O and Q4.

These personality differences give a picture of the female group

being more tender-minded (1+), emotionally labile (C-, 0+ and Q4+),

conservative (Q1-) and less dominant (E. -) than the male group.

DISCUSSION OF THE CROSS-NATIONAL AND SEX DIFFERENCES IN THE PRIMARY SOURCE TRAITS

Results for the cross-national comparisons on the primary source traits

support the hypothesis of no difference between British and American

119

groups on Factors B (intelligence), C (emotional stability),

0 (apprehension) and Q1 (experimenting) in females and Factors C

(emotional stability), N (shr&v. dness) and Q4 (tense) in males.

In addition British females scoring lower than the comparable

American group on Factor Ni (imagination) and British males being

lower on Factors A (outgoing), NI (imagination) and higher on Q3

sc-1-conc? )4- cor trot ) ere all differences in . P. ? hyno hesi sed

direction.

Departures from hypothesis usually involved statistically

significant differences in results where none had been envisaged;

only Factor E (dominance) in males with the British rather than the

American group scoring more highly showed the opposite tendency

to that expected.

Thus the hypothesis that the British group would show a more introverted

personality pattern is only partially supported by the data. Whilst

differences on Factors A (British less outgoing) and Q3 (British more.

controlled) were in the predicted direction, no significant differences

were found on Factors F (happy-go-lucky) and H (social boldness). As

we have noted above, one major extroversion component in males,

Factor E (assertiveness), even showed the opposite trend, with the

British group scoring higher.

Factors which show consistent differences between the American and

British groups are G (Americans higher on superego strength), I

(Americans more tender-minded), M (Americans more imaginative) and

Q2 (British more self-sufficient). These source traits do not group

120

together in any major second order factor such as extraversion or

anxiety nor are they factors like E, C, O and Q4 which are particularly

susceptible to the social desirability response sets (Cattell 1973)

and so it would seem less likely that these personality differences

merely reflect dissimilar research methodologies.

.. ýý.. cu, h and other primary source traits reach statistical

significance,, Ae must not lose sight of the fact that the groups in

terms of mean differences are reasonably close. The largest difference

in Z-score units is . 37 for males on Factor Q2 which still represents

considerable overlap between the two curves. Certainly most of these

differences would have little effect on the standard score tables

(n-stens) used in profiling results.

What is probably more important to the reliability analysis in the "

next section is the apparent greater dispersion of factor scores in

the Arerican female group. Fourteen of the sixteen primary factors

show higher variance in the latter group (sign test, p-<-'. 05).

The mean scores being relatively close suggests a higher incidence

of more extreme scores in the American female sample. This may well

have its origin in Cattell's method of collecting data by pooling

results from distinct and less typical sub-groups of the population.

It could also be true that variations in the motivational set of his

samples is at work here. For example, the large differences between

variances on the emotional stability factor (C) may have its origin

in a response set for those tested under client centred conditions

(eg. therapy) to fake low and those involved in organisational

121

decision making (eg. job applicants) to fake high. Combining such

groups would tend to polarise the data, perhaps having little effect

upon the central point but increasing dispersion about the mean.

As a concluding remark, it is important to remember that because of

the nature of the American data, too much confidence should not be

? ', ä Gi SV Ci oss-r Flt.:. cr _ý ?: _ ýl "7cß'1:! j3 (19-C.. )

1S

pointed out if it is not really possible to specify the exact modal

personality of a nation, then it is not possible to talk definitively

of personality differences between nations. Inferences regarding

cross-national differences can only make sense within the context

of sample characteristics, the research design and the conditions

of testing. We may conclude, however, though differences are often

statistically significant, their actual size is usually small and

for our purposes, we do not appear to have particularly unusual

data with respect to Cattell's. n

Turing to the sex differences on the primary source traits, ten of

the factors showed differences in line with those hypothesised.

These were namely Factors A (females more outgoing), E (males

more assertive), H (males more venturesome), I (females more tender-

minded), L (males more suspicious), *O (females more apprehensive),

Q1 (males more experimenting), Q3 (males more socially controlled),

Q4 (females more tense) and F (no statistically significant difference

between the sexes). The hypothesis of no difference between the sexes,

however, must be rejected in the case of Factors B (intelligence),

C (emotional stability) and Q2 (self-sufficiency), where males

achieved a significantly higher mean score than females. Moreover,

122

although the mean of Factor G (conscientious) was higher for the

fer; aie group, the difference failed to reach statistical significance.

Finally, on Factor M (imagination), with British males higher and

on Factor N, with British females higher, reversals of the expected

sex differences in personality are evident.

r_ st: nele , r: ýup scoring significantly ; nigher on average than

the female sample on Factor B (intelligence) is perhaps a little

surprising, especially in view of the fact that the females were

of some hat lower mean age than the male respondents in the study.

Reference to Cattell's (1973) norm supplements shows that in just

about every case, the American male and female groups have identical

means and standard deviations on this factor. This suggests that

Cattell uses this same trait as a classifying variable when

compiling his tables and, if this is so, these data call into

question the validity of the method. 0

The sex differences on the primary source traits were in fact consistent

across both Forms A and B, except for Factors G (males higher on Form A,

females higher on Form B), N (females higher on Form A, no significant

difference on Form A, males higher on Form B). Why these factors

should fluctuate in such a manner is not immediately apparent. It

may be significant, however, that as measured by coefficients of

equivalence between Forms A and B, Factors G, N and Q2 are three

of the least reliable source traits (Cattell et al, 1970).

It should be remembered that these sex differences are based upon

groups of mean age approximately 38 years. The nature and extent

123

of these differences can only be generalised to other ages with

great caution. There is evidence that the 16PF personality factor

age changes in the American culture at least, show dissimilar

patterns for the sexes, (Cattell, Eber and Tatsuoka, 1970).

For example, crystallized intelligence demonstrated a slight

fall in women on Forms A+B in the 1961-62 edition of the 16PF

^h er^. i . ̀-din for r males. al The . "'yamestL. ion o `Jerson iy is not d_sc :

... '

differences between the sexes at various points in the age

distribution will be taken up later in this chapter.

As concluding remarks, we can say that the British sex differences

on the primary source traits are roughly in line with Cattell's

model and results reported in the united States. The differences

between males and females in personality were more predictable

than the cross-national trends, although this could have been due

in part to comparing data from studies which used very different

AP

methodologies. It is certainly true to say that from these results,

the sex of subjects matters far more in understanding personality

than their nationality.

4

124

4I

H

I..

0 M

W U

0 EA

A W

:n

O Nm N m N V

N

O

NO N

OI N

Q Q

4 f0

m R

p Q °

\G

ö R can v cv

N ° ° r�

Z

O MR

N

( N

M I°

I

Q

p

-

V N

1

S

R^

' NV

N Jn

ý

Q _' N

Q

M Q

{

Qf R

O CJ

m O

Q

6

O

J

V L7

Q

O

C

M

ýA ;

ö

O

V ý

N O

u'f Q N

;n O 1

N ý�

m C7 f")

MO N

Oý tD

L! i N

OJ V

O

H ^

M

00

V

LL

W

R m 1_

N

O^ mO

t0

m C7 U

ýJ

lLi

Ct

(

C"ý

R

N N

R Q

M

N

°

N

R

v

M

O

O

r

p

O

ýI

N

V

V

G

öR

v OV

!JIQ

-I

i,

r

ÖI

r. ý

C]

"-

"

M M °

N

0

v

°

°

o

ci

[h _

Z

9qO Q mi N

I 6

fn 0..

G_>

CS C q C2 �

C G1

Q

v

I ! [1 G 1 CI '. ' V.

a

12ý

(V

W

W 4

C7 ß.

0

0

H

H E-4 H

z u

H

w Uz W a w w

Q

H ý-7 4 z 0 vý w a h Q U H HIS

H ýl

j-ý

UNý

14-4

O M

0

H U7

ýI

V O

OU V

p

t7

_ N _

cý c Of

T

OU N

C1 p

1

d

^ ps

v co NSW ýJ

7 ýQ ý- p

or .r `i

ý ^m N .. < °' 'Q ö

C 10

C7 N(ý

dU

NU

O

O 1

i

IS u CD _

` OO : ") r

_

MO

0

O

O

hfl 7

QN 4

L7

U N

O

I

I

O

ý

pO

ýn

O Ca

U

R-

v ', (O

:np

m

CJ

O 1

I

"

t7

'f N, m

Oa Lfl

T lý1

"mal ý.

O1 (7

O

O

9uG Cý h

O ICN

.nZ

©f G

yu

01

v

WO I'ý O

V

O

O O

jv

cJ

N

O O

V

n! O OO cc v

cý V

O Uý

VZ Ü

N Lý

O_

NV

(. l

U9 O nO

C., `i

6

UO O

QV

C

V

O

Ci

?I

ýJ

i

126

rý: a C1 H

z gW

ti

z

H

ro

r-1 H

V)

C) c O

C: r ,-

_ O

I

O

Ö to p `i

.n Ci : tt!

R O

c? N 'ý

d m QQ N r, ý ýr/

n V 'r

0 - `

r; n t

u ö N j

V

ý ý! r! c UO

N N V

n o2 ö LS N 7

tO q fý O

b M

ý

r

N

dO

(ý

r M

O m

C. 7 ý

Q N

ýr

LLý tý N

OH O

LL N NrV tL- . y V OG O

W O M! C ', C7 hO

V O

J ý rv 'n M ýO

o dNv N a ýJ

n c c

- c-i ^n ý M! 1/

`

m' NN , rý O

Q N O UG

N i v

L O y CO

L to

C `c

v=

C1 o Y c+

VN

jo

127

N rn x x

O vý v7 ý

ý:: ( ý H 1

m H

C. )

WI

H'I

W

Qli

ö

z CA V)

Hi

HI U2 f Ga

z

ri Q

CLI Q Ul U) V] (/l

z

z z z z

b v) vý U) Z z z

Q z ¢ w w

U U1 ", -I Vý W vý "d o)

E-1 a E, rý c a c a . U) ä äw ä C). ! )l

c

s4 (u 0 v "r-, a

-rA c ä 41 ö ä) ö 0 ý

W,

4 t)

.C 0 tr s4

.0 r1 0 to

0 "HI U

-P u 4

. .H 4 -1 i

V) C "rl U 4-i 0 rf "ri >1 "d CP 4-i

r-1 (' "d rl I ( a

u a a "d ä 0 0 Q ) 0. bý

- C 0 0) co

+3 U ul 0) N u1 "'. 4 "H eH

$4 -P v rs 9

vi d cn i ýý z

128

Aw

7. SOCIAL CLASS DIFFERENCES

129

INTRODUCTION

In an earlier section hypotheses were developed from Cattell's model

on personality differences between socio-economic classes. In this

study subjects were assigned to social groups classified on a similar

basis to that used by the Census General; these categories have been

described above.

Linear relationships have been hypothesised for the majority of the

primary source traits and such trends will be investigated by way of

conventional product moment coefficients. However, to test the

hypothesis that the regression of personality on social class is

non-linear, eta coefficients will be calculated on the data.

The correlation ratio or eta (Y) is a general index of association

particularly useful when analysing data in which there is a curved

regression. Eta allows us to predict on the basis of array means

and can be defined as the proportion of variance in the first variable

explained by and associated with the second. Eta assumes no particular

type of functional relationship between two variables and it can be

regarded as the maximum size of correlation between them. Like Linear

correlation, two regression lines are found for eta. However, these

may differ in shape and slope and so two eta coefficients can be

calculated:

dl y d

y

Eta for the regression y on x

130

N

N c4 H

W a H

a ö

0

H U)

a

a

a

0

0

z

z

a

H

Ö

Nx w

U) 0 ww

U

H Hq

U

O O 0) 1-4 %D N N I -i 'cr M ko ;: r N C) N i- CO 00 O O O. r-1 . -1 O O O

N M m l0 00 r-r O. N O CO 01 I- O 1.0

O "

r-1 O 9 O H O "

O O.

I 1 1 I 1 I 1 1 t

Cl v. co N CO lo off" S H CO M, l0 Lfl . N- N H r-1 O O O O O O O O O O

IT f'7 M: r CO t- IT N CO 110 % ö ON ý C ö e-1 r . -1

. C)

. C)

. H

p

" 1 1 1 t "

i

H H H O ri CV O r

Ö r i

CV N lJ Ol kD ('') ri fl O lfl tf) O ON L) f"1 CO Ln N O O O O O O- O O O

1 I

l0 N 01 Ln CV) ßl O C) LV) to C) 01 H" 01 tlý O i- N N N N r-1 N N, N N

1 1 ! 1 1 1 1 1 t

C) ýo ko (A r- C14 co 4w in C%j C) (n IV I MY en r- N (D O H r-I r-4 Cl O C) O j

O ý ko ON En L V r'r O . 0'. co

O O O. r "

tý . -r O O O

1 I i I 1 I 1 I" I

H CO %. 0 H O i -1 - O rI N r-{ r 1 - r-1

1 1" I 1 I" t "I 1ý

O in C) CO M C14 Ln c

In . 00 N C) Ö

O O O O O O. C) C)

1 1 1 1 I 1 t

' %D M lfl CO N co O r-r '-1 O O O O

" O O O O C)

i 1 1 t 1 N C

dý Ol H O co H CO "Cr r- A 43) - tD

" . Q, O O O O C) C) 1 1

H r Ln CV) CA C) 00 CO r 1 ß -I C) N ß in ý

O ri H N° 1 1" S 1 I I 1 t I"

00 Ö CO - 1` CO (W) %D H C*4 N f"1 l'M M

N CV)

M

1 I " 1 I I 1 I" U)

H M C . 1 - r-1 rl LO N

Ö CV)

Ö

O O O O O O O Cl

N O

In ýy^

co gOct

cr, h: ýv d aý ý: }

131

d1 Eta for the regression of x on y

Q x

where d1= standard deviation of the yl values predicted from x

d1= standard deviation of the x1 values x rr¬dIcted from y

d and d= standard deviations of the total distributions yx

We shall be concerned with the first of these, testing the hypothesis

that the regression of y (personality) on social class (x) is linear.

If we wish to test for the linearity of regression, Guilford (1973) has

proposed an F test based on an analysis of variance approach where only

a knowledge of eta and the product moment correlation coefficient (r) t

is required.

r2) (N - K)

(1- 2) (K - 2)

where K= number of columns or rows

SOCIAL CLASS DIFFERENCES: RESULTS

The product moment correlation coefficients between Cattell's primary

source traits and social class are contained in Table 7.1.. Because of

computer restrictions, these data are based on slightly reduced group

sizes in that only subjects with complete data records were included

in the analysis. The social class variable was coded by assigning the

i

132

S N

E-4

oa + d

O

y C1 b

Gý+

d' M pp rý t0 U1

O " z

co 0 I ri

" I

O "

Ol CO N N O O .N tt

a. z

rn cM 0 0 ýn ul

z O

co co C- O Q M cp

z z

N N r. Q

O N

Ln O ý-1 . -t rt ßl1 O

N CO O

O

rn rn ý

x o to

. LO Ln CD , -i o 0 0

o LO

` ö rn ö , -ý w O N

Ö d p ul

N N y

U " ö z t

t` c+1 M ul

z O N t 1

M iý

z

u ý v ýry! W

4-) u :: s u 41 -r4

4J -q v 44

44 ri

1J 4Ou 04 0.6 : E: U

133

character 1 to the highest social group (A) and the character 6 to the

lowest group (E). The number scale bears an inverse relationship,

therefore, to the social class variable. Those coefficients significant

at the 5M level or beyond are marked in the table. Appendix 7 also

contains the means and standard deviations on the primary source traits

for each social class, 16PF form combination and sex.

Turning to the product moment correlations, which of course are appropriate

only when investigating linear trends in data, it is evident that there

are sizeable correlations between social class and the primary source

traits of B (intelligence) and M (imagination). Two other factors

which show consistent associations. across form and sex are C (emotional

stability) and 0 (worrying, troubled). In addition, there are relationships

in the female group with Factors H (social boldness), I (tender-minded),

L (suspicious) and Q4 (tense). These findings are relatively consistent 40

across the two questionnaire forms although Ql (experimenting) seems to

be related on Form B but perhaps in the opposite direction in Form A.

As to the direction of these relationships, higher social class in general

is related positively with Factors B, C, H, M, I and Q3 and negatively

with Factors L, Q4 and 0. This gives the picture of members of the

'higher' socio-economic groupsas more intelligent, emotionally stable,

imaginative and controlled and less apprehensive. Additionally, females

in the higher socio-economic groups appear more socially bold, tender-

minded and less tense and suspicious.

Eta coefficients, representing the regression of the primary source

traits (Forms A+ B) on socio-economic groups are contained in Table 7.2.

134

0 Mý

N co H

a 0 6 H

H

8 En

Ö

En H H

En U)

aW Cry

0 Lv

N

N U') 0 rl H 0 0 O

O M

co Lr) -4 Cf) r1 O M 9 a I M

ul N 0

N O O O V1

a r-I ýD d'

H O r-I V]

ff i N

z

N N tp

O z N

0 1 d Ö

N

O 0 0 M M 0

z i z N m

O 0 01 a z

p ö ö -i 0

d'

O ö x " z

a ö H i N

e CC) 4 r1 H O lp ° " N

In fV W V " " " z

N

M

" ý

Ö 0 O ° `n Q . z N

U U C ý.

C N u

ý U (a 4-1 U

4 j ` V v - w to ä 8 ý' '-

4-4 (a 0 t o. ääßü U)

135

With regard to the data for females, the F values suggest that six

of the 16 factors show significant departures from linearity. These

are the primary source traits of F,, G, I, L, M and Q3. The corresponding

analysis on-the male data (Table 7.3. ) shows significant F values for

: 'actors ß, 1;,, F, C, I, N, Q3 and Q4.

SOCIAL CLASS DIFFEhENN; CES: DISCUSSION

The largest coefficient is between social class and Factor B (intelligence).

so as the cu t-egori_ attion of social group is based upon the

occupation of silo hoadý of the family unit, this is perhaps not unexpec'ýed.

As is evident from the correlation matrix reproduced in Table 7.4.,

social class is also positively related to the amount of full time

education received by the subject, which clearly is likely to have ý.

some effect on a verbally orientated ability scale like Cattell's Factor B.

Another primary trait with a significant relationship with social class,

is Cattell's Factor M (practical vs. imaginative). This variable shows

a significant departure from linearity and reference to the group means

and standard deviations (Appendix 7) suggests that the regression line

tails off slightly at the extremes with the maximum acceleration taking

place between social groups Cl (the lower middle class) and C2 (the

skilled working class). It is interesting that it is precisely this

>I I; ýti':. _ch _ I'csec' _ divisic be t. en the professional or 'ideas'

jots or 'thing' orientated occupations. These

results are in line, therefore, with the initial hypothesis that those

in the more practical occupations would score lower on this factor.

136

TABLE : 7.4.

CORRELATIONS BETWEEN AGE, EDUCATIONAL AGE AND SOCIAL CLASS

IN THE 16PF FORMS A+B ADULT SAMPLE: BOTH SEXES

(N = 2012)

AGE EDUCATIONAL AGE SOCIAL CLASS

AGE 1 -. 41** -. 01

EDUCATIONAL AGE 1 -. 37**

SOCIAL CLASS 1

40 correlation significant at . 01 level or above

40

I 137

Three other trends in the data which support initial hypotheses all

concern the major anxiety variables in the 16PF with the higher social groups

being more emotionally stable (C+), less apprehensive (0-) and, in females,

less tense (Q4-). Only Q4 for the male adults showed evidence of a non-linear

relationship . In addition tender-mindedness (I) in females showed a

significant correlation with social class, which again supported the

ot'r. e. is of a slight. positive relationship.

In all, results supported or tentatively supported hypotheses on some

ten of Cattell's primaries: -

FACTOR

A Outgoing

B Intelligence

C Emotional Stability

E Assertive

H Social Boldness

I Tender-minded

L Suspicious

M Imagination

O Apprehension

Q4 Tense

SOCIAL CLASS DIFFERENCE

No differences

Positive relationship


No differences

Positive relationship - (females) No differences (males)

No differences (males) Positive relationship (females)

Negative relationship (females)

, No differences (males)


Negative relationship

No differences (males) Negative relationship (females)

Of the six unsubstantiated hypotheses, probably the most serious

concerned the Factor N (shrewdness) where no clear relationship with

social class was found despite Cattell's contention that this trait

is closely associated with social status.

I

138

a

$. AGE DIFFERENCES

139

IN'T'RODUCTION

Cattel? (1972) points out that the curves of age trends in personality

need to be determined in order to make proper comparisons among

individuals of different ages. Beyond this consideration, reliable

information about the development of personality traits is also

fundamental for personality for as Ang]eitner (1971) has shown, the

age dependency of personality scores shows in every culture yet examined.

Cattell analyses age curves as the sum of endogeneous and exogeneous

components. The former describes the effects of internally determined

a maturation and involution and is virtually synonymous with a pure

genetic component as it manifests itself over time. The latter is

made up of all the various averaged environmental influences on the

individual and can be divided again into an ecogenic, representing

what happens to all people as an average across all historical human

ecologies and an epogenic component expressing the effect peculiar to

an epoch and a given culture.

Methodologies for separating the ecogenic components have been discussed

by Baltes (1968), Cattell (1970) and Schaie and Strother (1968) in terms

of comparing longitudinal (the follow-up of age cohorts with re-testing)

with cross-sectional data (the study of different age groups at one

point in time). However, Cattell's conclusions on age trends in

140

FIGURE-' 8 . 1.

ANALYSIS OF PERSONALITY AGE CURVES

ENDOGENOUS

genetic ccuponent)

EXOGENOUS

(nurture)

DCOGENIC

generalised effects across all human ecologies

Innate maturational and involutional influences on personality

Environmental influences on personality

EPOGENIC

41

effects peculiar to an epoch and culture

a

I 141

personality remain based on simple cross-sectional data where averages

on the primary source traits are available for a particular decade and

for groups of people at different age levels. These curves thus

confound the ecogenic with the epogenic components but are,

nevertheless, important as a first step in understanding age

differences in the primary source traits. Moreover, in that both

Iris Judy and Cattell's conclusions are based on a cross-sectional

design, it becomes possible to test for the similarity of age trends

in the primary source traits.

AGE DIFFERENCES: RESULTS FOR PRODUCT MOMENT AND ETA COEFFICIENTS

An indication of the hypothesised age trends in personality was

given in an earlier section describing the primary source traits.

Cattell (1972) has more recently provided a picture of the curves 0

for the primaries across an age range from about 5 to 60 years.

These curves are represented in n-sten units about the total group

mean and standard deviation. Two associated problems arise in the

comparison of Cattell's trend lines with those collected here.

First, Cattell's curves are for a different age base. In order

to ensure comparability, it will be necessary, therefore, to

restrict our considerations to changes over the adult years,

16-60 years. Secondly, this change in age range will need a

corresponding adjustment to the group means and standard deviations

upon which Cattell's standard score system rests.

}

142

* Eý

6 5 C4 w a w

a U) a

4 X H N

Ö w Ey H

9

a

w

4' I) N

0 10

0 10

0 M

O to

O CC)

O OD

O . -f

O CO

O tO

O In

In N

m (V

o O

C'J " t0

" I! 1 U1 V; If) If) ltd tý l0 l0 l0 N

" l4

" N

to In 0 0 0 0 In o 0 to 1n O 0 0 cn fV (V U) 0 O\ CO N In O CV N O N O Cf .

d' "

4l .

to .

t0 .

to If) .

d ..

trl r

lf) ul LI) LI) N

N In N

LO N

O to

11 N

Ln N

u1 N

O 0

10 IN

0 0

O 1-1 EN

o 0 lf)

cV C5

O O

i! ) to N It) to ul Lý I) 10 lD 11 V In N

{ o to

0 In

O In

0 In

0 Un

0 Le)

0 1n

0 In

0 V)

0 LI)

o II)

O In

0 to

0 O

Q . " . . . . . . . " " . "

N In I! ) I[1 In IA LA L1) In LA ý Ifl N

O !O

l0

O O

to

O U)

It)

I; N

If)

0 to

Ul

0 Lf)

CO

0

.n N

u1 Ca

N

0 O

N

i'') 5

l0

O Gý

10

u) CV

N

N d'

lD

O U

N

ýJ N

to

N

to

in

N

to

U)

0

lß

N

CO

0

tO 10

11

l0

t11

c0

O

N

0

E

N

l0

O

N

If) N

"

0 Oý

"

O u)

" to

0 ul

" In

0 0

" to

0 0

" d'

It) N

"

O In

"

O O

. ul

O O

" U)

to N

"

0

r 0

N N

"

p

" 0

N

ºa " %D

" 10

tin "

it)

d' "

to

I

" u)

" tO We

" Ott

0

. In

00) e

ýi'

O "

If)

0

" t[1

W "

0

O r

(%4

H O O

in

O r-i

In

0 in

in

tn N

in

0 O

l0

0 rt

l0

0 10

N

0 O

co

'0 H

a)

LI) C\t

O

0 0

co

O 0)

N

a0 N

to

0 p

N

x O d'

IT

O fý

It

0 tf

tn

in N

II)

In N

d'

0 to

cM

O 0

d'

O H

d'

0 c

d'

U) N

dý

0 0

. t'

0 c0

M

t[1 M

dý

N

N

%r

0 Ln d'

tin

in W0

to In

to

N

to

In

d' S

d' 0

In

N

to

0

to

0

to

0

u)

o

N

0

t0

N

u1

(D 1

in

in

0 0

to

d'

d'

u0'ß

c0

0

t0

)

to

N 4

N O)

M

0 O pip

.4

0

N

W O p

" In

O d'

" 10

O in

" U)

0 H

" U)

to N

"

0 N

"

O G0

" M

O 0

" 44

O CO

"

O It)

" M

O , -.!

" CO

LO N

" N

d' N

"

O O

" tV

0 p

d'

O U1

d'

O it)

ui

0 U)

(n

O N

tn

0 to

d'

0 CO

M

0 01

en

0 d" '

d'

0 mal'

d'

0 0

00

c*>

co

"

o

" N

O 0

" tIt

0 tO

" U)

0 o "

to

0 0

" In

O in

" [

0 "

d'

0 0

" ul

0 10

" to

O LO

r In

O 0

V"

0 LI d" '

0 0

ý3" '

N 0%

<! " '

0 0 N

AC Lý d'

O to

tý

to

LI)

to in

tO 0 U)

O I'll) U1

O 0

l0

O I lQ

001

U'1 If)

M

IA

p

N

C7 4

N r-i

N N

0 en

0 d'

0 u1

0 to r-I N CO d' U) cß t11

L S3'TVW

1 S3'IVw33 1

143

4 N

Wa

cn

Eý N a

6 A W N4

W',

N O H

W

a U W 4

13

ör M

W

O U

O d' O

" ('M '

" M I

"

N Ids

"

S IM "

N IN "

11 m N

"

r d'

"

o', N

"

O H

"

dý "

-i In "

M N "

O O

" tß N

m a

I 0 H d' m OM I d' (ý 'rA O O

0 0 0

" U1 N

Ný N O

"

N O

"

N 0

"

N 0

"

N 0

"

N O

"

lD I) "

N O

"

d' . -I

"

41 ,

"

CV O

" r-1 rI

"

tV N

"

0

" L7 N

Q E; O 0 O 0 0 O O O 0 0 0 0

0

tr! 0 Oi

U) (V

0 O +N O H i l)

cri ct

CV O

N LO

01 M

N N

d r

N cV

G M

N ý"

p

l0 N

z Ian I0

"

(M R "

IcM+1 .

I. -or) 1

"

Ö

" IrI

" O

co

" . -i

" ýT

"

(f) "

N

"

0

"

H lD N

d N

O O

O' M

M M

d' '!

t0 (M d'

IN r1 I CN I

d! , -{

4V O

l0 IM N N

0 O

[P CV

a r-1 dý r-i N

0 I

d' I . -ý Ol (O w

M M v O

ON N

O ýM d' M IO ýCf 0

ý! '3 N

m GO

d' (d0

d' I lfl

N I to

01 IM

Iq IM

rl d' rl lD

W lfl

d' N

r-I $p

W to

N O

D N

M 0

CC) . -ý

CD Ln d'

0 N

M I d'

ap I . -1 M ' . -1

M O

to IO 'M M O

N

O a $cn

" ýN

"

N

"

N C.

N

" H

" IN

" " '0

" O ce)

. O I0

"

04 O

" " Cl

Lf N

ý d1-1 '

M O IN Iw0 to to

M 10 t- IM ON $ '1 'Itt dl co OD 0 O

d' N

W M

" " 1O

" " N

" IO

" tV

" IH

"

(c'

" IcV

" Iul

" IN"

"N 0

0 N

V N Ö N N M

O ' IN IO !O N $M d' O

d' N

O "

M

"

N

" O 1w

"

r-4 IN "

ýD IV'

"

lq O "

M

" N

.

d' 0

"

H N f" tD

d' "

N M

"

O 0

" dý N

M ý Ö Ö (rl I'l Iý co Ö

cy) M (ý

m N ýp O

M O

u1 N

C7 4

tý r-1

N N

0 M

O d'

O In

O 0

N . -i

N N

0 M

0 d'

0 t! 1

0 t0

N C)

A "

s-nvw 11 5iVM3 I 144

Accordingly, a grid was prepared to facilitate the taking of readings

in n-seen units from Cattell's (1972) figures at the age points of 17,

22,30,40,50 and 60 years. These data were recorded for males and

females separately, treated as raw scores and then converted to total

group standard score units with a mean of .0

and a standard deviation

of 1 (z scores). It was these values (Tables8. sand g. 21 which were

plotted on the graphs contained in riqLtres 8.2. to 8.17. (blue lines

for males and red for females).

In a similar fashion the subjects of this study were grouped into one

of seven age cohorts with mean ages of 17,22,30,40,50,60 and 68

years. Raw score means and standard deviations were then computed

for each factor and age group by sex by 16PF form (A, B and A+B).

The 16PF Form A+B data for each sex were then converted to z score

units based on the total group (male + female) means and standard

deviation for each primary source trait (Appendix 8). Figures 8.2. to 8.17. 40

contain the results of plotting out these statistics in graphical form.

This scaling procedure allows comparisons to be made between those

personality differences predicted by Cattell and the results obtained

for the samples of male (N= : ='=l000) and female (N=1000) subjects in

this study. In addition, comparisons are now possible across the two

questionnaire forms as well as the different primary source traits.

The product moment correlation coefficients representing the linear

association between the primary source traits and the chronological

age of the respondents in the study are represented in Table 8.3.

Statistically significant linear negative trends are discernible on

Factors B, E, F, L, Ql and Q4, whilst Factors G, N., Q2 and Q3 show I

145

9

.8

.7

.6

.5

---- .4

.3

_ .2

H

0 o 0

FIGURE 8.2.

AGE TRENDS IN PERSONALITY:

FACTOR A

Ö \\

L

- rý

5I__.

-. 6

-. 7

-. 8

146

.8 t G: ý TRýi jS :N_ : iýs'JidiLI'_

-

14 FACTOR B

.6

ý- '-5I.

.3

1ýXý n

20 30 40 50 60 68

-,. _.. ý (YEARS)

Male Form, A --

----- 147- - --_"- .. __,

.8

. .7

- .6

.5

i1GU:,,.. 8.4.

AGE TRZIII)s

FLC oR C

/

0 O -- -----.... .

ý- -- --_ -'ý .4 --

- -. 7_

20 30 40 5O 60 68

AGE OL-EARS) F0 -r 7,1 ~- -- -_. _

Kiiiie Form.. A Form, B

148 -- --.

ýý... _ _--- ";

--_ _. _.. _.

ý __ _

__:.:.. _ _

__ `^

-, . ý.

.9

_8

.7

.6

.5

x ý.

x

FIGU 8.5.

FEGE TRENDS IN PL Gi. ý : ITY:

FACTOR E

ýý "`-yam _ ...

ý` `.

IS,

Ü

-- - -- N--. 1 - -. - -. ,x

4-

_ -- - 20 30 40 SO 60 "- - 6G

AGE NEARS) Temcl Farre F. ------ - -- -- psm F

.9 FIGUR"- 8.6.

$ AGB TRT-' DS IN PERSCNALIT' :

FACI'CR F

.7

.6 "\

14

-. 5 .

-. 6

I- ---

ý; - -. 7___.

\

\- ---ý

20 30 40 50 60 62

. _.. . . _. _... tiGE (17EARS) ... - ...

Fe. °-Gle Form 'h Form E

-1'ä1e Form Ä---------- " ... _. _ý. _... -- -- --- -ý--------

ý_. _ýý------ ---- __ --- - Form--B ..... _. ----

-...

.9 FIGURE 8.7.

-7 { FACTOR G .

.6

-. 5- ýc

.4

- ýý

-I-/ /1'

2aý

/

-__. _.. -.. xoe

ai . _..

-- ------- k/7 JbKc

ý. -. 7

20 30 40 50 60 68

AGE (: '£ARS) Y _.. _

-- _. _. __. _-_. _. _.. _ "---- c1E FC--7r, A

J_ 'ra Form. P-,

15-1

.9

.ß,

. 7_

.6

-. 5

FIGURE 8.8.

FACTOR H

a ýKý. _. ý_... _? c

.. _. X ý_ ýý

20 30 - 40 50 60 68

ýý -- __. --_- kC-E (YE; -. R£) Fend er crin A

Orr,

'' '--_ -- ---. -- --- --- ___ -- - .. Form- $

G

.B r

.%

.6

i

.Sý..

0 U Cl)

---- ; -. 4

-- --p

/

/

2 30 40 50 60 66

AGE (YLF, RS) F

- --- - --Made Form. A-_-

Jý. _...... _ _ý ._ -- F orni B --

FIGURE 8.9.

AGE TREK IS IN PERSONALI_- :

.9

.8i

.7

.6

.S

rx

/I, "-

.8 20 30 40 50 60 Ea

; G£ (YEARS).. pemcle Forrn i, Yon. h

-- -- _ ____- -- leF orii r------- Fern- B _.. _

.. _. __. 2.. 4

FICURE 8.10.

AGL IN PERSGNA_, IT'_ .

FP C^CR L

.ö

ý_

.7

.6

.5.

C. 8.11.

. =GL ; READS IN jTY.

FACTOR M

-4 - - TT It I_

_ '\ _

/ ýI

" - ý\- - '-- ------ --2-- . -

ý f. _ ------ -- -

- - i T t ý Y

----- 8----. 2-- -.

- .. _:. z. 3. -- - -- ---- - -- - ..... -_--.... -. ___..

__

- -. 6 -

20 30 40 50 60 68

" -- AG= (YEARS) F'e111c1 EF p271 A "Form

-_ _` -- --`- -- - --- - ---

ýý. _

^ý. _. --- -^ _Mae.

Form

.9 FIGURE 8.12.

_8

. .7

.6

AGE TREI DS IN P. ERSONALrY:

FACTOR N

5-

8__.

20 30 40 50 60 65

"-_. - ---- .-- jiGE (ice; RE) gi-ec1E Form i.

. --_. ''- -- ---___ Form F-

--- - -- Mal e . Form A - -"-_ -- "-- -- --. _. ---. ý.. _------------ __.. _. ----- ______

---- - ---- --- -,.. _ ý. -F crni B"--

.7

.6

.%

.6

.5

. -'1G; -F rö. 13 .

. GL TflrS IN ptisSOicALIT:

7z

N

20 30 40 50 60 68

AGE (Y-kRs) 1 elrýýle }'orr l:

_- -= --- ---... Fcrr,

\-ý -- - . _.. __- __-- ----

-- -- --. -- - Male Form--A--- -- ---- ý1

- ... -ý

---- _- Form B-"-., -

_ Ecu: X3.14 .

.8 ý\\ TAGE TRENDS I PERSONALITY:

FACTOR Q1

.7 \

-- --- . -- - - -

20 30 40 50 60 68

7. GE (YE; --RS) F'err cý E f'orr,, t, -

Form-B

"9 f FIGURE 8.15.

-8

+I t Gr i iu:,; ýJL5 I1: FýrcSONALi X.

FACTOR Q2

.7

.6

ý-- _---. 2 -- -. -...

ý- -6

20 40 30 50 60 68

- ýOY'Tt i AGE (YMRS) }orn. h -ý

fir` -----_ ___---------- _- -- --- -- --- ----------- -- ___. _.. _ _. ____. _ ; _.

------" - -- - --- ý Form .. B - -- --------- - --- ---

.9 rIC E 8.16.

.8 AG TRENDS IN FERSOXA. -I I'.

FACTOR Q3

.7

i

.6

_ -r

o

q, -

20 30 40 50 CIO 68

_ J, GE (YEARS) Fe ae Torn. A ---- _. _. -,

Form Ä

Q

.8

.%

.6

x __ .s

ýý g. 17.

AGE. TRENDS 1IN

FAC^"'C, R Q4

T"4 x"\ -T::. _. _3 x

__. _--

E, -- __ --_- -- -

o r 0 V i

20 30 40 50 60 6E

ý .. - -- AGE (YEARS) Female Form i;

`- - -- - _ } orri ao

w... - :- -- ---- - ---- Form

r Ct

c-+

N 0

0

0

z

X

H

r, '

Ow 0

AI L

z

w

CY)

HHV

I co 4 c, to o V)

I. D rl C3

t f co f{ C,

}1 -a o.

1i

ch tie -, crn .t I r-4

cý1 1 --4 LO N '+. " c7

ell I

04 co C*4 (7 N M N N N `

i 1 1 :' t

in %D O

ýIý

d' c'7 O "i t

to in O

1'

C) N O

i

"cT O

f.

N ýp O

t.

N Oý O

%D O

!t

CO Q

i" .

-+ tJD Ln Cß M N

N in Get

N CO) N CV N N N N

r4 V-4 ai r- n co . co

I ("

O O r-4 M - ö Q O n O O O O O ,

ý 1_ i" ý" 1 1ý

to "N a% M O N Co N l1w m OD co C3 r- (IN O Ö

r-1 . - O O rl O C) .4 " I 1ý - 1- 1. 1" 1 1" t 1

to ' Gl p C) "C' to to C%j M l1ý ! "ý "ý ch . N O ý4 ý- O r+ O O O

ý O'

i ! [- C) rn in w M 03

° in to ý ö ä ö ö ö o .

40 f- CV) Cý 04 410 cy) Ln N Lti LO LO C N C -1 N M N

M C '-)

co Gl O co /`') O N C"I O Xn (V . rl E? 1

in C') c" M ý

ýi`

1 1ý !f 1f f 1

"I Co . -# to t- M O N l0 -ý . to to n . 'W " t- at in N " -t N C-1 O . -ý r" 1 C)

" . -ý

t It 1 i i t F

fi C '' a in C) O r-f a C14 4 O O O O O O IO

" aJ O. O O

1 1 I 1 t ! 1

-co M l7 (N (n Co M N- CO

r4 N r -f "-! EEE i t" i i i i i ý I

N f) (D tf) M (n Ln l'ý E'er . -4 Ll) C' N lfl MC) .s CJ. , N O O Q O O O. O O Q

tý p

"-i t t! 1

Iv C 7 4rß

ti

162

W t7 4 Ö

W

O

R

U) Eel

w a4 U) C :4

W

ä Q) 0

ýýw

Q1 h

o

M M

N CO N

N : O lp z

C

Co M O

O N

ri O N C! ]

C) rn o co N M N vi

z z

0 o rn vý 0

r-1 r-I" CO O O M

H O d' to H " " " z

r-i

ý N M O N

x z

M ý'ý1 lp V!

r-A cm C%

w z r-,

rn C(I

w .. z d1 O

M O

M M ý

r1 r N N pý Ln O

1 N

0 Ln 0 L

ý, ar

v (J

tu p4 ý, ;2

0 4-1

tu >

ääzu 0 14

ý. ccn

163

0

En El H

9 H

ä

cA a+

.. N

W0

c'ý wý

`" " 0 z 1

M N cý O

" " " CY)

00 N "4 M O

M

M M lO VI ý5 " " " z,

a O -I

N l0 t0 O O N Z

0

M rl Y) M In bj

z z '--ý

O N N

ý+ o O z o . -+

to r1 i L v)

z 0

N N

ýý " " O Z

01) LO N 0 A

" " "

V

O

d' Ci O CD

r- (D co U-1 d' d' O

w O N

N N In N O

" " " " O N

N c CO O O 01 y

U " O" " z

C\l CV

OD 1 M

ý

CU " " 9

O M

N d' O O U)

4 " ö z

4-J 0)

(a z 4J Nu +' :3 U w 4 a s

w -1 a 04 - 0 4)

ää Ü cU)

164

significant positive relationships. The remaining six primary source

traits (A, C, H, I, M and 0) show no clear linear pattern.

In order to test the hypothesis that the regression of personality

on age is linear, eta coefficients were calculated on the primary

source trait means and standard deviations, classified by age and

con, ained in the appendix.

The eta coefficients representing the regression of personality on

age for Forms A+B of the 16PF are contained in Table8.4. -for females

and Table8.5. for males. F tests of the significance of departures

from linearity were then undertaken between the etas and product

moment coefficients and the results of this analysis are also

presented in these tables.

i In the female group Factors B (intelligence), L (suspicious) and

Ql (radicalism) show significant departures from linearity. In

males, six traits have significant F values, these being B (intelligence),

E (dominance), F (happy-go-lucky), G (superego strength), Q2 (self-

sufficiency) and Q3 (social precision).

i 165

CATTELL'S QUADRATIC AGE CORRECTIONS

It was noted in an earlier section that in addition to graphical

representations, Cattell (1972) has reported precise mathematical

expressions of the relationship between age and the primary source

traits. These are based on cross sectional data and are provided

for the benefit of the researcher who wishes to compare the

personality of individuals or groups which differ in age. These age

'corrections' which reflect both the endogeneous and exogeneous

components are usually quadratic in form and of the type: -

Yadj =Y- (Y' - Y'

=Y- bl(X - XS) - b2(X2 - XS)

where X= age Y= observed score

v

If the primary source traits represent the skeleton, then these age

expressions are part of the flesh of Cattell's theory. In his earlier

work, 1940-1960, Cattell was primarily concerned with the identification

or 'discovery' of the source traits. Although considerable effort has

been spent in defending and extending the system, since those dates

Cattell has been concerned to provide a meaningful contextual

explanation of his traits. Indeed, it is only reasonable to expect

that once a theorist has made claim to certain constructs, he will

seek to define their nature vis a vis external realities.

166

But in Cattell's case this need was more acute because he was

particularly vulnerable to the charge that his work lacked clinical

impact or, in more extreme terms, that his theories were nothing more

than matrix algebra in search of meaning.

Any mathematical expression representing the relationship between

`ü 'v: l contain scre degree of error. YV ýi a;. per: - oý litt' iL. h- lst of

acknowledge the existence of such error, these expressions are

constructed with the aim of reducing the common variance between a

personality variable and age. One expects that such formulae

produce results which at least show a lower correlation between

age and personality than was the case before they were applied.

RESULTS ON CATTELL'S QUADRATIC AGE CORRECTIONS sr

The descriptive statistics (Forms A+ B) for the male and female adult

groups and for each age cohort were corrected by the formulae provided

by Cattell (1972). These data are contained in Tables 8.6. and 8.7.

respectively. Following these calculations, eta coefficients were

computed (Tables 8.8. and 8.9. ).

In the male group the two factors of F and N show significantly lower

eta coefficients after application of Cattell's age correction weights.

Factors B. E, G, I, Ql and Q2 show little difference between the two

sets of coefficients. The remaining eight factors show etas which are

in fact higher for the age 'corrected' scores; these being Factors A, C,

H, L, M, 0, Q3 and 04. Most of these discrepancies are not only

167

statistically significant but very large; the most extreme example

being Factor 0 (age uncorrected eta = . 07, age corrected eta by Cattell's

formula = . 78).

Turning to the female analysis, Factors F and N again showed significantly

lower eta coefficients after application of Cattell's weights.

Differences in et-as were small for Factors H, E, G, H, I, L, Q1 and Q3.

The primary source traits of A, C, M, 0, Q2 and Q4 all showed

significantly higher eta coefficients on the age corrected data.

CATTELL'S AGE CORRErTIONS: DISCUSSION

The results of the comparison of eta coefficients between corrected

and uncorrected primary source traits and age, show a fair degree of

consistency between males and females. In both groups Factors F

w (happy-go-lucky) and N (shrewdness) showed a significant decrease in

correlation after application of Cattell's weights, whilst B (intelligence),

E (dominance), G (superego strength), I (tender-minded) and Qi

(radicalism) demonstrated no appreciable differences.

Those source traits with the most marked increase of correlation with

age were Factors A (outgoing), C (emotional stability), M (imagination),

0 (guilt-proneness) and Q4 (tension). In fact eta coefficients on the

corrected data were so high in certain cases, especially with males,

as to make one wonder at the very origin of Cattell's age expressions.

In the case of C, 0 and Q4 for males, the 'corrected' data seem to

bear little or no relationship to Cattell's hypothesised trend lines

reproduced in figures 1 to 16 and so we must conclude that errors

of mathematics are at work here.

168

9 TIi A

a 0

a w H

Ö

U W

4

rn

4

w Hy

N a w a rl

,pR C6

a

IN to lfl rl Ol M M O N OD 14, N

a N d' M H Co Co N CV N N /-1 c1 d' Oý Lý m Co rn al ýt' a, a' 1ý () O ri C Y) ul to a)

N N N N N CV Co

lf) In C\I Co lf) N 00 If, a1 d' O N N

cr CO O O (7) N N r-I N N r-1 <V N

N Co C 1S ^ 0 ! fl co

. CJ H 711

Cr r4 co Iz CV H H H r-1 r-I

d' Co O) Co Co Co N E- Co N O " . " W 0) 0) w

N N N ý1 C \I, IV . 4t 00 ON f) NT Co N OD M tf )

Ln co N Co Co 0 CV CV CV N CV c*)

C4 0 cr% (A N H

rl 1 H

Q\ if) 0, N a) H H . " " " . "

H H H H iH H.

to N N Cý H 0 Co d' d' 0) N 0 x M N H

CV CV M

[` N H 0 t0 Ol M N O

Co CO cm CV N N

M

IV 01 [N U) q; r

Co M (V N N

N a% N N l0 Op O l0 M C\l N Q) W " " CO Co Co [,

CV N N N N R

N H O N O u) N N

M O CNJ c C Y) cn M

N N O c'1 0 H " co " " " " Ln in in d' C') S M

H H H H H '-4

N C\j Co N M N

r co N H -4 r N

N

qz?

Q i o e uý t t If) c p ý-I (V M LO

169

A W H

Ö

U

W

ca

4

O Gv

z H

A

ýi

ä

E'ý ýý

LO N o "

ON "

co "

(V "

N "

Cl) "

co N N N

N r-I

M lc: t to

O O1

N . -I

C) U)

"1 l0

to M

r4 N 4 l() (- N N N N N N

N O

N N 1' a% to r- N

co

O N

ON 1-1

ON ri

a1 r-1

O CV

v CV

(Y) 0)

In rn

U) o

CV M

co p

O . . . co (A

CY) N N LO N N

0 ý

CO 0) '1) N Oo

N N N N cm N

Lr) CV

M (V

N

r-I CV

rl

CV CV

O

M N.

U) N

Ol

Lp CV

CV dI

(5)

Un

O

to H

i

In cM

N

"

H " N " N

C.

N

C.

N

0 IIZP

N

N M

N G" r-f

(5) N

r-I CO

(+) CV

d' rl

d' N

N CV

C') C\ I

pl

N M N

M

(n N

U1

N

9

14V N

-

r-1

N

d'

N

LL0 "

N

N N N N N N

W " N

r4

N

" co

N .

p

" 0

r. 4

N

d0

N

u

ul

0 M

co N

CD N

IC) CV

N

N M

co

N N.

N "

M

-i M

"

M

OD co LX)

" -

to hfl

"

rl

to "

-4

N

, -4 nj

5-4

O N 5 -4

tto N

d'

C\l N

C

N

n

-4 N

t0

C\l

O

N

4 ui r--I

ui I

CV

d'

Lo M ýn

IV LA Ui

in

170

4 Ö W

v

ac ? I-fl 1-4

H

U)

p w ua

4

kD rn N M ri H

"

M M O M " " " a

N N O G" H M

CM " . a

(y CV p

M 0 M M r-1 N

0 "

H H

rn N

Co

z

In co N O

N C` In N S O

01 H N O H O

H " " .

0

N M co

N

W " ,

N r-I 0 W

IWO G CM en C U :

fV N

'4 " "

O

l 0 M ri N

4 "

0) 0 4 Ö te 4 -1 4U 0) 'a ' Q5 Qý q

171

Aº

w ö

cý

u w u 4 Ö W y

ä

oý w °°

C1 ý WW

Lr)

CO l0 N d' M

cyl

a ri N r-I N N O

N " " " O I

M p rI O

N

O " "

ri d' M

. -+ M N z "

N O N ri M

d' N N " " O

H

cm Co O M

O O O t7 "

N O

W

N

N tp

N O

N N 0 O . /ý W " "

Co N O

78 ü

.. i i i

e0 4-3 0) u (0 4-4

Qaý q

AP

172

In summary, Cattell's age corrections had little appreciable effect

on about half of his primary source traits. For the remaining factors

they introduced biases which were far more extreme than those in the

untreated data. It would seem that these corrections have indeed

been poorly developed and should not be used until such time as

new formulations can be researched.

In the interim, two possibilities are open to the researcher who

wishes to take account of age differences between one or more groups

of subjects. First, he may use the separate standard score norm

tables prepared for each sex and age cohort contained in Appendix 6.

Alternatively, it is possible to use the product moment coefficients

in a conventional regression approach, which despite certain but

usually small departures from linearity, is likely to be a more

reliable method than Cattell's age correction formulae. In any

event, there can be no doubt that the quadratic age expressions have

produced anomalous results, even against Cattell's (1972) own

graphical trend lines.

SUMMARY OF RESULTS

FACTOR A (OUTGOING)

Consistent very small negative correlations with age, although

the majority are not statistically significant.

Eta was not significantly higher than the product moment coefficient -

that is no evidence to suggest that a curvilinear relationship is

a better fit to the data.

i

173

Only female Form B curve approximated the hypothesised

increase to 30 with levelling off thereafter. Males on Form B

showed a slight linear decline from 18 years as did females on

Form A. Males Form A showed a decline up to 30 years and then

slight but non-significant increase.

Gene-rail sad s rate-: e. over form and sex seems to be a flat

or perhaps very slight linear negative relationship.

Obviously sex differences are more important than age variation

here - at every age significant differences remain with females

seen as more outgoing than males.

FACTOR B (INTELLIGENCE)

There was a significant negative correlation between this primary

source trait and age for both males and females. Tests of the

departure of linearity of regression were also significant, the males

showing a rise up to about 25 years of age, a levelling off until

about 30 years and more rapid decliKie after 40 years. Females on

the other hand showed a relatively shallow gradient up to about

50 years of age, when the decrease accelerates considerably.

In all the data are in rough agreement with Cattell's trend line,

although the point of deflection appears somewhat later. at 40-50

years and not at 30 years as postulated by Cattell.

i

174

FAC' RC (EMOTIONAL STABILITY)

This factor showed no significant linear trend in either males or

females. There was some tendency for adults of both sexes of about

50 years of age to have lower scores on this factor, although the

overall line demonstrated no significant curvilinear results. The

data do not in fact shop: any real correspondence with Cattell's

hypothesised differences in age; his curves being a marked

exaggeration of the empirical results found here. From Figure 8.4.

it can be observed that sex differences on this factor appear

more important than age differences.

FACTOR E (ASSERTIVE)

This factor showed significant negative linear correlations with

age in both groups. For the male sample there was a slight but

statistically significant departure from linearity (p----. 05).

Although Cattell's quadratic age correction formulae in fact had

little effect in reducing the co-variation of age with this primary

source trait, the visual impression from Figure 8.5. suggests that

Cattell's graphical' representation of age differences are relatively

close to those discovered here. In males, there is an increase in

means up to about 30 years of age, when there is a decline in

assertiveness. Females show a similar pattern only the decline is

somewhat earlier at 22 years.

p

275

FACTOR F (HAPPY_GO--LUCKY)

The trend in this trait is for a very pronounced negative correlation

with age (r = . 4) for both sexes. We have already noted that this

was one of only two factors where Cattell's age correction formulae

significantly reduced the co-variance with age. The graphical plot

of age cohort means shows a relatively close similarity with the

trends postulated from Cattell's model and these data support the

hypothesised direction of differences.

FACTOR G (SUPEREGO STRENGTH)

The strong positive correlation with age is clearly marked. in the

graphical plot of the 16PP means. This trend is strong in both

males and females. _ "

One explanation of this trend is of course a tendency for superego

strength to increase as an individual ages. We more usually think of

the process of socialisation and the development of superego strength

as occupying the years from birth to adolescence. But here is evidence

that perhaps the process continues considerably past these early years

in fact throughout life up to some 70 years of age.

This fits in well with popular prejudice, ie. reckless expedient youth

versus the conscientious moralism of the later years. In fact, this

is probably one of the major discriminators of the younger versus

the older personality. An alternative explanation of this phenomenon

is that the younger individuals in this study have experienced a

176

particular epoch where superego strength has in some way been eroded.

They will certainly have lived their adolescence' through the nineteen

sixties' where there was a revolution in the attitudes and morals of

society. And such changes in attitudes no doubt had more effect on

the younger 15-45 than the older 50+ age groups.

The furý. er possibility exists trat 'he sýmpiing method employed

in this study was responsible for this trend, for example, that the

young people in the sample were unusually low and the older people

of particularly high superego strength. One sector of the community

which was not tested for example, was prisoners who are no doubt

low on this primary source trait. This explanation does not seem

very likely however, there being-no clear reason why such an

unusual sample bias should have been introduced.

f With regard to hypothesis, the evidence does support a pattern of a

moderate but decelerating increase with age.

FACTOR H (SOCIALLY BOLD)

The results for this primary source trait, one of Cattell's major

extraversion components, show a relatively flat pattern of differences

across the six age cohorts. When Cattell's age differences are

superimposed on these data, results for females are relatively close.

However, the male curves are very dissimilar in shape, there being

no evidence of the pronounced inverted U shape proposed by Cattell

This discrepancy was also noted in the section on quadratic corrections

in that Cattell's formulae had little effect upon the female age

177

differences whilst eta on the male group was considerably inflated

after application of his weights. We must reject the hypothesis

then of a positive relationship between the source trait and age.

FACTOR I (TENDER-MINDED)

The general trend of differences on this factor is slight, positive

over age group, particularly in males. Cattell's postulated curves

are in fact relatively closely related to our findings but he may

tend to exaggerate sex differences somewhat. Nevertheless., the

variance associated with the sex of the subject is very large.

We have found some support., therefoize,. for the hypothesis of a.

low positive relationship between this primary source trait and age...

FACTOR L (SUSPICIOUS) i

Cattell has postulated aU shape relationship between this factor

and age in a study of cross-sectional design. More explicitly he

suggests (1972) that L (suspicious) declines up to 40 years of age

and then increases.

The data contained in Figure 9 show a slight decrease in scores for

males over age with no significant departure from linearity. Females.

differences on the other hand did depart from the simple linear

association, suggesting that scores decline until about 50 years of

age where they begin to increase again. This was about the strongest

non-linear trend in the personality data. The picture which emerges

on this source trait is best described as a form of rotated "D".

178

where differences between the sexes are relatively small at the

age of 20 years, slowly decline in males and more rapidly fall in

females thereafter, until approximately 60-68 years where the

differences between the sexes are again relatively small.

The exact reasons for these findings are difficult to establish

i. a study like this. Perhaps marriage in females has amore marked

effect on this personality variable than it does for males. It is

possible, for example, that young single females tend to be more

suspicious and hard to fool, that the level of this trait declines

through and as a result of marriage and as the years of widowhood

approach, begins to increase again as a natural defence. mechanism.

Obviously explanations such as this remain speculative; this study

was not principally designed to investigate such possibilities. However,

we can say that Cattell's suggested U shaped relationship between this

personality trait and age was supported in the case of females but

not with the male group. Indeed, the overlap between age and the

quadratic corrected scores in the case of males was very much higher

than eta based on uncorrected scores.

FACTOR M (IMAGINATION)

Results on this source trait indicate relatively small differences

across both age and sex groups. Product moment coefficients were

generally non-significant as were statistical tests for the

significance of departures from linearity.

179

Cattell's claim for an inverted U shape relationship was not supported

by these data and his corrections merely increased the correlation

between this variable and age.

As an additional observation, it is interesting that despite this

trait's relatively high correlation with Factor B (intelligence),

it dos not da-ionstrate the characteristic sharp decline in

performance at about 40 years, shown by many ability tests. It is

not improbable, therefore, that this scale represents the overlap

between the ability and personality domains; what some researchers

have termed'cognitive style'(Witkin 1970).

FACTOR Ni (SHREWDNESS)

This factor shows a moderately strong relationship with age with no

significant departure from linearity. In all, Cattell's mathematical

corrections were reasonably successful in reducing the common variance

between this trait and age. We have evidence. to support the hypothesis,

therefore, of a positive association with age.

FACTOR 0 (APPREHENSION)

The overall trend is for no significant relationship between this

primary source trait and age. Figure gi3 . is in fact similar to the

data contained for Factor C above, ie. a relatively flat profile of

age cohort mean scores with perhaps a slight rise at about 50 years

in the direction of increased anxiety of which this trait is an

important component. Again the effect of 'correcting' by the use of

180

Cattell's weights was to increase the correlation between age and this

particular personality trait. In the case of males the weights for

Factor 0 were so inappropriate as to produce a corrected eta value

of . 78; so high a result as to suggest some serious mathematical

error in the equations.

Ln view of these results, we must reject our initial hypothesis of

a fall in the personality trait of apprehension after adolescence

followed by a slight increase after 40 years. It would seem that

there is no clear linear or curvilinear association between. age and

this variable.

FACTOR Q1 (EXPERIMENTING)

This primary source trait intended to represent experimenting

behaviour, appears to be very much like Eysenck's concept of

radicalism-conservatism. (Eysenck 1969 , Wilson 1975. ) It is

rather curious in that it is the only factor where Cattell claims,

and we have hypothesised, no significant relationship with age.

Our results suggest, however, that this factor is one. of the most

strongly negatively correlated traits with age, both in the male and

female group. We must conclude, not unexpectedly perhaps, that older

subjects in this study tended to be more conservative than younger

subjects, especially if they were male. There was evidence of some

curvilinearity in the relationship for females but not in the male group.

I

181

FACTOR Q2 (SELF-SUFFICIENT)

Cattell has postulated that this variable tends to be positively

related to age for female subjects but to have no clear trend for

males.

Thus the hypothesised differences give a contradiction of the data

collected in this survey. Here we found that age was significantly,

positively correlated with Q2 in males but only slightly, if at all,

for female subjects. In males, tests. of departure from linearity

were in fact significant.

The general trend is for scores to increase on self-sufficiency after

adolescence until about 20-25 years where they level off until about

55 years. From that age onwards up to 68 years, the group means 410

tend to increase.

FACTOR Q3 (SOCIALLY PRDCISE)

The Q3 source trait, representing controlled, socially precise

behaviour, was hypothesised to have a strong, positive association

with age, levelling off in the middle years of life.

The results represented in Figure 15, show a relatively strong

linear relationship with age (r. =. 35) but little evidence of any

levelling off at 40 years. It is mainly this discrepancy in middle

age which is responsible for the male age 'corrected' eta to be higher

than eta calculated on the original data.

182

FACTOR Q4 (TENSION)

The hypothesised slight negative relationship of Factor Q4 with age

was in fact found in these data. However, Cattell's quadratic age

corrections again introduced more systematic bias into the data than

need existed in the first instance.

In the case of females, it was Cattell's postulated increase after

50 years of age, not present in these data, which was mainly

responsible for this distortion. Differences between the sexes

are quite evident from Figure 167 as is some slight rise in tension

level around 50 years of age. Overall, this source trait shows a

somewhat higher correlation with age than the two other main anxiety

factors of C and 0.

aR

R

183

AGE DIFFERENCES AND THE PRIMARY SOURCE TRAITS: GENERAL DISCUSSION

In all, the correspondence between Cattell's graphical descriptions

and the age differences investigated in this study is not impressive

(See Table 2.11. Whilst it is true that five of the primary source

traits (B, F, G, I, M and Q2) show reasonably similar age differences

to those hypothesised, the remaining ten factors vary considerably,

particularly QI (radicalism) and M (imagination).

In part these discrepancies could be due to cross-cultural influ ices

although it is improbable that the more extreme variations could be

explained by such effects alone. Both this study and Cattell's

research are based on a cross-sectional design with data being

collected within a few years of each other, so it is unlikely that

conflicting findings originate in these simple methodological Am.

differences.

A further possibility is that the British age cohort samples (N..: =--2100)

are in some way very unrepresentative of their respective age groups in

the adult population. Because we know that the total sample is

reasonably representative, we are forced to conclude that any such

bias is compensatory in nature. Moreover, the very fact that the

age difference analyses were conducted holding sex and age constant

means that the most likely source of bias involves the social class

composition of the age cohorts. Whilst bias in the social class make

up of the age samples cannot be completely ruled out, there is no

clear reason why this should have occurred.

184

If we turn to Cattell's data on the other hand, we have noticed that

the description of age trends show considerable inconsistencies. His

verbal description of the relationship of the primary source traits

with age are at odds with his own graphical representations and in turn

these are often inconsistent with his quadratic age expressions.

On Lhý' Crucs . ion of the

C; L? adratic relationship of curtain of Catttel l 's

primaries with age, there can be nothing to recommend his formulae.

The amount of additional variance explained by such procedures is

minimal and not sufficient to justify the considerable errors that

they may introduce.

GROUP DIFFERENCES AND THE PRIMARY SOURCE TRAITS: CONCLW IONS

Although there were some inconsistencies across forms of the 16PF,

earlier sub-sections showed that the majority of the , sex, cross-

national and social class differences in the primary source traits

were in line with those hypothesised from Cattell's model.

There is reason for much concern, however, on the question of age

differences in personality. Only two of the primaries demonstrated

a reduction of common variance with age after application of Cattail's

age 'correction' weights. These age results support the general

hypothesis, therefore, that the primary source traits have been

derived from inadequate data sources and are untenable in terms of

the verifiability of group differences postulated from Cattell's theory.

185

Having observed inconsistencies both with regard to questionnaire

forms and the verifiability of group differences, the next section

will consider the important and controversial issue of the

reliability of measurement afforded by Cattell's primary source

traits.

"

i

186

SECTION 4:

RELIABILITY AND THE

PRIMAR YS0URCETRAITS

gý ALTERNATE FORM RELIABILITY OF THE PRIMARY SOURCE TRAITS

410

INTRODUCTION TO RELIABILITY THEORY

The reliability of a test or questionnaire concerns the accuracy,

consistency or precision of measurement without prejudice to the

dimension being measured. Every measuring device is subject to some

degree of error whether it be a simple ruler or a complicated

electronic instrument. Reliability of measurement is a principal

aim in science because it is commonly assumed that reliability is

a necessary, if not sufficient, pre-requisite of a valid set of

observations.

In the classical test model the score of any given individual on a

test or questionnaire is said. to be made up of a true score and error

component.

That. is: -

OS TS +E

where 0a observed score s

Ts= true score

E: error ca iponent

On the assumption that the action of error is completely random,

it follows that:

mo

(a) the mean of all error scores is zero.

(b) the variance of error scores on all tests of the same length is

the same.

(c) the correlations of error scores with true, obtained and other

error scores are zero.

188

Having made these assumptions, we can make the statement that

observed score variance on a scale is made up of true score variance

and error variance.

(i) ö dt + de (basic classical test model formula)

where d2 = observed (total score variance).

0

= true score variance

d= error score variance e

The reliability of a test is oftenexpressed therefore as the

proportion of the observed or total score variance which:

is true score variance.

That is: -

0-2 (ii) rt (basic reliability formula, classical test d model where rt = reliability coefficient)

0

This, the fundamental reliability formula in the classical

test model, has the limitation that we never directly know

the true score variance.

But knowing that:

t (basic reliability formula) r dt

d2 0

derived from the and dt : dö - dE (basic classical test model formula)

then rt = dö 02

d2 0

189

and

(iii) rt =1- 6e

d2 0

Formula (iii) is a more convenient expression because it is possible

in practice to estimate error variance by a number of different methods.

This is achieved by correlating one or more sets of scale scores for

a group of individuals with another set of scores for the same group

on the same scale or variable. Correlation coefficients calculated

in this way are known as reliability coefficients and express the

extent to which a scale agrees with itself in producing consistent

or reliable scores.

Three main methods have been distinguished for estimating the

reliability of a test or questionnaire. Each of these techniques.

isolates different elements contributing to the scores and counts some

as true score variance and others as contaminating error variance.

INTERNAL CONSISTENCY

This method more usually involves assigning the items contributing to

one test score to one of two sub-scores, on some basis which will

produce two equivalent half-tests. Commonly, odd and even numbered

items are the two sub-tests. This method of calculating reliability

has the advantage of requiring only one testing session: the Spearman-

Brown formula is applied to correct for the effective reduction in test

length entailed by the split-half procedure.

I

190

In its purest form, this type of reliability is based on item statistics

and counts as true score variance only that part of item variance which

is common to at least one other item. This is known as the homogeneity

of a test and in a sense estimates reliability on all the possible halves

into which the test could be divided. Statistics developed to facilitate

this process are the Kuder-Richardson coefficients and the more

generalised coefficient alpha (Cronbach 1951 ).

Cattell (1972,1974 ), however, has pointed out that his measures

possess internal consistency without homogeneity. They are deliberately

heterogeneous in content in order that items have a high multiple

correlation on the factor on which they load. Moreover, items of

low homogeneity, he claims, avoid the problem of bloated specifics

which are to be avoided when measuring broad personality traits.

For these reasons, it would seem that conventional homogeneity 4P,

estimates are not applicable when investigating Cattell's primary

Source traits.

TEST-RETEST RELIABILITY

Test-retest reliability is estimated by repeat administration of the

same test. Irregular change in scores between the two administrations

is treated as error. Cattell (1970) has distinguished between the

'dependability' coefficient where the retest session follows too soon

for the subject themselves to change and the 'stability' coefficient,

where two months or more may elapse between the two administrations of

the test.

Stability is a very important concept in trait psychology: in the

191

calculation of stability coefficients both relative shifts in score due

to real changes in the individual and shifts attributable to random

measurement error are treated as error variance. Clearly, the researcher

will want to be able to distinguish between these two sources of 'error'

so as to be able to evaluate real changes in the individual. Attempts

have been made to separate traits which are relatively stable from

shorter-term in-Eluences which are more bound to particular situations

and change in strength over a short period. For example, the StateTrait

Anxiety Inventory (Spielberger et al., 1970) seeks to isolate those

aspects of anxiety which are situationally dependent (state) fron

those which are a more stable feature of the make-up of the individual

(trait). From the point of view of reliability theory, it is important

to recognise that the length of time which elapses between test and retest

may be an important determinant of the size of the test-retest coefficient.

Score variation, due to changes in personality and mood between test

and retest sessions, which will be treated as error variance, means

that great care has to be taken in specifying the exact conditions

or facets (Guilford 1970) of the reliability study. Further, the

stability of measurement over time tells us little or nothing about

the underlying structure of the item content and the very personality

model upon which it is based.

ALTERNATE FORM RELTABILLTY

The third commonly used index of reliability is the alternate form

coefficient, otherwise known as the equivalence or parallel forms.

coefficient. The different names all represent - from different

I

192

theoretical standpoints - the correlation of scores on one form of

a test with scores on another form designed to measure the same

characteristics. That is to say, the alternate form coefficient is

a measure of the extent to which two distinct forms which are designed

to measure the same thing actually do so.

:: eilst difficulties may arise in interpreting test-retest and

conventional internal consistency reliability estimates, there can be

no doubt of the importance of alternate form reliability to Cattell's

theory of personality.. Because of the difficulty. of establishing

really adequate external criteria of Cattell's primary source traits,

the model very much rests on the extent to which it is internally

consistent with regard to replicability across questionnaire forms.

As there was such a short time interval of about 10-15 minutes *º

between administration of the two alternate forms, there are unlikely

to be undue influences in the data collected in this study as a

result of changes in personality over time and for these reasons the

reliability analyses which follow are based on the alternate form approach.

THE STANDARD ERROR OF MEASUREMENT

The reliability coefficient is a useful statistic when a judgement

has to be made on the consistency of scores afforded by a particular

test or questionnaire. It is often the case, however, that we wish

to set a band of tolerance around a given score and make estimates

as to the likely error component.

i

193

From (iii) rt =1 ße

02 0

then r+ (52 e=1

62 0

and (ýe -1-r t

d2 0

Thus: d2 = Uö (1 - rt)

The square root of this value is 'generally known as the

Standard Error of Measurement (SEM)

SF4 SD 1- rt

AP

This statistic is much used in educational and psychological

measurement and can be said to represent the standard deviation of

obtained scores around true scores. It is also conceptualised as

the standard deviation of error scores.

In standard psychometric practice it is often assumed that the SEM

is constant for all score levels and that the standard deviation of

error scores is normal. It is then possible to set up confidence

limits from the normal probability curve, where 68% of observed scores.

will lie 11 SEM units and 95% 12 SF4 units about the true score.

I

194

ALTERNATE FORM RELIABILITY COEFFICIENTS: RESULTS

The alternate form reliability coefficients representing the product

moment correlation coefficients between the primary source

traits of Forms A and B of Cattell's 16PF are listed in Table 9.1.

Results are given not only for the total sample but also for the

sample classified by sex. The standard errors of measurement (SEM)

in n-ste-n units for : ach factor have also been calculated and are

entered in Table 9.2.. But it should be noted that the values are

for the questionnaire forms used separately and have not been

corrected for two form length by the Spearman-Brown formula.,

The first observation is the general order of these coefficients

which are low by standards of cognitive tests. Factors H (social

boldness), F (happy-go-lucky) and Q4 (tense) are the most reliable

primary source traits which is in accord with Cattell's findings (972).

Primaries with coefficients around .5 are A (outgoing), C (emotional

stability), E (assertiveness), G (superego strength), I (tendermindedness),

0 (apprehensive) and Q3 (social precision). Next are two factors which

show rather low coefficients, namely B (intelligence) and Q2 (self-

sufficiency). Finally are source traits in the very low reliability

range of 0.2 to 0.3; these are the factors of L (suspicious),

M (imagination), N (shrewdness) and Q1 (radicalism).

It is of interest that most of the major introversion-extraversion

(A, E, F and H) and the main anxiety factors (C, O and Q4) show tolerable

reliability for variables in the personality domain. The same cannot

be said, however, for a relatively well documented variable in other

195

H H

W U

O

a

a a x U

a O L"a

CQ

4

u.

En H z

H

U U

W

cz N w E

ßf1 to CV rr N. N Ol to to Q1 O CO h co -e to r 01 r c: > tO O Cfl N- O Ln n') IS) q} L2 CD, ; ;

r- Cl Lc r

Nnt r

rLn r

Ci r F-

d

r

m C) M Qn to ct qt N- Ol O lD U) O to N 1ý

m to M cm tO U') LA rý mt r- ;; { C) r-- LO

NM r

rM r

NM r

NM r

rM r

rM r

t7 n O IA N n %Zt r LC) 0) U, Lo r O d' Co CV

CV Cf d tL ) Ln N 110 Ch N- M (ý) L! ) (ý7 r om

rM r

Co M r l") r

U) M r- Ch r

Co M

CM U) M U') U) (F) CV tO Ct LO CO Nr d' tO N- NM

r N rM 1.0121 N N N-O N rO MCQ C: Dr

alt CI) U)N - OM r

Co en OOM OpN

U) al M O N- clM er ul CO a' r (F) N- tO (F) 00 CV

O ýp MM Co LA rN if i Ln LC) LO O 1ý to

r 'ct r Od Oý cf Nd N «z;

(V Q1 ct U) O N- U) C% M O Ql 1- r l0 r %; t C') N

z N rr- to 00 CV U)O cPCD cV N-r- to N.

rM r

ON r

OM r

OM r i---M

F- C) vu r

M 01 1ý O) U) to O M Ida N 00 N U). N 0) (F) N OU) NN- N

- dd U) N- M fam. to Ch (M

CV M I

OC7 r

CVM r

OM r

r. M r

OlM

M Co CO MCo N-O N- im OD LO CO pM Q1M

M Net' UA N- N tD if O tD c' U) «t' O to

(ID M ON OOM 01N Iý. M Co cm

Lr) 0% CM If) tO Co to to to ",: t Ct 00 U) OO U) M tD to N Iý O it qi' r- ch CV M M N- tO CO

rM OM r

Q1(M 00 m M CV r

CJ r

M Mý ýr MO NM

M C`)t- to Co

= 1- m OO 00 N- "

CM to a1 tO N- "

th (F) 1- N-

Nl1) CM to r-

MLA r

MU) r

rLt) r-

" to

r

h 00 0' NN 00 N- Lil O

O a' 00 P- C xr rU) U)O er M NO -e OM N-C)

N NM NM CV M NM NN

d Co it tO 00 ßf1 t0 cf OO ct CD M N N- ýy tG q& O CO N- lD - mr r Co O tO U) Ol Q' U)

N Lff r) (V U) M U) CV ýk M Ct

r toto Nr O0 F N Co to CVI -e- le

W to CO ýt rr t1 f ý"ý M 10 O h ý ct O l0 N.. Oit r-

0-A& r

CM q& I

cm «Z- Ql'- cl

M M U) r OM Co Lin 0) O "-- tt Co O CO 00 rl_ V LA f- 01 N- O U) cV Ql CO Oº O

M r

'ý et r

Ln M F-

tC) ýt r

ýt M r

MM '-

M t'7 t0 r st Co N NN it C) ct' e-- C% I LO

m O F- M Qº N . - «t- O Oº O N C% N- eV N- - N- cV N- N t0 N fr

cu Ln N . - NtD MM a' to N- N0) cm cn 0 -el U'>

Q U) (V O it tD ct c2 O NW Ct O CO ý . -: OM ý

c%M 01M Com rN r

OM r

M in d Ö

d m -i cm Q C

v f C

i

Qr Z: C C N 10C RfC J 11 m MC du(s Z II

1

du C3 tud L N N L

Oº. CL) L+- L A-

iv; LL- 2c L LOi. to

i v)

Ci , o; N ca 9

U)

z ö

Un

E4 0

z U)

q

U W

U)

U)

u) N H

t-ý

O Ul

r v co r} rl N

I Cl I}' Cl ' V Or " "

M C') Cjý N U) LI Lf) 0'

r4 4 14 M ý' C

l0 M H 0 (V M Cl

r-1 -4 H

Cr) N z r r r

rH

r-1 ON

r-1 H r-i

(D o Lr)

H M ý1 lJ

r1 1 H

O O O

un Cr)

H H .1

co r H

N

Cr) M

r-I H r{

O 0

rI H rj

O pp -

Li `" ý"

O N C\l

r Sr ö ö

En En ö

ý 114, Au g2 2 O 7 W ..

40

197

researches (Eysenck, 1954, Wilson, 1972), Cattell's primary source trait

Ql of radicalism.

In order to investigate differences in the reliability of the primary

source traits for the male and female adult groups, the coefficients.

in Table 9.1., were converted to Fisher z-statistics (Table9.3) and

the statistical significance of the differences calculated.

The results of this analysis show that the male group had significantly

higher reliabilities on Factors A (outgoing), I (tender-minded),

Q3 (social precision)"and Q4 (tense) whilst the female group had a

significantly higher coefficient on Factor M (imagination). When the

means of Fisher z-statistics were calculated across all the primary

source traits, there was no significant difference between the two

groups at the 5% level (z for males = . 523, z for females = . 498).

aw We must conclude, therefore, that although the sex of the subject was

related to the reliability of five of the factors, there was no

significant trend for either sex to demonstrate consistently

higher reliability coefficients across all the primary source traits.

To test the hypothesis that the adult reliabilities would tend to be

lower than those reliabilities collected for student groups on the

primary source traits a comparison was made not with Cattell's American

student groups but with data based on a more relevant sample of British

second year undergraduates collected by Saville and Blinkhorn (1976, pp 78).

For the total group of undergraduates (N = 1148) and adults (N = 2007)

the student reliability coefficients were significantly higher on eight

198

H H

N

aw x

Hz

W[z,

Cif

cn

Aw H il Ö 4'

mz cn

cli

ý>ä Co Ä E-+ U4

0 r-I CO

w

t0 M H

r-1 0

M m

lJ

lppp V'

Cr)

H 0 " " " "

N d' i co

/

0

Cy rQ c\j o n

N M

Cr) in

H

o 0

I N °

z N N 0 Z

C\j r-i

CM 0

N O

In O In N

" cyl 0 V3

z rn O

H M co tN , -1 H In Cr)

" r-i

" 0

"

c C

In

C) " " o

t. vn z

N M

O H

N N

C7 un u1 O vl

10 CD in co N 0 Vl

in o

In N

0 co

10 in 0

10 C (Y) ý

u 6f'ß In O fn

H LO

d M m

Pq d' "

O z

i I

0 M

mot' O O

G] E4

vl ýc7 ,

ýý

W

Ü

H V

W 1 N

(W Z0

ý - N H 4 H

Gý Q vn

4s

199

oý w m H

x H H

N M d, O1

"

t0 ýD

"

l0 cri

""

N 0

"

Lt} 0

"

N 1M Q1

M

Of Co

"_ 110 l0

" LO Ln LI

"" 1

a) 0 Ln 0

" N d'

Cr)

Lo

1

I 14; r N co 0

! to to O

NN ""

Ln N

r-4 O

" .

%D "0 En

1 to

Ip tD

uý Ln LO

H r-1

z N N C. (V O" Z

Cr) M' NN d' E N N N1N 0

t

Z

1 N c '(

H ýD 0

1.4 co c n c c4 1 z

O M O

M Cr L

0

H t0 ý 1 ""

Ö

" z

S

O 1 0)

ON MN O

r-f x ao 1o N1O '-+

H

ID M H CMM1 ,... I t7 ýn to d'

0 N

N W CO

N LO t- ID N

N 0) 0

U) O

Oý N1 fit' 0 0 to lO ý1 1 l0

I O. vl

d' 1 d'

0 I rI

z 1N

r- N

N4,1- It) O In pq c+) M ýP d' O

1 O

CO M -4 ko FrT- H I r-4

. eC to Ln I-n 0

Cl) N 11 N

'-" W WU 2

pQ C\j 1 U 0

H W

GH

V

G] tl ý7 (I (s4 H

¢. zz Äz Iý

ý `ý A M

200

of the sixteen primary source traits. These were Factors A, F, G, H,

Q1, Q2, Q3 and Q4. The adult group showed higher alternate form

reliabilities on Factor B whilst the remaining seven factors showed

no significant differences. When the Fisher z-statistics were

averaged across all the primary source traits for the two groups,

the difference was significant at the 5% level. Thus, with the

exceptions noted above, we can conclude that the overall reliability

of Cattell's model as represented by the 16PF is in fact more reliable

on the undergraduate sample than on the general population adults.

On inspection of the descriptive statistics for these adult and student

data collected by Saville and Blinkhorn (1976), it is evident that these

are differences not only between mean scores of the two groups but also

between the scale score standard deviations, the students showing greater

dispersion of scores for the majority of factors. The restriction in

range phenomenon could be responsible therefore for these differences in

reliability coefficients and indeed when restriction of range is taken

into account by use of the SEM statistic , the differences between the

groups are relatively small.

The finding that students showed greater variabilty on most of the primary

source traits is perhaps a little surprising especially as undergraduates

are very selected sub-group of the population. However it must be remembered

that although restriction in range can arise through some influence extrinsic

to the reliability of a test or scale ( for example members of a sub-group

being in some real sense more alike, or perhaps some artificial

truncation of distribution), restriction of range can also as a result of

an influence which is intrinic to scale reliability (inappropriate or

unreliable items tend in themselves to restrict the overall dispersion

of test scores).

201

We should be cautious therefore in making the assumption that correction

for differences in the variability of scores by way of the SEM statistic

enables direct comparison of the reliabilty of a scale in different samples.

It could well be that greater dispersion of scores in the undergraduate

sample is a result of more appropriate, meaningful and reliable item content

for student groups.

201a.

All

10. AGE AND SOCIAL CLASS AS MODERATORS

ON RELIABILITY COEFFICIENTS

202

INTRODUCTION

It is not sufficient that the measurement of personality be reliable

for the adult population as an entirety or for that matter for groups

of students. It also needs to be shown that scale reliability is not

adversely effected by variables such as the age and socio-economic

class of the subject. Low reliability for sub-groups of the

population might suggest that a trait lacks stability, that there is

restricted range, that factor structure is not invariant across

subjects or that item content is simply inappropriate to certain

types of individuals. But whatever the origin of moderator effects

upon scale reliability, they remain a crucial concern to the trait

theorist, especially if his theories make claims of general

applicability, yet had their genesis in small and perhaps atypical

academic groups.

In an earlier section it was postulated that Cattell's primary source

traits would be less reliable on older, lower socio-economic class

subjects. In order to test this hypothesis, tables in Appendix 10

contain alternate form reliability coefficients broken down by the sex,

age and social group of the respondent for each of the primary factors.

Tests yielding only one score are amenable to the investigation of

changes in reliability with moderator variables without recourse to

special techniques. The Cattell inventories, with as many as sixteen

scales to be considered, present a more complex task. One approach to

this problem, however, is to rank the equivalence coefficients within

203

each factor over the classifying variable, and sum the ranks over

all factors. The resultant sums of ranks would be an index of the

relative magnitude of the equivalence coefficients for all factors

with respect to each classification.

AGE AS A MODERATOR ON RELIABILITY: RESULTS

Accordingly, Table lO. llontains the results of ranking the alternate

form reliability coefficients from highest (rank 1) to lowest (rank 4)

on each factor and producing a frequency count of these values across

all 16 primary source traits for each of four age groups (16-24,25-44,

45-64 and 65+).

TABLE: 10.1.

RANK FREQUEINCIES OF THE ALTERNATE FORM RELIABILITY COEFFICIENTS

CLASSIFIED BY AGE GROUP

Most Least Reliable Reliable

anks Age

1 2 3 4 MEAN SD

18-24 6 5 4 1 2.00 . 98

25-44 4 7 2 2 2.09 . 97

45-64 3 3 6 4 2.72 1.06

65+ 2 2 3 9 3.19 1.11

It is evident from this table that there is a gradual decrease in

the chances of a high reliability coefficient for a given factor as

s

the age of the respondent increases. This trend is significant at t

204

the 5% level (x2 = 9.00) when the data are represented in the

2x2 contingency table below.

TABLE: 10.2.

SUMMARY TABLE: RELIABILITY CLASSIFIED BY AGE

"Higher" "Lower" Reliability Reliability

Ranks 1&2 Ranks 3&4

16-44 years 22 10

45-70 years 10 22

x2 = 9.00

We can conclude, therefore, that there is a significant trend for

Cattell's primary source traits to show lower reliability coefficients

for sub jectu of higher ages.

An alternate approach is the use of the parametric Fisher z-statistic,

a logarithmic transformation which enables. the inspection of product-

moment coefficients for significant differences. In Table lO. 3the

reliability coefficients for each age group have been converted to

z-statistics and the number of significant differences calculated on the basis of two - tailed significance tests.

It is evident from this analysis that seven factors, with a total of

fourteen significant differences between reliability coefficients,

gave results consistent with the hypothesis that older subjects

would show lower reliability on the primary source traits. These

205

m ö

H

cn

E-4 t H E-i

rx

U) H W 0 E-1

UI E-f

H

W W 0 U

H H

H L¢n

Ha

N cri t

x E1 O C-l

CO N ri r- a r- ý c ca

0 O

ýp CV 01 U) a 0

" L

" " 0 0

0) LI) co

or d' d d' co 0

N N

"

CN CV

"

CO 0

"

CO ri

" N 0

00 ca

M Lr)

N -4

O V E- 'D 0 O

W C\l

Lo -1 CIO

OD r-

z N "

N "

CV "

O "

0 O

co N M 0

H M Co

N co 0) N

co "

M

" N co

"

oN

."

O 0

M lD N

H M w

LS N U

If) ca 0 0

N N H

X. 01 0l (71 N N 0

C7 ý

" " M

" M

" N O

M CO

1 0CP\ co cl-

H

k+ .

CO "

Ln 0

Ln "

ý") 0

M 0 0

ý) to H 0

N H 0) M

0

W c ý d 0 "-1

° cC ,n

LO U') ý0 Ln

rn CO O N

z co -4 I ý 4 H

" d

" d '

" -73 Ö -I-ý a

ý J (

U C1 4 M

U "d "di

s4 ý0v ., i a Q 4-3

o Icr unul°0 inUlP'

W N I w U r ti 4-t "r -º 4-4

U

r N ý ONO O , -I . ca

ON 41 NN

A 4a O 4-4

4-i E-i 4-4 Z

ý Ä Q-

206

R4 ': ß FREQUENCIES OF THE ALTERNATE FORM RELIABILITY COEFFICIENTS

CLASSIFIED BY SOCIO-ECONOMIC GROUP

Most Least ReliableR eliable

e, SS

AB 1 19 6 0 1 1.56 . 81

Cl 6 6 3 1 1.94 . 93

C2 1 3 8 4 2.94 _85

DE 0 1 5 10 3.56 . 63

27

were namely the Factors of F, G, H, M, Ql and Q2. The three

primaries of A, B and C on the other hand demonstrate a total of

four significant differences in reliability which were in the

opposite direction to that hypothesised.

SOCIAL CLASS AS A . NOD: BATOR ON RELIABILITY: RESULTS

moderator class as a on rý.! _iabilitV, the

coefficients were first ranked from highest to lowest in the same

manner as with the age data. The results of this analysis are

contained in Table 10.4. and in the 2x2 contingency table reproduced

below which is significant at the 1% level (x2 = 30.25).

TABLE: 1O. 5: SOGTAL CLASS AS A MODERATOR ON TI-CE RELIABILITY OF THE PRIMARY SOURCE TRAITS: SUNLMARY TABLE".

"Higher" "Lower" Reliability Reliability

Ranks l&2 Ranks 38: 4

Social Group A, B, Cl 27 5

Social Group C2, D, E 5 27

x2 = -30.25

The corresponding Fisher z statistics are presented in Table 10.6.

In this case, thirteen of the sixteen source traits show a total

of 37 differences which are in line with those hypothesised, (-'-,.,,

Only Factors C, 0 and Q4 do not show significant differences

between the reliability coefficients of the four socio-economic

groups. In addition, the mean of the sixteen reliabilities for

the AB group is significantly higher at the 1% level than that . 11

of the DE group.

208

C;

H

En H

U) rl

N

H Ga O H

zl w

O

H

WQ

aH

W

O

H a a

U

a

8 I)

Iý

P4

N co O o

Cl Co N N N

m '

M t-

I CY)

M CM I O Q N lfl d

LN In

N I

Q' om Co O ci

,

1-4 In C\i

N -4

N O

Co d O

a M d' N N

C) C C) N

C) C) J

Z 0) -

N N N r!

N r-i

r-1 M cM CO

N

N qlo

N 0

x O

Q

c3)

C% "

Co C) oý

"

N I

" N 0

ri

ýO

N

ul M ýP

O d' M O

Gy p N

Co

O, % 0,1

00 N

N 0

G1 M N (Y)

ý0

M

u1 c`ý O

C") M

l0 (Z)

Co LO M

U un Ln O O

N N N M

Ql (

CU 0

O

N Co 0

N O

Q2 d' (D

O CD

Ct A

'O

O 4

U N

ß IN r 4i SI

äi v1 .

r!

WON

vl 0l( Ö

M U)

r-1 0. " 0) 'ßx5.

u N 4-4 ON

4-f O

a v V fý 41 0, G)

NN ab ab

°O .

4-1 ÄQ Ä

41

209

AGE AND SOCIAL CLASS AS MODERATORS ON THE RELIABILITY OF THE PRIMARY SOURCE TRAITS: DISCUSSION

Results give tentative support to the hypothesis that the older age

groups (45-70 as against 16-44 years) would show lower reliability

coefficients on the primary source traits. This was by no means a

particularly strong trend however, but was mainly as a result of

the action of six of the sixteen primaries (F, G, H, M, Ql and 02).

It was interesting that the non-parametric approach of ranking and

using a chi-square analysis on age as a moderator variable gave more

positive results than the parametric Fisher z statistic. Although

the mean z statistics for each age group across all 16PF factors

were generally in the expected direction, they did not reach

statistical significance as did the chi-square analysis. This may

have been because of the large differences in the size of the alternate

form reliability coefficients across the various factors which perhaps

mask the more subtle influence of age. The conversion of these data

back to correlation coefficients shows that the mean reliabilities

across all 16PF factors were . 490 for the 15-24 year age group,

. 495 for the 25-44, . 460 for the 45=64 and . 435 for the 65+ age

group. in view of these relatively. small differences, it is clear

that weilst there may be some slight trend for lower reliability in

older people, variations as a result of differences in factor

reliabilities are likely to be a more important consideration than

differences as a result of age.

We also saw that there was a tendency for the adult group to show

lower reliability than data collected for university students S

210

(Saville and Blinkhorn 1976). This is despite the fact that

undergraduates are a restricted group both in terms of age and

academic attainment. Thus, the evidence supports the original

hypothesis and, to some extent, those who are critical of the

applicability of questionnaires in general. It would seem that when

scales are developed on specific sub-groups of the population such

-is __..; _nt:, that se'_aýcted items are not such a good fit to the

model in the adult population as had been the case with the original

development samples. This is an important point to the personality

theorist who should seek to avoid item content which may be more

appropriate to student groups than adults in general. After all,

a generalised model of personality should provide reliable

measurement not just across specific accessible student. groups

but across the entire population range.

When the effects of social class were investigated as moderators

upon scale reliability coefficients, the results were more positive

than with the age analysis. There would seem to be good evidence

that the 16PF scales showed less reliability with lower socio-economic

groups. Both the parametric z statistic and the non-parametric

chi-square analysis gave very significant results here. In fact, only

Factors C. 0 and Q4 did not show significant difference between the

reliability coefficients..

Apparently then, social class has more effect upon scale reliability

than the age of the respondent. Perhaps this is not altogether

surprising when one considers that social class is some, albeit

indirect, measure of educational attainment and, therefore, academic

211

ability and we may merely be saying here that those of lower scholastic

ability tend to be more unreliable in their responses to questionnaire

data. It is certainly true that the lower social classes will have

had less experience with questionnaire techniques, with form completion

and all the assorted clerical activities which are more usually the

part of higher social economic jobs. Indirect evidence in support

of ß: i0 explanation is the positive correlation between social class

and Cattell's primary source trait of intelligence (Factor B), which

we observed at an earlier section.

The Fisher z comparisons between the reliabilities of the four

social class groups employed here did in fact show significant

differences and when these were converted back to correlation

coefficients, it was found that social class AB had a mean reliability

of. 540 across all 16PF factors, Cl a mean reliability of . 535, C2 a

mean reliability of . 465 and the DE social class a mean reliability

of . 435. Quite clearly these differences are large enough to be of

importance in scale interpretation, even allowing for the large

variations of reliability across the primary source traits.

Certain factors had particularly low reliabilities for most measurement

purposes. It would appear that Factor L (suspicious) especially in

the lower social classes, Factor M (imagination) in the lower social

class and higher age groups, Factor N (shrewdness) in the lower social

classes and higher age groups and Q1 (experimenting) mainly in the

higher age groups, were particularly vulnerable to these moderator

effects. For these source traits, the more extreme age and social

class groups possessed such low reliability as to bring into question }

212

the very meaning of scores on the primary source traits obtained by

such subjects.

As a final question we might ask what the effect of restriction

upon range of scores within certain sub-groups had upon the

reliability patterns reported in these various tables. Although

1972) to e t-L-Tl . te -41e reliäb. 'i1ityr

of a scale based on different group standard deviations, they make

the assumption of the consistency of standard error of measurement

across the two groups. Clearly, this is an assumption we cannot

make because this is the very effect which we are investigating.

What is more important is the fact that 'correcting' group statistics

for restriction in range can often be missing the point. It is a

perfectly legitimate aim for a test constructor to increase the

discrimination of his scales to a maximum. All things being equal,

this will tend to help the reliability of his scale. If, therefore,

certain personality source traits show markedly reduced variances

across a certain age group or social class, correcting for such

inconsistencies would be fallacious. We would only gain statistics

based on artificial assumptions, which will have no practical or

theoretical value in understanding the effects of moderator variables.

Nevertheless, it would be remiss not to investigate the extent to

which restriction in range of scale scores affected the reliability

coefficients calculated here.

I

213

11. VARIATIONS IN THE STANDARD ERROR OF MEASUREMENT

214

INTRODUCTION

Up to this point we have researched variations in the values of alternate

form reliability coefficients. The problem of interpreting such data

where restriction in range is suspected within specific groups of subjects

has also been introduced. The statistic which is commonly supposed to be

r..... ^. nrdeni= of Va iat! iity the sample is the standard error of

measurement and perhaps we should consider this statistic in a little

more detail.

It was stated in an earlier point in this section that the standard error

of measurement (SEM) is commonly supposed to be constant at different

score levels. We have previously defined the SEM as:

SEM SD --r

Guilford (1973) states that the SEM is ablanket index and that if the

score distribution approaches normality and if obtained scores do not

extend over the entire range, the standard error of measurement is

probably uniform at all levels.

Surprisingly, few investigators have challenged this assumption, and the

published studies which do exist seem to be restricted to the testing of

educational attainment. Gulliksen (1950) concludes that the SEM should be

constant for all score levels when the distribution is not skewed and when

kurtosis is three. Mollenkopf (1949) investigated the effects of skewness

and variations in kurtosis on the SEM and concluded that slight skewing

could be tolerated but not departures in kurtosis from three. Lord (1352)

points out that in most distributions the dispersion of errors will be

smallest at the tails of the distribution. Thorndike (1951) concludes,

on the other hand, that the precision of measurement will be smallest at

that point in the score distribution where item difficulties are concentrated

215

which for a general purpose instrument is likely to be the mid-point

of the scale.

The lack of published work on the possible variability of the SEM in

personality scaling is perhaps the more remarkable, when one considers

that the question as to whether certain personality traits in reality

constitute continua is open to more doubt than the more tangible

variables met in educational test data.

A particular problem is one of the dimensionality of the middle response

where, as with Cattell's inventories,. the subject is required to choose

the most appropriate of three possible responses to the item stem. In

such an item format, the middle response could easily become the catch-all

for otherwise unclassifiable responses including item inapplicability and

uncertainty of meaning for the subject. In view of these points and Lord's

work (1952), it would seem reasonable to suppose, therefore, that the

extreme categories of the primary trait score scale will show the

greatest reliability.

Turning to the mechanics of calculating the SEM at different parts of

the score scale, Guilford (1973) has suggested that if there is reason

to suspect that discilnination is actually unequal along the scale, this

can be examined by preparing a scatter diagram showing the relationship

between two forms of the test or questionnaire. The standard. deviations

of the columns and rows at different score levels will indicate where

predictions have the greatest accuracy.

One problem with this approach is that it is more akin not with the standard

error of measurement but the standard error of estimate which is defined as

the standard deviation of errors of prediction from correlated measures and }

is given by the formula:

216

SEE = SD 1- r2 X xy

Comparing this formula with that for the standard error of measurement:

SEM = SD x-r x xy

it is apparent that the SEE isVi + r11 as large again as the SEM. "IV

What is not generally realised is that the SEM is in essence the

SEE corrected for unreliability (attenuation) in one of the variables.

That is to say, that if we take the SEE formula and correct it for

attenuation in one variable by the standard formula we can derive

the SEM.

PROOF

(i) SEE = SD 1- r2

(Standard formula for the standard error of estimate)

(ii) Correcting r, for attenuation in one variable then: -

r=r xy xy

Ny yx

(iii) Substituting:

(correction for attenuation formula)

SEEC SD 1_Y2

ryx ý _.

217

2 SD 1 -r r yx

(iv) But in a reliability study the best estimate of yX = rXY

thus substituting in the formula then:

SEE SD 1-r2 cx

r XY

= SD 1 -r XY

which is the standard formula for the standard error of measurement.

Essentially the SEM is a hypothetical value which apportions error

variance to the two forms in a reliability study and makes an estimate

of the unreliability in the form assuming the other to be perfectly

reliable. Calculating the standard deviations of columns and rows at

different score levels of the scatter diagram as proposed by Guilford (1973),

takes no account of this and if such data are to be interpreted against

the overall standard error of measurement, then the suitable correction

factor would need to be applied.

Lord (1952) has proposed a simple formula for calculating the standard

error of measurement along different parts of the score scale.

SEM = to (n - ta) p (n-1)

where SEMp = standard error of measurement at a particular }

point on the score scale

218

to = total of items correctly answered

n= number of items in the test

For our purposes, however, rather restrictive conditions apply to this

formula in that it assumes that items are scored 0 or 1 and have been

assigned either randomly or in some stratified fashion to the alternate

forms of the test. It is ä1so assumed that one is sealing with

homogeneous tests and not the deliberately heterogeneous item content

of Cattell's inventories.

Fortunately, it is possible to take another approach to the problem

by reconsidering a basic definition of the standard error of measurement

as the standard deviation of observed scores around true scores.

That is: -

SEM = 7- (x - xt) 2

N

where x= obtained score

xt = true score A

N- number of people in the group

Whilst we cannot ever be absolutely sure of an individual's true score,

it can be estimated by calculating the mean universe score, which in this

case will be the mean of the Form A and Form B factor scores. That is: -

ASS xt A+B

2 }

219

where xt = estimated true score for an individual

SA = sten on a given factor for Form A

SB = sten on the same given factor for Form B

By substituting in the above formula then: -

SF1, i SA _ ($A s) -2

2

N

which reduces to:

SEM (SA _ SS)

2

4N

By the use of this formula it is possible to make an estimate of an

individual's true score, calculate its departure from the observed score

(either Form A or Form B) and then compute the standard deviation of

error scores around true scores for each point on the N-sten score scale.

It will be noticed that this formula makes no assumptions about using

binary data or that the test is homogeneous in item content. Moreover,

the use of both forms in estimating'the true score means that the

values gained by way of the above formula are more analogous to standard

errors of measurement than the simple standard deviations calculated from

the rows and columns of a scattergram.

The foregoing demonstrates some of the problems which arise

when investigating errors of measurement. The most obvious source

of confusion is the definition i

220

of what really constitutes error.

obtaining column and row standard

in fact based upon the logic of d

estimating (predicting) the score

parallel test. We shall refer to

an observed score where:

Guilford's (1972) method of

deviations on a scattergram is

fining error as the error made in

on one test from the score on a

this as the error in estimating

si_ std /1-r2

x xY

An alternative but equally acceptable approach is to define errors

of measurement as the error involved in substituting a score on one

test for a score on a parallel test.

In this case we can write the formula:

SE 1d d2 /(XA XBi 2

where SEMd = standard deviation of the error in

substituting a score on one test for a

score do a parallel form.

XA = score on the first form of the test

XB = score on the parallel form

d= difference between scores on the two forms..

Finally we have the more commonly used definition of error developed

here and that is the error made in substituting the observed for the

true score:

221

I

SEM =X _X It2

N

SD Vx-r

x xy

On the assumption that the best estimate of an individual's true

score on a given trait is the average n-sten on the parallel forms

of the 16PF, then we have established above that: -

SEM (SA - SB)2

4N

This statement bears a close resemblance to that derived for the

SEMd (the standard deviation of the error in substituting a score

on one test for a score on a parallel form) and in fact is 1 as large. 4T

We can see, therefore, that the SE14 will be half as large as the SEMd

on the same data base.

There are other possible definitions of error but we shall investigate

how the conventional definition of the standard error of measurement

as the standard deviation of observed scores about the true score

(SD a1 --r ) affects our conclusions on the differential reliability

of the primar' source traits. This particular statistic has been

adopted here because it has the distinct advantage of being based

upon the most generally accepted definition of error according to the

classical true score model.

222

AGE AND SOCIAL CLASS AS MODERATORS ON THE SEMs: RESULTS

SENs were calculated in raw score units on the primary source traits

for each age and social class group. The results of this analysis

are presented for the total sample and for the questionnaire forms

separately (Table 11.1. for Form A and Table 11.2. for Form B). Age and

C l_ ? class were divided Into the four ca tego'ries listed in the

tables.

Because the data are represented in raw score units, no meaningful

comparisons can be made between the SBMs on the different personality

variables. To ensure comparable results, therefore, SEMs for each

social class and age group were ranked on each factor independently

from highest to lowest. Summary tables were then prepared and a

chi-square analysis undertaken (Tables 11.3 to 11.7. ).

AGE AND SOCIAL. CLASS AS MODERATORS ON THE SEMs: DISCUSSION

By referring to Table 11.7 we can see that overall the Form A

scales had lower SEMs in older than in younger subjects, whilst in

the 16PF Form B there were no significant differences. This is in

direct contrast with our previous analysis utilising reliability

coefficients where older subjects tended to have lower reliabilities.

Because the only other statistic appearing in the SEM formula is the

standard deviation of scores, we are forced to conclude, therefore,

that most of the difference in reliabilities amongst age groups is

due to a restriction of scale score variability amongst older subjects;

223

this effect being particularly pronounced in Form A of Cattell's

16PF questionnaire.

Turning to social class, we find that whilst those classified as

being in the lower social classes tended to show lower dispersion,

they were still significantly less reliable in terms of SEMs on the

trait scales of the 16PF than those in the A, B and Cl socio-economic

groups. This trend was significant on Form A but again no differences

were found on Form B of the 16PF.

It would seem, therefore, that a distinction can be drawn between

differential reliability as a result of age and differential

reliability as a result of social class. In the case of age,

reliability coefficients were lower with older individuals as a

direct result of restricted scale discrimination with such subjects.

It is possible that older subjects have a central tendency response

style (Cuilford, 1954) and choose the middle response to the various

questionnaire items more frequently than their younger counterparts.

However, if this is the case, they do not act with any less

consistency and indeed they may act with more. The situation with

social class seems very different on the other hand. Although both

questionnaire forms showed restriction in range in lower social

class individuals, when this lack of variability was taken into

account by use of the SEM statistic, the C2, D and E groups still

appeared on Form A as the more unreliable group in terms of their

alternate form scale scores on the primary source traits. Again,

there was no significant trend on Form B. In this case the smaller

dispersion of scores in lower socio-economic classes and the

224

Q H

w x

x 0 aI

H d` O a0 CJ CO OD Co Co C " . . . " " .

M N CV in CV CV CV N

H C: ) m M V CV N N M M M

N CV N CV N N CV N

w N N in in L G l0 U-1 (X)

N N N N N N N N

N ch N (0 to Co N r1 00 Oa 1-0 N N Q'. N

N N N N N N N N

J N i9 Co i' ltd L) lý ýý

N N N N{ N N N N

Co i «e CM ýD

z pp to (0 l0

N N N N N N N N

h r-f p i H N c,

p O

N N M M N N M M

N d' CV L0 t a)

a o, ao rn w N 1o o co N N cV N

" N (N «i N

N N 01 M lo H to N ttl d' to to

(i CV N N N N N CV

OD (D d' 0 N IN M d`

x pp N N M N CO C) O1

N N N Cr1 N N N N

H M N N Q'. a) V'" ýp l(ý Ln to LO Ul

N N (N N N N N N

N r-1 N H 0) LID In O

W N O "

O O 01 rn CD N M N M M N N ("')

W d' aO

0 O

1-4 OD t0

M O

00 N 0)

N M N' N M N N N

Oý Gp [1' ýO Ci l) CD ý. uD Co N N

N N N N N 'CV N

N ap

in 00 ul

r- 1. D G% N m (n

LI)

r1 O N

O rI

O O

Oý r-1

-1 . -I

a) r-1

r" O

Q N CV N CV N N N

, aM 31

'n ý 4 U U Ä

H N om ( , 0) ü

225

C\l

H

U) U)

U a

8 U) A z Q w 0 Q

0 z H A f1' O UI Qi

£V1 W N

A

z O. G. fsr

w

0 w

lfl H M U) t0 11) rn o) rn o o, rn co rn N N N M N CM N N

OD t0 d' t() d' N d' l0 d'

N N N N N CM N N

[N N CN LN l0 to U) U) N M d' to N to dý M U)

" " " " " " " " CM N N N N N N N

cm CY) o M ) M r-{ L,; U) l p

c M U

CM N N N 0

N N N N

co '. p (p N U) N N N

,ý o rn o 0 0 o o o M N f') C'M CO M M M

C\j 04 to dý ý4' to Co V' d' mal' CM N N N N N CV N

to 0 M %44 ka . ý. H ri N H M N H O

M C') C') ('') M CM f7 M

H N H ko U) IN H N N N N N H C') H N N N N N N N N

CM U) 0 M N l0 a1 H

N N d' rI cß'7 N N M

N N N N N N N N

Oý [l' r1 rI Ql '. p N N

x o ao 0 '. o co m o 0 M N N M N N c'') cM

to 0 H M O U) lp cr H N M N N H H l

N N N N N N N N

H CO CO s r-1 O a)

0) 9

H N

N N (V) N N N N N

r-I O kO m C) l0 N N W l0 00 N U) Ný l0 l0 [ý

N N CV N CV N N N

kD N N O LO kp N OD CO N OD 00 OD co 00 N N N N N N N N

d' d` d' aD raj N

l0 0'% CO H U) %D' O' C') hfl 1l d' M U1 % ul Uf)

N N N N N N N CM

Ef) C N d l0 Ui 8 4.

+ Q U U U) r-i

u) CM

11 d'

l0 O ClU

226

TABLE: 11.3.

SEN RANK FREQUENCIES FOR SOCIAL CLASS

BOTH SEXES - FORM A

RELIABILI'T'Y

SOCIAL Low > High

CLASS 3 4

AB 4 3 3 6

Cl 2 1 7 6

C2 6 3 4 3

DE I

4 9 2 1

TABLE: 11.4.

SEM RANK FREQUENCIES FOR SOCIAL CLASS

BOTH SEXES - FORIIi B

SOCIAL

RELIABILITY Low High

CLASS 1 2 3 4

AB 6 5 2 3

Cl - 5 4 7

C2 5 3 6 2

DE 5 3 4 4

227

TABLE: 11.5.

SUMMARY TABLES: SEMs AND SOCIAL CLASS AS A MODERATOR VARIABLE

FORM A

SOCIAL RELIABILITY (SEM) CLASS Low High

Higher Social Class 10 22

Lower Social Class 2.2 10

x2 9.00 pC . 01

FORTI B

SOCIAL RELIABILITY (SEM) CLASS Low High

Higher Social Class lý 16

Lower Social Class 16 16

chi-square =0

228

TABLE: 1.1.6.

SEM RANK FREQUENCIES FOR AGE

BOTH SEXES - FORM A

AGE

RELIABILITY Low High

2 3 4

15-24 6 7 - 3

25-44 3 5 4 4

45-64 1 4 8 3

65+ 6 - 4 6

SEM RANK FREQUENCIES FOR AGE

BOTH SEXES - FORM B

AGE

RELIABILITY

Low High

1 2 3 4

15-24 2 8 2 4

25-44 4 2 6 4

45-64 5 4 5 2

65+ 5 2 3 6

I

229

TABLE: 11.7.

SUMMARY TABLES: SEMS AND AGE AS A MODERATOR VARIABLE

FORM A

AGE RELIABILITY (SEM) Low High

15-44 21 11

45-65+ 11 21

chi square = 6.25 p . 02

FORM B

AGE RELIABILITY (SEM) Low High

15-44 16 16

45-65+ 16 16

chi-square =0

230

unreliability of their responses as gauged by the SEM, suggests that

it is perhaps the irrelevant item content which is at the root of

differential reliability here. Moreover, Form A of the 16PF seems

to be the main propagater of this bias.

The foregoing seems to suggest that age per se is not as an

tmajor rr; od gyrator vzriab e on the reliability of Cattell's

primary source traits as social class. Any tendency for reliability

coefficients to be lower in older people can be accounted for by

lower dispersion of scores in such groups... Exactly why dispersion

should be lower is difficult to establish. We have noted that for

certain factors, e. g. L (suspicious), that males and females are

relatively close in mean scores at about twenty years of age,

increasingly depart from each other up to forty-five years of age

and then begin to move closer together. If sex differences were

less extreme at older ages, this would tend to decrease the

dispersion of scores. On the other hand, it could be purely a

matter of response style, older people prefering to take the safe

middle response in preference to the extreme alternatives.

In contrast, social class is seen as an important moderator on the

reliability of the source traits. Those classified in the C2, D and.

E categories not only showed restriction in range but considerable

unreliability within that range. An important point is that the test

constructor should be particularly concerned with the applicability

of items with lower socio-economic classes.

In conclusion, we must state that there was conflicting evidence

231

on the effects of age as a moderator on reliability. However, our

hypothesis that the lower social classes show lower reliability is

supported by the data collected here.

RELIABILITY AND THE PRIMARY SOURCE TRAITS: CONCLUSIONS

In this section on the reliability of Cattell's source traits, based

on a sample of approximately 2000 British adults, we have seen that

few significant differences were found between the sexes. Similarly,.

there were few differences between results reported here and those

published by Cattell in the United States. The adult sample tended

to show lower reliability coefficients, however, in relation to

university undergraduates. This was mainly as a result of less

variability in the primary source trait scale scores for the adult

group.

The most reliable of Cattell's primaries, with reliabilities. in excess

of . 5, were A (outgoing), C (emotional stability), E (assertive),

F (happy-go-lucky), H (venturesome), I (tender-minded), 0 (apprehensive),

Q3 (controlled) and Q4 (tense). Factors B(intelligence), G (superego

strength) and Q2 (self-sufficient) were of moderate reliability in

the range .4 to . 5. The remaining source traits, L (suspicious),

M (imaginative) and Q1 (experimenting) all demonstrated low reliabilities

below a value of . 4.

On the study of moderator effects in reliability, it was found that

older and lower social class subjects had lower alternate form

232

coefficients but when account was taken of restriction in range in

the scale'scores of older subjects by use of the SEM statistic, the

older age groups were possibly more reliable than the rest of the

sample. The conclusion that social class was an important moderator

on scale reliability still held after correction for restricted

variances however.

Of course, the problem of moderator effects on reliability is not

peculiar to Cattell's questionnaires and it is useful to take a

retrospective look at the data to see which particular source traits

are most prone to social class influences on reliability. The factors

most influenced were E (assertiveness), G (superego strength),

I (tender-minded), M (imaginative), Ql (experimenting) and Q3 (controlled).

On the other hand, the reliabilities of the three major anxiety

components of C (emotional stability), 0 (apprehensive) and Q4 (tense)

were least affected by the social class variable , none of them showing

any statistically significant differences.

That is not to say that other factors were unaffected by the social

class variable. We should not lose sight of the fact that other

traits with unusually low coefficients also showed a decrease in

reliability in the lower socio-economic groups. For example, although

the difference did not reach statistical significance at the 5% level,

the coefficient for Factor L (suspicious) was . 393 in the middle class

and . 261 for semi-skilled, unskilled and unemployed subjects. A

decrease in reliability which, for many practical purposes, calls into

question the meaningfulness of scores across the two 16PF forms.

233

Our final conclusion must be, therefore, that reliabilities calculated

on studen* groups give a better impression of the consistency of

scale scores across the 16PF than is the case with adult samples.

That is to say, that the power of measurement afforded by Cattell's

model as represented in his two major questionnaires is less impressive

on general population adult groups than was realised hitherto.

This conclusion makes a detailed investigation into the internal

structure of the scales, which purport to measure the primary source

traits, even more necessary and this topic will be taken up in the

next section.

I

234

SECTION 5:

INTERNAL CONSIST ENC YAND

T ,HEPRIMA

RYS0URC E TRAITS

12. INTRODUCTION

I

235

INTRODUCTION

Up to this stage our researches have involved the descriptive power

and reliability of Cattell's source traits as represented by scale

scores. We have noted that certain scales which purport to measure

the primaries at times show inconsistent relationships with external

variables and demonstrate poor reliability, especially in the lower

socio-economic classes. Clearly it is now necessary to investigate

the internal consistency and item characteristics of the model,

an area of much controversy in the past.

There are two basic problems which face the test constructor at the

outset of his enterprise. The first is to discover, choose or

otherwise define the nature and names of the dimensions he sets out

to measure. The second is to develop items which measure these

dimensions efficiently. Of course, this logical separation is not

necessarily mirrored in the actual practical business of scale

construction. One might for instance start with a set of items,

factor than and develop tentative scales and then by a process

of distillation arrive eventually at a set of scales and items

to measure than. Nonetheless, there is a logical distinction to be

upheld between dimensions or traits in a theoretical model and

the device used to measure than.

There is the paradox in scale construction that if one starts with

a set of dimensions and sets out to find items which measure then

successfully, one runs the risk of being accused of producing

entirely arbitrary scales, whilst on the other hand, to start off }

with a totally unstructured approach creates difficulties in that

236

no criteria are available for the selection of items for the initial

item pool. Moreover, if scale development begins without any

hypothetical structure, there is the risk that the emergent factors

will be obscure, difficult to name and of no practical value to

the clinician or other test user. As Nunally (1967) has suggested,

the best approach is probably to start off with hypothesised

'Jimensi, Jns, and refine pese in the course of scale development

by the use of factor analysis or some similar technique.

Where ability testing is concerned, in most applications at least,

the acid test of value is predictive validity: does the test

successfully predict performance on some criterion, and if"so,,

how efficiently(Cffinpbell and Fiske, 1959). ' It has been suggested

(e. g. by Vernon 1963) that the same is true of personality tests.

However, it is a fact that most of the controversy raging around

the question of personality measurement in fact concerns construct

validity. By this we may mean one of two things: either how

valuable the construct (ie. proposed dimension) can be shown to be

as an explanatory tool, or alternatively how well the scale developed

to measure a dimension actually does so. Where factored personality

questionnaires are concerned, notably those devised by Cattell?

Eysenck, Comrey and Guilford, the question of construct validity

in this first sense comes down by and large to the matter of

techniques adopted in the course of conducting the factor analysis.

Argument rages not so much over what sorts of items to include in

the finished inventory, indeed whatever the author, items tend to

be remarkably similar, as how to assemble them into meaningful scales.

Persistent lack of agreement on the question of which techniques to

237

employ has led some to the view that as yet factor analysis is

insufficiently understood, and that it is premature to hope that

it will provide any imediate answers to the problem of personality

measurement.

PRCJLEMM CF RF.? i, ICATING CATTELL' S FACTOR SOLUTION -', F,

In Cattell's (1964) paper "The Importance of Factor Trueness and

Orthogonality in Test Scales", direct validity is defined as the

multiple correlation of the items in a scale with the pure factor

which those items represent. If we accept this as a meaningful

definition of validity, it follows that scale validity is

maximised where: -

(1) each item has a high loading on the factor it purports to

measure

and

(2) the intercorrelation of the items in a scale is kept to a

minimum, ie. there is low homogeneity of item content.

This type of scale construction will also mean that suppressor items

can play a part in suppressing unwanted variance in an item which

loads on more than one factor. This is the concept of purifying

measurement by'means of what Cattell (1964) has termed 'buffered'

scales.

238

A number of problems arise in testing out these claimed construction

methods for Cattell's inventories. First, as Eysenck (1972) has

pointed out, other researchers have repeatedly failed to replicate

Cattell's factor structure. Thus, if our aim is to test how well

questionnaire items measure Cattell's model, then utilising loadings

from a factor analysis as a measure of 'direct' validity would not

seem viable.

Moreover, it is known that no two factor solutions are identical

each is unique either in terms of the number of factors or their

definition by factor loadings. The factor solution is in turn

based on an intercorrelation matrix of 16 variables. Such a matrix

is influenced by the sample, the mental state of the subjects and

many other variables. The concept of the 'pure' factor is rather

like that of 'true score' in the. classical test model - it is a

platonic ideal - not directly measurable but with reasonable

assumptions, it can be estimated.

Similarly, the 'pure' factor can only be estimated. Even if one

were to factor these data and 'find' 16 factors, they will not be

identical to Cattell's solution. But because many investigators

have failed to find such factors, we seem to have reached an impasse!

One way round this problem might be to undertake an arbitrary

factoring procedure (Saville and Blinkhorn, 1976) but even when this

method was tried it was not possible to replicate Cattell's solution,

The only possible solution seems to be the acceptance of Cattell's

scale scores as the delineation of the model; in other words to use

the 16PF scales as our factor estimates. If the scales are inadequate

239

as factor estimates, the only possible conclusion is that the utility

of the model is suspect. It is true to say that response styles,

attenuation and other effects may tend to alter the magnitude of

intercorrelation of scale over factor scores but the fact remains

that the scales are the vanguard of the model. If a theory exists

only within matrix algebra and allows no conversion into useful

eas tr i: E'ý; ý, than we have every right to q' estion The very value

of that theory.

It is also relevant to point out that Cattell's direct validities.

merely tell us how well the scale in the questionnaire represents

the factor in the factor analysis. As such, this tells us nothing

at all about the usefulness of the test in practical situations;

the test user does not have access to the pure factor, only to

scale scores. Any correlations between the pure factor and

performance on outside criteria will usually be mediated by the

scales as found in the questionnaire: hence for the test user, the

scales represent the factors, they must be expected to represent

them efficiently and consistently and ultimately he must judge

whether the assumption that they do so is borne out by such facts

as he has available.

Further justification for this approach, ie. using scale scores as

our factor estimates, IA " Cattell's claimed direct validities of .8 to

.9 between scale and factor scoret. This means that for many scales

the overlap with the 'pure' factor is so great as to suggest that

scale and factor are virtually identical. For these reasons it is

proposed, therefore, to ignore Cattell's strictures on the use of

240

estimates of the pure factor, since in the published inventories

there are provided means of estimating true-factor scores, namely

the factor scales themselves. If these are themselves poor representations

of the factor space, it is difficult to see what

justification exists for their development as means of

measuring personality.

THE PROBLEMS CF RESPONSE FORMAT AND CORRELATIONAL METHODS

Thus, the intention is to correlate item scores with Cattell's

personality scales as the analogue of his direct validities

within a factor analysis. We can view such correlations as item

validities. To quote Guertin and Bailey (1970): -

"The rationale for obtaining factor scores is analogous to

obtaining a total score for a person on a test that has been item

analysed to yield biserial 'r' validity indices between total score

and the items. An item-loading on a factor can be thought of as a

validity coefficient correlating the item and that reference factor

and analogous to the biserial 'r' between an item and total score

in item analysis. "

Empirical evidence for this view has been provided by Henrysson(1962)

where he showed the close similarity between factor loadings and

biserial correlations in item analysis. Unfortunately, this is not

simply a matter of calculating a large number of correlation

coefficients. Questions arise concerning the relationship of items

to scales, which must be considered before proceeding with any item

analysis.

241

The Cattell questionnaires under consideration have three response

categories for each item, generally corresponding to two extreme

choices with a middle 'unsure' or 'in between' category, with the

single exception of Factor B scales which, since they represent an

ability measure, have one right and two wrong alternatives provided.

The question arises as to whether the middle, 'uncertain' response

can be assumed to lie on the same underlying dimension as the two

extreme responses. Circumstances in which a subject might mark

the 'b' response include, besides genuinely wishing to appear as in

between the two alternatives, finding the item altogether irrelevant

or meaningless or possibly unclear and ambiguous. It may be that

in many instances the 'b' response is an escape route -a means of

avoiding a straight answer.

Thus, it might be argued that 'b' responses should be ignored as

representing 'noise' rather than 'signal'. There is certainly a case

to be made out for totally discarding items which have more

than say 80% of subjects responding in the middle category. Having

said all this, however, were we to take the course of ignoring the

middle response altogether, we should not be investigating the 16PF

as it is actually used. The item weights are given on a continuum by Cattell

and are an integral part of the construction of the-test. Furthermore,

by discarding the 'b' responses, we should be not only discarding a

good deal of useful information, but also reducing the comparability

of the coefficients in the final matrix, since different item-scale

correlations would be based on different samples. Whilst the individual

interpretation of many middle responses in the 16PF may be difficult,

Cattell has essentially decided the issue for us by including them

in his scoring scheme and eonsequently it is not proposed to discard

responses in the middle category. 242

It has been suggested that the Pearson product moment correlation

coefficient is not the appropriate statistic for item scale

correlations of the kind where three part data are involved. When

data are entered into a product moment correlation grouped into

as few as three classes, the resultant coefficient will be an

underestimate. Peters and Van Voorhis (1940) have suggested a

for Lh: e ur. &er°_. ^. timate caused by group ng; however

there are assumptions underlying the use of their correction

factor which may well not he met by item data of this kind.

Levonian (1961) suggested that the problem could be circumvented

by dichotomising the item data. Since the instructions on a test

booklet encourage the subject to respond in either the 'a' or 'c'

category, and to restrict his use of the 'b' category to one in

five items at the most, the distribution of item scores will tend

to be bi-modal, thus violating one of the assumptions underlying

the product moment coefficient. Levonian achieved dichotomised

scoring by placing the middle category in that extreme-category

with the fewer case. "This procedure tends to equalise the split,

to maximise the correlations to be computed, to allow analytical,

reproducible results and to introduce no systematic bias. "

(Levonian 1961) However, Levonian was computing inter-item

correlations, using the phi coefficient, a procedure which is not

appropriate at this stage. Furthermore, such dichotomisation does

in effect revise the scoring procedure to an extent which may

produce results whose importance is impossible to estimate. In

the case of item-scale correlations, either the point-biserial

or biserial ccefficients can be applied to dichotomised data: f

since the biserial

243

coefficient is simply an upward revision of the point-biserial,

applicable in circumstances where stronger assumptions can be made

about the underlying shapes of the distributions, it is unlikely

that the rank order of coefficients would be significantly altered

by the choice of one rather than the other. The actual decision

as to whether or not to dichotomise was made on the basis of an

e ,p rl.: al st ud'i with actual q.: estionnaire data and will be discussed

later.

When a scale score is correlated with the score. on an item which

forms part of it, whatever the nature of the scale, the correlation

coefficient which results will be spuriously high, since the item

variance is included in both the variables to be correlated. This

effect increases rapidly as the length of the scale diminishes.

In the construction of a mental ability test, this effect may

reasonably be disregarded, since the number of items in the scale

will generally be large and the effect will be more or less constant

for all items. The Cattell inventories, however, characteristically

consist of a number of short scales and we cannot allow this spurious

inflation of correlation coefficients to contaminate our data.

Furthermore, since the intention is to compare the correlation of

each item with both its own and every other scale, it is important

to avoid such confounding effects. Since in the factor loading

matrix such inflation will not arise, as all the items in the

questionnaire contribute to the factor, swamping the effect, we

must introduce some correction, so that our results are the best

possible estimates of the item true factor correlations.

I

244

The most obvious course is quite simply to correlate the scores

on a given item with scores on the appropriate scale minus that (the item remainder method).

item scoreA The objection to this procedure is that each item will

be correlated with a slightly different scale score and one,

furthermore, which is composed of one fewer items than the true

scale. To overcome these objections, Henrysson (1963) proposed

; Dme now formulae to provide for the estimation of the correlation

of an item with the same total composite of items, with the

spurious overlap effect removed.

For the biserial correlation the formula is:

rb(H) = rbSt - (pq/y)

1S+ pq - 2t rbSty

where: rb = non adjusted biserial between item and scale

rb(H)= corrected biserial coefficient

St = standard deviation of scores on test t

p= proportion passing item

q= proportion failing item

y= ordinate in the unit normal distributions corresponding to p

Wolf (1967) attempted to evaluate empirically the effect of these and

other formulae devised for this purpose. Under conditions of varying

test length it was in fact found that the results obtained by the

Henrysscn formulae agreed perfectly in all cases with the item

remainder method. It, therefore, seems that there is nothing to

choose between the two methods, and consequently, for reasons of

245

convenience, the item remainder method has been chosen here.

Cattell has written in scathing terms of the efforts of those

adopting an itemetric approach to questionnaire structure. By taking

each item in turn and examining its correlations with its own and

every other scale, we may well overlook the fact that the scale may

be 'Hewed not only item by item but also as a structured whole.

As the source traits measured by Cattell's scales are in any case

oblique, ie correlated,. we should expect to find significant

correlations between items and scales other than those on which

they are keyed. In some cases, these correlations may become

disturbingly large. It is, however, a feature of these-scales that

such complications are catered for at the time of construction.

We have already noted Cattell's claims that 'suppressor' action

has been built in to the scales, so that a substantial correlation

between an item and a scale on which it is not keyed is compensated

by the correcting effect of the relations between other items and

that scale. That is to say, where as the individual items in the

scale may show overlarge correlations with other scales, the scale

as a whole is a balanced composite largely free of spurious

correlations.

Interest in suppressor action began with the search for suppressor

tests to partial out unwanted correlations and their effects.

However, as Levine (1952) reports, these proved difficult to develop

and more recently interest has tended to centre around suppressor items.

Lord and Novick (1968) have commented on the fact that few suppressors

246

have nonetheless been found, and Conger and Jackson (1972) conclude

that new predictors will always prove easier to find than new

suppressors. Lubin (1957) showed that suppressor weights are even

more fickle than ordinary regression weights when cross validation

is attempted.

Ra 1e han plunge into te intricacies of whether a rar : icular item

has or does not have suppressor action, the simplest way to assess

whether scales are working in the way they are intended to is to

compute the multiple correlations of the items, constituting the

scale with that scale and with every other scale. If all is well,

the multiple correlation between a set of items and the scale they

constitute will be higher than their correlation with any other scale.

This is the empirical analogue of Cattell's 'direct validity',

viz. the correlation of a scale with the pure factor found in the

factor analysis. Since without replicating in every detail the

factor solution derived by Cattell, we cannot gain access to the pure

factor, for purposes of calculating this correlation, we must accept

the questionnaire scale as the best estimate of this factor.

At this point the importance of using item remainder rather than

ordinary item-scale correlation coefficients emerges. Were the item

scale coefficients not corrected for inflation due to overlap, the

multiple correlation of the items in a scale with that scale.. would.

necessarily be 1.00, since all the variance in the scale score is

to be found in the item scores. However, by using item remainder

coefficients, we obtain the best unbiased estimate available of the

item true factor correlations, which then serve as a basis for

247

calculating the multiple correlation between the items and the scale.

Of course, when it comes to calculating multiple correlations between

sets of items and scales in which they are not included, the item

scale overlap problem does not arise.

It is important to realise that comparatively little weight will be

attached in the ensuing discussion to the absolute size of the

item scale multiple correlations. Our arguments will rest largely

on the relative size of the coefficients. Thus, the error inherent

in the use of the factor scale as an estimate of the pure factor

does not affect our argument, since this error will affect all the

multiple correlations. Quite simply, we shall look for the highest

multiple correlations between sets of items and scale scores. As

in any case the multiple correlation coefficient tends to overestimate

the true correlation, any depression of these values as a result of

the procedures we have adopted is not likely to produce serious

distortion of the relative size of coefficients, or indeed their

absolute size. We shall not, therefore, concern ourselves with

corrections for 'shrinkage' and the like.

Our requirement will be quite simply that for a scale to be acceptable,

it should be seen to have a higher multiple correlation with its own

items than with the items constituting any other scale. This

requirement is lenient in the extreme. Because we will be calculating

the scale score analogue of Cattell's factor direct validities, we

are merely asking that items in a scale measure their own trait

better than they measure another factor.

i

248

THE HSPQ STUDY

We have seen that items of the type where subject responses are

expressed in terms of a choice from three possible preferences

have been analysed by one of three methods:

i1C_h: '.: JiiLl. o. _. CJ ifn}

data .ii

such a 1731 ,%

that i the eiddl e responses \viý. um r1..

(designated the 'b' response in Cattell's inventories) are

included under the smallest of the two extreme responses,

using the biserial'correlation coefficient to represent

the association between item and scale.

(b) dichotomising the item data as in (a) but using the point-

biserial coefficient in order to correlate item and scale.

(c) analysing the item data directly as coded by the subject

without dichotomising data, with use of the conventional

product moment coefficient.

In order to investigate whether conclusions on item adequacy may

in any way be affected by these different approaches, item data

were collected from British standardisation of another of Cattell's

inventories for 14-17 year old school pupils, the High School

Personality Questionnaire (HSPQ). The background to this study

has been described by Saville and Finlayson (1973) and involved

the administration of the HSPQ to some 2000 secondary school pupils.

These data provide a useful experimental base to investigate these

different methodologies, especially as 12 of the 14 primary source

249

traits in the IHSPQ are common to adults and the questionnaire

espouses identical construction principles to those of the 16PF

The second major aim of this study was to investigate the effects

of overlap between personality scales and the various items loading

on that scale. Thus, in addition to biserial, point-biserial

and pr, --duct rnomer. t coefficients being calculated between items

and the total scale scores, the corresponding coefficients

corrected for spurious overlap between scale and item were also

computed. The corrected coefficients were based on an iterative

procedure which involved calculating the correlation between item

and scale with that item removed from the relevant. scale score.

THE SAMPLE

A total of 366 female pupils was sampled by the fixed interval

technique from the British standardisation group (Saville and

Finlayson, 1973) by taking every third record.

PROCEDURE

An outline of the item analysis procedure is contained in Figure 12.1.

RESULTS

Appendix l2contains the results based on dichotomising the HSPQ

item data and includes the corrected for overlap point-biserial

correlation of each of the 140 items with its own scale. In addition, I

250

W

WU O a a

r' ä N

W

(cý H

U, w a ý A H öw HG',

H T' o0

H3 äc1) Q+ý QWU

0

U Öq

w: ý 0u c6 .

EA OH

01. un ü3

HHý ý r :Z

G] xW

kt H

Ö0

Ua cm W a U)

H

w vi HW H ý/1

xö w

aä

cr AH cz x wý

ma > to

E-4 11

z

= E-4 2

HH

ä ýý

zö HQ

HHA +ý

10 O°b Umaa

OH

H. EA OH ý7

UW

ä

U)

ýzq

W Ef) Ö

H ul Or

H HC,

Ll F-! U1

Q fk

. ý? äi rCf ýj EH

C) L4 -4 W HO c71 Wpý

W Ht0 VÄHÖ

(a

"d

ÖE 12 HR

WÖHx

A 14

ý E-1 Z C) ýAI

O2 xO

VAH3CGfý..

dN

pö", ýö

u u2

w Hwo zaH

cý x HÖ WOa

R3. HWö

ýWýC7zýHCi

Q

E--f ul n4 4.1 0

xx HHHäý äaä

w Ö bx

0% Hx H

E-f 3 2:

W Ö

«1

UcH3 fý W

ý-

251

the corrected biserial coefficients and the biserial coefficients

between each item and the remaining thirteen HSPQ primary scales

are included in the table.

The corresponding data based on product moment coefficients without

dichotomising item data are contained in Appendix 12. For both

sets of tables, coefficients under a solid line are negative in value.

A further analysis conducted on these data was to correlate the

corrected dichotomised. biserial with the point-biserial coefficients.

The results of this analysis are contained in Table 12.1. below.

As is to be expected from the mathematical relationship between the

two coefficients, their intercorrelation is very high.

TABLE: 12.1.

Cc*RELATION BETWEEN THE POINT_BISERIAL AND BISERIAL APPROACHES TO

HSPQ ITEM SCALE CORRELATIONS

POINT-BISERIAL BISERIAL ITEM-SCALE ITEM-SCALE CORRELATION CORRELATION r

Mean . 167 . 215

. 99 S. D. . 092 . 117

Whilst the point-biserial correlations between an item and its owrr

scale are lower than the corresponding biserial coefficients, the

point-biserial of an item with an external scale are similarly lower.

It is quite clear therefore that as long as we are not concerned with I

252

the absolute values of the correlation of an item with its own

personality scale, the point-biserial and biserial statistics give

us virtually identical results.

If we adopt the biserial coefficients in that the item data are not

in any real sense a true dichotomy, it is possible to summarise

nu m bcýr of high`r extra t can i. n ra scale correlations for e ach

factor. For example, we can see from Table 12.2. that the first

item on Factor A correlated more highly with its own scale than

any other scale and the second item showed one correlation which

was higher with an outside scale than its own corrected correlation.

However, item 7 on Factor A showed twelve higher extra scale

correlations; in fact, it was only with Factor B that this item

had a lower coefficient than its own scale correlation. The

correlation with all of the remaining twelve scales in the HSPQ were

higher than the correlation upon which it is loaded.

Turning to the second set of tables in the Appendix, containing the

item analysis results using the product moment coefficient on the

item data in its original three point response format, we can see the

very marked effect of correcting the coefficients for spurious overlap.

This is no doubt due to the shortness of the HSPQ scales and their

heterogeneity.

Table 12.3. provides summary statistics for the number. of higher extra

than intra scale correlations for each factor, based on the product

moment approach. We can see that the overall relationship with the

corresponding biserial table is high; Factors A (outgoing), I

253

N

a

m

a

L.

U W cl: O

V)

C7

U-

ýJ dW

ýO O C. ) 4.

LLI

Ent J Q

cy- V

CD t- i- H In. Lu

0' =

1-1 Lý J F- c"C X

LU O Nw a w C-

NJ "Q

N

W JS 00' C7' Qw k- 2

w

ýI S

U.. O

ix 1. l. ß m

. rte..

-41 q7' co r ct CV d- ; f* CO 4 C L() r..

C") O G*' ct' d' r C) N N. O lO (V

CV O'

N r- r N r r M In Lfl CV. v

p ) (

O L[) V-- M O Cl Q1 CV cm -M 'rr d' r--

ID ch o O M aT c LO O N. tD r-. , --

ý-+ Ul d' O r U7 O N O O s. ý r `y- N

L) O r r CJ C3. fit': LC) CV CV N nj N

C3 r O r Cl O O N ¢}' p t, O I-

Li.. r- N N N O O C) -y- to r-. r

ivy

W N 00 . 4- Q1 N Ql CO r r-- Cif M p r '. p

O. LO O cr) cr O r O N . p tO qtr

U M O r e- cd- N. r- tD c, N. CD cat

m w O O r Q O O O C-D tV to

Q O r- r '4) M O N M , ._ ^ w N

W hN

°ý1-4 SW 31I ' N d Z: OCH -j L+-1 '+-+ Li U. 1 IAJ

ca : _C .JC: LU 4a . .., -cri Q

m : r: zv)

C) 4-

0

fu E rt D N 4

f

254

w 0

I-

C

CC 0 I--

U

w

LL

(/) to JZ wo

C3 CC

wJ

rr_ CD C) L)

w

Itl Q WU º-4 N CZC zM

än LLi

I-- H ¢ ýc

w 0

ws a c3

._ <n o C) Ll- NO r-I CJ

"N W LJ JZ co m CD Z:

ý-- sz

C+- kp N cV ' C; r N M N O CV) z: - co r M

M. O r l+) ý- d- CQ CO cV lD O (3 tV C

CV d' CV) (-j Cl) r- r- C. 7 r LO cr. ) C)

O U) e- C') O CO Co. N Cri C9.. .. M i . r-

,

7 d' tC O ýF c M to N% Ln cr) Q

º-ý CV d- O O O N C) C) O Cý) to r N

= %0 O .Y r C7 O Cf' Rj C`) C4 N CV)

(3 O O O N O C) d' d' O r"s C+)

L. r- r r CV O . O C) O c M

W CV Q1 O CO CV O C) e- r OD O O r r"- r-- N.

Cl U) ý-- r U) O O C) N p U") r-

U N O r- ,M r,. U) N In O ' .. 01 CV

CM

m (V O O r O O O O O N L[)

M Cr) CD lfl CV) (V) r- N r- r- M

O N

a+ W ED C

O I L, C) ct V W N

SW3II N =ý-- (/) ý ct: LLJ (n ýw H-)-O

1- C) C)

I

255

C (emotional stability), E (assertive), H (venturesome),

I (tender-minded), J (circumspect individualism), 0 (apprehensive),

Q2 (self-sufficient), Q3 (controlled) and Q4 (tense), all showing a

relatively large number of items with high extra over intra scale

correlations.

Whet: the biserial and product moment methods are compared over both

items and factors, it can be seen from Tables 12.4. and 12.5. that

the overall association is high and we may conclude with a fair

degree of confidence that the two methods of item analysis give

essentially comparable results.

I

256

TABLE= 12.4. THE HIGH SCHOOL PERSONALITY QUESTIONNAIRE: FORM A, GIRLS

CORRELATION BETWEEN PRODUCT MOMENT AND BISERIAL APPROACHES (ALL ITEMS)

PRODUCT MOMENT BISERIAL NUMBER OF HIGHER NUMBER: OF HIGHER

EXTRA THAN INTRA SCALE N EXTRA THAN INTRA SCALE CrýRitýt ,' TLO: Y CORRE1 ATICNS CORRELAýE IONS. .

MEAN S. D. MEAN S. D.

3.03 3.21 140 3.02 3.16 . 871

PRODUCT MOMENT BISERIAL NUMBER OF HIGHER NUMBER OF HIGHER

EXTRA THAN INTRA SCALE N EXTRA THAN INTRA SCALE CßRRELATI01`ý CORRELATIONS -CORRELATIONS - MEAN S. D. MEAN S. D.

30.29 17.65 14 30.21 16.20 . 970-

r

'_I

257

13. ESTIMATING THE DIRECT VALIDITIES

OF THE PRIMARY SOURCE TRAITS

258

INTRODUCTION

From the studies on the High School Personality Questionnaire,

we have established that the biserial and point-biserial

approaches after dichotomising the item data, give very

co"' Arab' e ri sui Ls c the use of the product moment coefficient

on the trichotomised data.

Because of this finding and in that the trichotomised data after

all represent Cattell's original response format, it was decided

to use the product moment technique on the adult. item data,

correcting the statistic for spurious overlap between item and

scale.

A further point has to be borne in mind however. Cattell (1972)

advocates the use of multiple correlational techniques in

calculating the composite correlations of his items with the

factor estimate. It is the case, however, that such techniques

are likely to give the upper bound to scale validities mainly

as a result of their tendency to capitalise on chance effects in

the data. What is needed is not the direct validities calculated

on optimal weights but on the weights as they actually exist in

the item data. Thus, the values '0', '1' and '2' were assigned

to the item responses as laid out by Cattell in his 16PF scoring

key and thlse used to calculate the multiple correlation of sums

between items and scales.

I

259

That is: -

res rci I

dcl �- ii II

Where: -

rci = correlation corrected for overlap between the item xi and the scale (i. varies from 1 to n)

0= standard deviation of the item

rij = correlation between one item xi and any

other component xj, when j is a higher subscript number than i.

In summary, we are to calculate the multiple correlation of sums1

between items corrected for spurious overlap and the relevant

personality scale. These values we shall term composite item-scale

correlations. The scales are taken as factor estimates in order

to establish how well the items measure the underlying model.

This approach, it has been argued, is the analogue in item analysis

terms of Cattell's direct validities and is to all practical purposes,

the only method by which the internal structure of Cattell's model

can be investigated. With this in mind and in the light of Cattell's

claims, a number of hypotheses can be tested.

HYPOTHESES

(1) The composite item-scale correlations, representing direct

validities, will tend to be of the order of .7 to . 9.

(Cattell, 1972) }

260

(2) Because of the method of scale construction used in the 16PF

(Cattell, 1972), which deliberately sacrifices item homogeneity

in favour of suppressor action, these composite item-scale

correlations will tend to be higher than those derived from

conventional homogeneity formulae such as coefficient alpha.

(3) After the principle of 'buffered' scales advocated by Cattell,

the composite item-scale correlations (direct validities) for

any given scale will be higher than the intercorrelation of

that scale with another. (If this is not so, then it suggests

that the items in one scale measure another factor rather

better than they measure their own scale. In such a case,

there can be no real justification that the scale in question

exists as a measurable separate identity. )

DIRECT VALIDITIES OF THE SOURCE TRAITS: PROCEDURE

The responses to each item on Forms A and B of the 16PF for the

adult group (N = 2008) were punched on to cards and scored by

computer according to the keys provided by Cattell to produce

an item score matrix. Scale scores were then created on each of

the primary source traits for each of the two forms of the 16PF.

Item responses were then correlated by the product moment statistic

with the relevant 16PF Form A+B scale, having corrected that scale

I

261

for spurious overlap. ( These data are held separately from the main

thesis. ) The multiple correlation of sums of these values was then

calculated. This procedure was followed for the male and female

groups separately (Table 13.1. ).

DIRECT VALIDITIES-01F. THE SOURCE TRAITS: RESULTS AND DISCUSSION

The first observation we can make regarding the table of composite

item-scale correlations. concerm our hypothesis on their absolute

size. In fact, the data are comparable to the multiple correlations

of scale its with the pure factor as presented. by Cattell (1972),

(see Table 13.1). It should be remembered, however, that Cattell's

values are based on Form A only whereas our item composite

correlations are for two form length (Forms A and B). The only

major difference between the two separately derived estimates

is for Factor I, where Cattell's value is somewhat higher than

the corresponding data reported here.

We can see from the table that the primary source traits of

highest reliability are F (happy-go-lucky), H (venturesome),

0 (apprehensive) and Q4 (tense). Low internal consistency estimates

were obtained on the other hand for Factors L (suspicious),

M (imagination), N (shrewdness) and Ql (experimenting). These

findings are very much in line with those reported in the previous

section where the parallel form method was employed. That is to

say, that the primaries of low reliability as measured by the

alternate form approach are precisely those factors which show }

262

TABLE: 13.1.

COMPOSITE ITEM-SCALE CORRELATIONS (DIRECT VALIDITIES)

FOR CATTELL' S PRLN! ARY SOURCE TRAITS

COi iPOSITE ITE:, i_SCALE COEFFICIENT MULTIPLE R OF " f CORRELATIONS ALPE=. SCALE ITEP; S

FACTOR i

MALES FEMALES MALES FEMALES FAWITH E

CTORU(CATTELL)

A . 601 . 532 . 626 .. 554 . 62

B . 581 . 561 . 599 . 579 . 60

C . 673 . 653 . 695 . 675 . 54

E . 665 . 646 . 687 .. 667 . 53

F . 786 . 763 . 814 . 790 . 76

G . 628 . 575 . 654 . 599 . 69

H . 840 . 848 . 871 . 880 . 79

I . 591 . 454 . 616 . 471 . 73

L . 471 . 506 . 488 . 525 . 44

Ni . 374 . 446 . 385 . 459 . 51

N . 369 . 364 . 382 . 377 . 40

0 . 723 . 706 . 748 . 730 . 65

Qi . 448 . 424 . 465 . 439 . 47

Q2 . 570 . 549 . 593 . 570 . 57

Q3 . 660 . 622 . 688 . 647 . 64

Q4 . 791 . 750 . 819 . 776 . 73

corrected for two form length by the Spearman Brown formula

t

263

low internal consistency, either in terms of homogeneity or by

the 'direct validity' approach.

Referring back to Table 13.1, we in fact notice that the item-composite

correlations are in every case, for both the male and female groups,

lower than the alpha homogeneity estimates. This must lead us to

reject our second ht, -pot; esis that suppressor action in the 16PF

scales would lead to higher composite item-scale correlations

than would be obtained by conventional internal consistency methods.

We are forced to the conclusion, therefore, that no suppressor

action of any marked extent is present in the 16PF scales and

that the internal consistency reliabilities of the primary source

traits are not underestimated by conventional techniques, as has

been claimed by the proponent of the scales.

Despite Cattell's theoretical standpoint on the production of

scales with balanced specifics, the evidence reported here suggests

that suppressor items are fickle in behaviour and that an author

would generally do better to consider the addition of more predictor

items in a scale. It may be the case in Cattell's theory of

personality that the liking for suppressor action is merely a

reflection of the fact that it has been found extremely difficult

to construct simple predictor questionnaire items for some of the

primary source traits. This problem might well have arisen, of

course, as a result of Cattell transferring a model of- personality

from his studies on ratings which included inadequate or arbitrary

traits.

264

The third hypothesis in this section concerned the question of

buffered scales; the principle of items suppressing unwanted

variance due to the influence of other personality traits. The

vital test of this claim is the size of the composite item-scale

correlations in relation to the factor scale intercorrelations.

If a set of items has a composite item-scale correlation which is

ýJ'výý than that scal'ýýS re.. ationship with amothei pYi_rn ry, it

in effect means that: -

(a) the items measure another scale better than they measure

their own scale.

(b) the construct validity of the model is in doubt because

the scale in question is not de-nonstratably a separate

identity from other scales.

Table 13.2.. represents the intercorrelation matrix of the 16PF factors

as scale scores, based on the British adult group, the matrix for

males being above and for females below the main diagonal.

Information from this and a previous table have been included

in Table 13.3-which shows the item composite correlation for males

and females with the highest intercorrelations of the relevant

scale with the other primaries.

It is evident from this table that the items of Factor C (emotional

stability) measure Factors 0 (apprehension) and Q4 (tension) as

well, if not better than they measure their own scale. Indeed

these three factors, although of tolerable reliability, all show }

265

TABLE: 13.2.

CORRELATIONS BETWEEN THE SOURCE TRAITS OF THE 16PF AS SCALE SCORES

FORMS A AND B

A B C E F G H 1 L M N 0 Q1 Q2 Q3 Q4

A -07 08 14 31 02 42 -22 01 -02 -03 -12 04 -43 01 -04

B -04 17 15 07 -05 09 -05 -03 28 -14 -21 22 02 03 -06

C 11 20 06 08 11 41 -14 -37 12 11 -70 02 -12 42 -66

E 10 10 01 48 -15 45 -18 38 32 -44 -16 43 -10 -15 09

F 27 17 12 50 -33 50 -15 30 -07 -42 -09 32 -42 -30 12

G 04 05 09 -18 -30 08 -04 -14 -15 29 -15 -31 -03 60 -24

H 34 17 40 49 58 -01 -03 00 18 -12 -51 16 -44 22 -31

1 08 15 01 -17 -07 10 01 -08 25 05 18 -08 04 -07 12 U'

z L -01 -07 -41 38 19 -09 -01 -16 03 -30 31 19 04 -29 44 äo

w M -10 23 18 34 19 -10 33 18 00 -22 -12 33 08 -11 -04

N 09 -08 09 -41 -42 28 -19 04 -24 -28 -13 -38 05 38 -28

0 -15 -14 -67 -13 -14 -09 -48 07 28 -18 -08 -07 12 -50 71

Qi 07 11 -01 38 32 -23 25 -05 22 27 -31 -08 00 -25 10

Q2 -31 -12 -13 -06 -43 -04 -39 -04 07 00 07 07 00 00 06

Q3 05 09 36 -20 -31 56 09 02 -26 -12 37 -39 -20 02 -56

Q4 -12 -03 -64 09 11 -15 -27 05 40 -07 -24 69 05 03 -48

FEMALES (N=717)

266

M

oý N

Vjl

HI

HQ'.

9j

z H

tn Ö

H I

U

H

U)

l

cr Q ýD 11

C- LD

Cl) Cf 0 U') CV)

N

a

o m

N mo t O t o

11

z

z

a

H

M

U O 'o 110 N

co (f

C\i M 1i' d' M

11

cr1 t0 to to l0 co c1 Ln r-4 11 N N t! 1

" U)

" ýO Q

" L-

" ul

" CO

" Iq d'

" in ýn

" " m

" " d'

" in

" .o

" CD

" Hj V) H 0 E-4 P4

O p N

ºý O N N p)

L- c7 l0

dý 00

rn to

N N M

N M

cV N

to r' in

lo tD N

H

Ö v

4 w u W w 0 x H a z 0 ä

267

such high intercorrelations that their existence as separate

identities seems doubtful in the extreme. This possibility has

been suggested previously by Eysenck (1972) and Saville & Blinkhorn

(1976), where as a result of their researches into student

samples, the C, 0 and Q4 primary source traits seemed to form

one underlying vector.

Two other factors, unassociated it would appear with the anxiety

primaries considered above, but which show intercorrelations about

as high as their own direct validities are G (superego strength)

and Q3 (social-precision).

The picture is a little more confused with the remaining primaries

of L, M, N and Ql. Factor L (suspicious) shows intercorrelations

almost as high as its direct validity estimates with the two anxiety

factors of C (emotional stability) and Q4 (tense) and with

E (dominance). Similarly, Factors M (imaginative) and Ql (experimenting)

show a strong enough association with the Factor E primary source

trait, as to make one wonder as to their functional independence.

Finally, the data demonstrate that Factor N (shrewdness) is a

glaringly weak factor, with the items of F (happy-go-lucky),

G (superego strength) and Q3 (social precision) having higher

correlations with the scale than the items in the scale itself.

Thus, there is the nonsense that the best estimate of this

particular source trait is provided not by itself but by three

other scales.

} 268

DIRECT VALIDITIES OF THE SOURCE TRAITS: SUMMARY AND CONCLUSIONS

We have seen that the direct validities calculated here on two-form

length scale scores are about as high as Cattell's claims for Form A

only. We have utilised the statistical methods advocated by Cattell

with the exception of dealing with scale scores rather than the

more esoteric factor estimates. This approach was adopted because

of the great difficulties of replicating Cattell's factor structure

and in the belief that scale scores are in any event more scientifically

defensible criteria against which to validate items.

The values obtained for scale score direct validities were comparable

to the alternate form reliability coefficients reported in an earlier

section; particularly weak factors being L, M, N and 0-1. We found

no evidence to suggest that the alternate form reliabilities severely

underestimate the reliability of the primary source traits (Cattell, 1972,

C1 Krug personnt . communication, 1977).

When the role of suppressor action in the 16PF scale scores was

investigated, the direct validity estimates showed no tendency to

be higher than alpha statistics calculated on the item data. We are

forced to conclude, therefore, that the suppressor action as a means

of enhancing scale validities is not operating within the primary

source traits, as represented by Cattell's latest questionnaire forms.

Moreover, it was clear from these analyses that certain primary

factors were untenable as separate identities. Namely, the traits

of C (motional stability), 0 (guilt prone) and Q4 (tense) seemed i

269

to form one anxiety constellation and G (superego strength) and

Q3 (social precision), a measure of the extent to which an

individual adheres to conventional social standards.

The position with a number of further 16PF primaries was more complex.

Factor M (imagination) appeared to have'a good deal in common with

Factor E (dominance), as did Q1 (experimenting) and N (shrewdness).

The N trait also had a good deal in common with Factor F, so as to

exhaust most of its meaningful variance. Finally, Factor L (suspicious)

was highly associated with both the anxiety dimensions and E (dominance).

270

14. FACTOR ANALYSIS AND THE PRIMARY SOURCE TRAITS

271

INTRODUCTION

There would seem considerable evidence here to suggest that Cattell's

model of-personality as represented by the 16PF first order factors

is open to serious criticism, especially as we could find little

support for Cattell's contention that the primaries all represent separate

identities.

Because, moreover, no suppressor action of any great influence was

discovered within the item structure of the 16PF and because the

scales were, therefore, very much to be judged by their homogeneity,

there can be no real objection to a conventional factor analysis of

the primary source traits as scale scores.

It is beyond the scope of this thesis, however, to construct an

alternative personality theory to that of Cattell or the other major

factor theorists. The literature is strewn with many one-off analyses

which add little to a more general understanding of the structure of

personality. There is also the point that it is inappropriate

here to become heavily involved in the controversial issues of the

type of factor analysis best employed.

Nevertheless, it is feasible to test whether as few as three factors

are sufficient to account for the 16PF traits; for example, are the

three dimensions of the Eysenck Personality Questionnaire (Eysenck

and Eysenck, 1976) sufficient to explain Cattell's primaries?

For this purpose promax factor analyses of the 32 scale scores of

Forms A and B of the 16PF was undertaken on the male (N = 1004) and

272

female (N = 1003) samples collected for the purposes of these

researches.

Promax (Hendrickson and White, 1964) involves first rotating to an

orthogonal varimax solution and then relaxing the orthogonality

of the factors to better fit simple structure. From the orthogonally

rotated matrix, an ideal factor matrix is constructed in which high

loadings of varimax are made higher and low loadings smaller. This

is achieved by normalising the orthogonal matrix by rows and columns

and taking, in this case, the fourth power of each loading. Finally,

the least squares fit to the ideal matrix is found using Hurley

and Cattell's Procrustes technique.

Because of the nature of the hypothesis. that as few as three factors

will account for Cattell's primary source traits, a relatively

conservative criterion of the number of factors to extract was

adopted here. The criterion was that of limiting the factors to

those with eigenvalues greater than one (Kaiser, 1960).

273

TABLE : 14.1.

16PF FACTOR ANALYSIS: MALES:

EIGEN VALUES AND PERCENTAGE OF TRACE

FACTOR EIGEN VALUE

PERCENTAGE OF TRACE

CUMULATIVE PERCENTAGE

1 5.6249 17.58 17.58

2 4.9684 15.52 33.10

3 2.5615 8.01 41.11

4 1.9918 6.22 47.33

5 1.6511 5.16 52.40

6 1.2370 3.87 56.36

TABLE: 14.2.

16PF FACTOR ANALYSIS: FEMALES:

EIGEN VALUES AND PERCENTAGE OF TRACE

FACTOR EIGEN VALUE

PERCENTAGE OF TPACE

CUMULATIVE PERCENTAGE

1 5.2104 16.28 16.28

2 4.9571 15.49 31.77

3 2.1967 6.87 38.64

4 1.9962 6.24 44.88

5 1.6954 5.30 50.18

6 1.2906 4.03 54.21

7 1.1149 3.48 57.69

274

TABLE: 14.3.

PROMAX FACTOR ANALYSIS: MALES: TABLE OF FACTOR LOADINGS

OBLIQUE PRIMARY FACTOR LOADINGS FIRST ORDER 1 2 3 4 5 TRAIT

A 0.6873 -0.2582 B C 0.8345 E -0.8141 F -0.3226 0.5100 -0.3027 G 0.9366 H 0.2832 -0.4938 0.5041 I -0.7739 öL

-0.4453 -0.5825 M -0.3899 N 0.4395 O -0.7494 Q1 -0.4639 -0.2966 Q2 -0.7673 Q3 0.2976 0.6337 Q4 -0.8688

A 0.6303 -0.3266 B C 0.6988 E -0.8619 0.2033 F -0.3747 0.4240 -0.2852 G 0.7970

czý H 0.4111 -0.3892 0.4500 I 0.2322 -0.7710 09 L -0.4005 -0.5420 0.2089

44 M -0.4364 -0.4572 N 0.2185 0.3124 0.2676 0 -0.7732 Q1 -0.2700 -0.2364 Q2 -0.2132 -0.7699 Q3 0.3739 0.6190 Q4 -0.8170

6

0.8743

0.3121

0.8668

0.2821

f

275

TABLE: 14.4.

PROMAX FACTOR ANALYSIS: FEMALES: TABLE OF FACTOR LOADINGS

FIRST ORDER TRAIT

A B C E F G H I L M N O Q1 Q2 Q3 Q4

A B C E F G H I

OL P4 M

N 0 Q1 Q2 Q3 Q4

OBLIQUE PRIMARY FACTOR LOADINGS

234567

0.7304 -0.7959

-0.7854 0.731j7 0.3094 0.5232 -0.2310

0.9093

-0.2363 0.6126 0.3822 -0.7455

0.4600 0.5066 0.2194 0.1825

-0.2441 0.3360 -0.4630 -0.4677 -0.2577

0.7478 -0.2177 0.2132 -0.2118 0.2531 -0.4180

-0.7436 -0.2496 0.6291

0.8472

-0.6391 0.8278 0.3575 0.4670 -0.2470

0.7847 -0.4227 0.5005 0.3186

0.2273 -0.7462 0.3400 0.4412 -0.2147 0.2018

0.6303 -0.3406 -0.2024 -0.2925 0.2944

0.8231 -0.2052 0.5525 -0.3029 -0.3145

-0.7873 -0.3722 0.5769

0.7935

0.2804

0.3277

0.7948 -0.8176 -0.2734

-0.2187 0.2287

0.25 34

276

TABLE: 14.5.

CORRELATIONS OF THE OBLIQUE PRIMARY FACTORS

MALES

1234 5 6

1 01 16 -08 32 23

-24 03 41 -32

3 06 -07 -10

4 -03 -18

5 -18

TABLE: 14.6.

CORRELATION'S OF THE OBLIQUE PRIMARY FACTORS

FEMALES

1234 5 6 7

1 01 -21 21 -25 18 -08

2 24 03 -42 -22 05

3 -10 -15 -09 15

4 -08 15 26

5 00 10

6 -04

I

277

FACTOR ANALYSIS: RESULTS AND DISCUSSION

The promax factor analysis of the 16PF scale scores produced a

six factor solution for the male groups. Factor loadings are

coma r. ßä in 1abic 14.3. -and the correlations between the oblique

primary factors and the percentage of the variance which they accounted

for are contained in Table 14.5.. The corresponding data for the

female group are contained in Tables 14.4. and 14.6..

If we first turn our attention to the results, based on the male

sample, we can see that the first major factor to be extracted

from the matrix and accounting for some 17% of the variance,

appears to be a general anxiety vector with high loadings on

C (emotional stability), 0 (guilt prone) and Q4 (tense). The

second and third factors appear as two extraversion-introversion

factors; the second factor having high negative loadings on

E (assertiveness) and to a lesser extent, L (suspicious) and

H (socially boldness). The high factor loadings for the third

factor are with Q2 negatively (group dependence) and A positively

(outgoingness). Also loaded on this factor are F (happy-go-lucky)

and H (social boldness).

}

278

Factor four is largely Cattell's I (tough minded) with some

relationship with M (practical). As we might have expected from

our previous section, the fifth factor is largely G (superego strength)

and Q3 (social precision). Finally, there is an ability factor

highly loaded by B (intelligence) with smaller loadings on

M (imagination).

Turning to the factor analysis undertaken on the female group,

we find that seven factors had eigenvalues greater than unity

and that there is a good deal of agreement with the male study.

Factors one and two seem largely to be an anxiety and assertiveness

factor respectively. Factors four (tough mindedness), five (superego

strength) and six (intelligence) are also closely in line with

results found for the male group.

The important departure between the two studies is on the third

factor in the male group. In males this factor has high loadings

on A, F, H and Q2, whilst in the female sample A (outgoing) seems

to form a separate factor.

The results of these factor analyses are interesting in that they

do not support the hypothesis that only three factors are required

to explain Cattell's primary source traits. It would appear that

a minimum of six to seven and perhaps a maximum of nine are needed.

If we are'to accept the trait labels provided by Cattell for these

studies, the following structure evolves from the promax factor

analyses. I

279

PROVISIONAL FACTOR NAME

I Anxiety

II Dominance/Assertiveness

III Group Adherence

IV Tender-mindedness

V Superego Strength

VI Intelligence

and possibly

VII Outgoingness/warmheartedness

HIGH PRIMARY SOURCE TRAIT LOADINGS

C- O+ Q4+

E+ L+ H-F

Q2- F+

I+ M+

G+ Q3+ Q1-

B+

N-

A+

Some confusion exists as to the dimensionality of the extraversion

components of the 16PF but it would certainly seem that Cattell has

managed to break this factor down into meaningful sub-factors, even

if not as many as he would claim. Clearly what is needed in future

research is to further investigate this domain using more sensitive

factor analytic techniques, perhaps working on item statistics rather

than scale scores.

CONCLUSIONS ON THE FACTOR ANALYSES

The results of the factor analyses conducted on the male and female

samples produced highly comparable results to the calculation of direct

validities on the 16PF scale scores. By using a promax technique, six

factors provisionally named anxiety, assertiveness, group dependence,

tender-mindedness, superego strength and intelligence with the

possibility of a seventh factor of outgoingness were extracted.

From all the evidence presented in this section, we are forced to

conclude that Cattell's primary source traits are not the most

parsimonious description of personality but that perhaps as many

as seven to nine factors are required to explain the domain as

represented by Cattell's questionnaire scales.

280

}

SECTION 6:

SUMMARYANDC0NCLUS10NS

15. SUMMARY AND CONCLUSIONS

281

The general hypothesis developed in this research was that Cattell's

theory of personality rests on a system of primary source traits

which is untenable in terms of the verifiability of important group

differences, the reliability of measurement which it affords and

its internal consistency. It was also hypothesised that despite

the existence of some independent, reliable and descriptively useful

factors, Cattell's model, as represented by his primary questionaire

scales, is not as parsimonious as he has previously claimed.

It was also pointed out in the introduction to this thesis that

there is unlikely to be a once and for all test of Cattell's theories.

As Howarth (1976) has stated "what is now needed is not a system

defence but a number of investigations to form a 'concensus buttressed

by co-ordinated studies".

What evidence have these studies provided then on the existence and

value of Cattell's primaries as identifiable constructs? We shall

seek to answer this question by considering each of the primary

source traits in turn, observing how well the model actually fits

the data base of some 2000 British adults collected here. We shall

also take the opportunity at this stage to paraphrase the 16PF items

which best define Cattell's primaries. For the

purpose of this section, items which have reached an arbitrary

criterion of a correlation of . 25 or more with the scale it was

designed to measure are included. Items, which are paraphrased here

merely for reasons of copyright, are expressed in this format where

an affirmative response corresponds to the positive pole of the

primary source traits.

282

FACTOR A (Reserved vs. Outgoing)

1, GROUP DIFFERENCES

I. I. Cross National: - British adults-tended to be significantly

lower on this factor than American adults.

1.2. So,, D rferonces: - As was hyaothesised, the female sample

obtained a significantly higher mean score on this factor.

1.3. Social Class: - As hypothesised from Cattell's theory, no

significant differences were found between the social

classes on this factor.

1.4. Age: - No evidence was found for Cattell's claim of an increase up

to 30 years with subsequent levelling off. A relatively flat profile

of differences was in fact found across. the age cohorts.

There was no evidence of any curvilinear relationship of

this primary with age.

2. RELIABILITY

2.1. Group Values: - The alternate form reliability of this factor

was approximately . 5, with significantly higher reliability

in male over female subjects.

2.2. Social Class & Age as Moderators: - Reliability was found. to be higher

in the higher social classes. Age did not act as 4L significant

moderator here.

I

28 3

3. INTERNAL CONSISTENCY

3.1. Estimated Scale Direct Validities: - Values of . 60 ,

(males)

and . 53 (females) were found as estimated direct validities

for this scale. Coefficient alpha was . 626 and . 554 for males

and females respectively.

3.2. Items Defininct the Factor: -

Item Number Form A

51 I would rather be a secondary school teacher

than a forester.

Form B

26 At work I would rather be in charge of

interviewing people than mechanical matters.

51 I would rather be a hotel manager than a

research chemist.

52 1 enjoy selling things or asking for funds.

53 I'd rather be a business manager than architect.

176 I'd prefer being in charge of children than

helping a watchmaker.

3.3. Factor Analyses: The varimax factor analyses of the female data,

where A (Outgoing) emerged as a separate variable gives support

to the view that this trait could conceivably be a first order

factor of a second order extraversion/introversion factor. In

males, however, this primary source trait was absorbed into a

more generalised extraversion measure. We can see from the items

defining this particular trait that they are in the main concerned with

284

occupational preferences and it may well be that the social

stereotype of women being attracted to people orientated jobs is

responsible for differences in the factor solution across the sexes.

It is also possible that this factor is mainly a 'thing (job)'

versus 'people' orientated discriminator. It is not altogether

surprising, therefore, that Cattell has claimed that Factor A

is important in distincruishing h& we n, for ? Xd: i l, sci nt is is

and sales persons. The content of high loading items on this

primary does not, suggest that this factor is necessarily

closely associated with softheartedness, adaptability and

easy-goingness. There is also the considerable danger that

this factor's apparent power to distinguish between

occupational groups is based upon teleological causes.

FACTOR B (Less Intelligent vs. More Intelligent)

1. GROUP DIFFERENCES

1.1. Cross-National: - Whilst results were not significant with

the female groups, the British males were found to be significantly

higher on this factor than their American counterparts.

1.2. Sex Differences: - On this factor males were found to achieve

a significantly higher mean score than the female group.

1.3. Social Class: - As was hypothesised, the higher social classes

tended to achieve higher scores. This was a relatively strong

trend. I

285

1.4. Age: - There was a significant negative correlation between

this primary source trait and age for both males and females.

Tests of the departure of linearity of regression were also

significant, scores showing a more rapid decline after 40

years of age than they had shown thereto.

2. RELIABILITY

2.1. Group Values: - This was a factor of relatively low

reliability of approximately . 44.

2.2. Social Class and Age as Moderators: - Some evidence was found

for lower social classes and higher age groups to have lower

reliability on this factor.


3.1. Estimated Scale Direct Validities: - The estimated scale direct

validities for this factor were . 58 for males and . 56 for females.

The corresponding alpha statistics were . 60 and . 58 respectively.

3.2. Items Defining the Factor: - Most of the adequate items on

this particular scale were of the conventional verbal analogous

type. Three items, however, (103 and 152 on Form A and 177 on

Form B) showed negative item scale correlations.

3.3. Factor Analyses: - Despite the relatively low reliabilities of

this factor, it still emerged as a separate identity accounting

for some 4% of the variance. In addition, this trait generally

behaved consistently with regard to group differences hypothesised

from Cattell's model and the estimated scale direct validities

suggest that with the exception of Factor M, this primary is

286.

relatively untouched by the other 16PF variables. Not surprisingly,

these researches suggest that Factor B (Intelligence) is an

important and independent variable within Cattell's 16PF. Its

low reliability, however, only makes it suitable for use with

group rather than individual comparisons.

FACTOR C '(Affected by Feelings vs. Emotionally Stab-)


1.1. Cross-National: - In line with the hypothesis, no significant

differences between British and American subjects were found

in this factor.

1.2. Sex Differences: - The female group were found to be significantly

lower than males on this particular trait. This was contrary

to the hypothesis of no difference between group means.

1.3. Social Class: - There was a significant positive linear trend

between emotional stability and social class in this sample.

This finding is in line with the prediction made from Cattell's

description of his primaries.

1.4. Age: - No evidence was found for a slight positive relationship

between this factor and age.

2. RELIABILITY

2.1. Group Values: - The group reliability coefficients. for

Factor C were approximately .5 with no significant differences. }

between the male and female subjects.

28 7

2.2. Social Class & Age as Moderators: - Neither social class nor

age were seen as significant moderators upon the reliability

of this particular scale.


3.1. Estimated Scale Direct Validities: - The estimated composite

item scale correlations obtained for this study were . 67 for

males and . -65 for females. This compares with alpha coefficients

of . 70 and . 68 respectively.

3.2. Items Defining the Factor: -

Item Number Form A

55 1 have hardly ever been let down by my friends.

79 People do not seem to ignore or avoid me.

80 People do not treat me unreasonably.

129 After looking forward to something, I

always, feel like going.

154 My sleep is not disturbed by vivid dreams.

Form B

41 never feel 'shut in' when in a small space.

5I never find it difficult to put trivial troubles

out of my mind.

80 The way people say things never hurts me.

130 I am ready for life and what it demands.

154 I never find it difficult understanding some

people because of their unusual use of words.

179 1 do not want a more sheltered life.

t

288

3.3. Factor Analyses: - This primary was highly loaded on an-

anxiety factor extracted from the intercorrelation matrix of the

32 Form A and B scale scores. Direct validities estimated from

the data were about as high as the intercorrelation of this scale

with Factors C and Q4. From the evidence collected here, this

trait would not seem to constitute a separate identity from the

o, -. her anxiety vsriaUles within the y PF.

FACTOR E (Humble vs. Assertive)


1.1. Cross-National: - Whilst there were no differences between

the female American and British groups, the British males

were significantly higher on this factor than the average

American male.

1.2. Sex Differences: - As was hypothesised very marked sex

differences were found on this primary source trait; males

being significantly more assertive than females at every

age group.

1.3. Social Class: - No significant correlation was found between

this factor and social class.

1.4. Age: - This factor showed significant negative linear

correlations with age in both groups. For the male sample

there was a slight but statistically significant departure

from linearity. This was in line with the hypothesised age

289

differences although the decline would appear to commence

somewhat earlier than 40 years of age as has been proposed

by Cattell.

2. RELIABILITY

2.1. Group Values: - The alternate form coefficient between Forms

A and B of the l6PF was . 57 for the total sexes but somewhat

lower for the single sex groups.

2.2. Social Class & Age as Moderators-- When account had been taken

of restriction in range across the various groupings, no

important differences in reliability were found across either

age or social class.


3.1. Estimated Scale Direct Validities: - The composite

item scale correlations for this factor were . 66 for males

and . 65 for females. The corresponding alpha reliability

estimates were . 69 and . 67 respectively.


Item Number Form A

61 do not hold back from criticising others.

7I make clever, sarcastic remarks to others.

56 In some ways I feel definitely superior to others.

81 Foul language does not disgust me, not even in mixed company.

106 1 am best described as 'forceful. '.

290

156 1 am pleased when I become the one in command.

180 I almost always put forward ideas.

181 1 am good at meeting challenges.

Form B

6I tend to speak out when another person's

reasoning is in error.

31 1 do not avoid saying things that embarrass

others.

32 I would not feel nervous with a loaded gun in my hand.

57 It is important to get your ideas into practice.

106 I usually tell a superior when my opinion

differs- from his.

131 I am more energetic and ambitious than many

equally successful people.

3.3. Factor Analyses: - In both the male and female groups Factor E

had high loadings on a second oblique primary factor, which

accounted for some 15% of the variance. It seemed to have

little in common, for example, with Factor A and appears to

constitute an important primary within Cattell's exvia

second order factor. Group differences were generally in

line with those hypothesised on this particular factor.

I

291

FACTOR F (Sober vs. Happy-go-Lucky)


1.1. Cross-National: -Group differences between the American and

British subjects were not significant either for males or

females on this primary source trait.

1.2. Sex Differences: - Similarly, there was no significant

difference between the sexes.

1.3. Social Class: - Whilst there were no significant product

moment correlation coefficients between this factor and social

class, there was some slight tendency for eta coefficients

to reach statistical significance; the higher social classes

scoring more highly on this primary.

1.4. Aqe: - The trend in this factor is for a very pronounced

negative correlation with age for both sexes.

2. RELIABILITY

2.1. Group Values: - The Factor F primary source trait was seen

to be one of the more reliable variables in Cattell's

personality model; the reliability being estimated at

approximately . 65 for both sexes.

2.2. Social Class & Age as Moderators: - No systematic variation

was found as a result of age or social class acting as

potential moderators on the reliability of this factor. I

292


3.1. Estimated Scale Direct Validities: - The estimated direct

validities for this primary were . 79 for males and . 76 for

females which compare with values of . 81 and . 79 respectively

by the alpha statistic.


item Number Form A

33 I am rated as an amusing talker.

58 I go out for entertainment more than average.

82 I have more friends than most people.

107 I do not avoid social functions.

132 I spend a good deal of my time talking

with friends.

133 I like doing daring things for fun.

157 I like dressing with eye-catching style.

158 I prefer a lively party to a quiet hobby.

182 I am considered an enthusiastic person.

183 1 like jobs which offer change and travel.

Form B

8 I often tell amusing jokes.

33 I enjoy playing practical jokes.

58 I enjoy the principle of 'laugh and be merry'.

83 I like being in exciting situations and bustle.

107 I do not avoid embarrassing topics when

talking with the opposite sex.

132 1 would like more excitement.

182 I tend to act on impulse.

183 1 eng best described as 'happy-go-111cky'.

I

293

3.3. Factor Analyses: - In the varimax factor analysis, this

source trait did not emerge as an independent factor but

rather was absorbed into Assertiveness (E+) and Group

Adherence (Q2-). This factor also appeared to have some

negative loading on a factor well defined by Cattell's G

(Superego Strength). It would seem, therefore, that whilst

this is a relatively reliable trait, it is not independent

of certain other extraversion/introversion variables within

the 16PF.

FACTOR G (Expedient vs. Conscientious)


I. J. Cross-National: - Contrary to the hypothesis of no significant

differences between the two nations, American adults achieved

significantly higher results on this factor.

1.2. Sex Differences: - No significant differences were found

between the sexes on this variable.

1.3. Social Class: - With regard to social class, no clear

relationship was found.

1.4. Aqe: - A strong positive correlation with age was found in

both males and females.

294

2. RELIABILITY

2.1. Group Values: - This source trait was of moderate reliability

with an estimated value of approximately . 47.

2.2. Social Class & Age as Moderators=- When account was taken

of differences in variability between the sub-groups, no

consis ter: L c-lid: -nc.. :,; es found for age and social class

acting as significant moderators on scale reliability.


3.1. Estimated Scale Direct Validities: - The estimated scale

direct validities for this primary were . 63 for males and

. 58 for females with corresponding values of . 65 and . 60

by the alpha method.


Item Number Form A

34 1 get annoyed when I see untidy people.

134 The sight of an untidy room annoys me.

160 I always refer to the rules of right and wrong.

184 I insist on always doing things correctly.

Form B

134 It is good to plan and avoid wasting time.

184 1 do not like seeing disorder.

f

295

3.3. Factor Analyses: - This trait, with Q3 (social precision),

emerged with high loadings on the fifth factor extracted

from the correlation matrices. The serious question emerges,

however, whether the items which define the factor can

in any real sense be regarded as measures of superego

strength. Rather they would seem to centre on tidiness

of dress, order and perhaps a love of structure. It is

not surprising then that this variable shows a considerable

increase over age; adults, after all, generally become less

accepting and more critical of the dress of others as they

mature in age. Cattell (1972) has claimed that G is

negatively. related with. creativity and it may well be

that the variable is merely a liking for orderliness and

routine in daily life. In any event, if we are to judge

this factor by the adequate items within Factor G, there

can be little reason for viewing this source trait as

representing superego strength within the generally

accepted definition of the term.

FACTO1 H (Shy vs. Venturesome)


1.1. Cross-National: - American females were seen as higher on

this factor whilst there were no differences between the

American and British males.

1.2. Sex Differences: - As was hypothesised, males scored =

significantly higher on this factor.

296

1.3. Social Class: - With regard to social class, a positive

relationship was found in the female sample but no

significant trend was discovered for male subjects.

1.4. Aqe: - The results of the age analysis for this primary

source trait showed a relatively flat pattern of differences

across t tz, six ag_ cohC . rts. The hypothesis of a positive

relationship between this source trait and age was rejected.

2. RELIABILITY

2.1. Group Values: - This was the most reliable of Cattell's

primary source traits with an estimated alternate form

coefficient of . 73.

2.2. Social Class & Age as Moderators: - Some evidence was found

that age was a significant moderator on the reliability of

Factor H; older age groups (65+ years) tended to have higher

standard error of measurements than younger subjects. No

significant effects on reliability were found, however,

with social class.


3.1. Estimated Scale Direct Validities: - The composite item

scale correlations for this factor were . 84 for males and

. 85 for females. The Cronbach alpha statistics were

calculated as . 87 (males) and . 88 (females) on Forms A and

B of the 16PF.

I

297


Item Number Form A

10 I come forward at social occasions.

35 I like being the focus of attention in a group.

36 I like to join a gathering of people.

60 1 do not keep quiet in the presence of

senior people.

61 I do not find it difficult to speak to a

group of people.

86 I am not content to let others do the

talking in a small group.

110 It is easy for me to mix with people at

a social event.

111 I can use tact and persuasion to get people moving.

135 I am a sociable, outgoing person.

136 I show my feelings in social relationships.

161 1 do not mind people watching me at work.

Form B

10 I have been active in organising social groups.

35 I am not unsure of myself socially.

36 I like getting into conversation with others.

60 I am not troubled by feelings of inferiority

in my social relations.

61 I do not feel awkward in company.

85 I am not embarrassed by people watching

me in a shop.

86 I avoid joining in a conversation with

other people.

298

110 I usually succeed in impressing people

favourably.

135 I am usually involved with what is going on.

136 I fit into a new group immediately.

161 It is not difficult for me to start

conversations with strangers.

186 1 am not bothered by shyness.

3.3. Factor Analyses: - Like Factor F, this variable has loadings

on a number of different factors extracted in the male and

female analyses. It seems to be related to anxiety,

assertiveness (negatively) and group adherence (A+, F+, Q2-).

We can see, therefore, that despite its very high reliability,

this scale is unlikely to be uni-dimensional and is related

to a rather complicated cluster of extraversion/introversion

and perhaps anxiety components.

FACTOR I (Toughminded vs. Tenderninded)


1.1. Cross-National: - In both male and female groups, American

subjects achieved a mean score which was significantly

higher than that of the British samples.

1.2. Sex Differences: - This is one of the major discriminators

between the sexes and in line with the hypothesis, females

attained a considerably higher mean score than males in the

British sample.

299

1.3. Social Class: - No differences were found between the social

class groups in the male sample. There was, however, a

significant positive relationship between this variable

and social class for the female sample.

1.4. Aqe: - Some support was found for the hypothesis of a low

positive relationship between this primary source trait

and age.

2. RELIABILITY

2.1. Group Values: - The reliability coefficient calculated on

the total sex group was. approximately . 55. However, the

sex of the respondent was a significant moderator here

with males and females showing reliabilities of . 47 and . 32

respectively.

2.2. Social Class & Aqe as Moderators: - Age had no significant

effect as a moderator on the reliability of this primary

source trait. Whilst there were differences in the

reliability coefficients for the various social class groups,

(the higher social classes showing higher reliability), there

were no significant differences between the groups when

account was taken of differences in variability of scores.



direct validities were . 59 for males and . 45. for females.

The corresponding alpha statistics were . 62 and . 47

respectively. i

300


Item Number Form A

11 I would rather be a playwright than a

construction engineer.

62 1 do not have a good sense of direction.

87 I prefer reading a novel to accounts of

hat-11 es

112 Being a guidance counsellor appeals to me

more than an engineering manager.

138 The beauty of a poem appeals to me more

than that of a well-made gun.

Form B

87 T do not tend to be interested in

mechanical matters.

112 T would rather be a philosopher than a mechanical engineer.

163 I would rather watch a concert artist

than a program on new inventions on the

television.

3.3. Factor Analyses: - This scale was highly loaded on factors

extracted from both the male and female intercorrelation.

matrices by the promax technique. Other 16PF variables

with sizeable loadings on this factor includedA (negatively)

and L (negatively) in males and L (negatively), M (negatively)

and Ql (positively) in females. Perhaps the similarity of the

items defining the I and A primary source traits should not go

without mention. Both possess a high vocational content with

particular emphasis on practical versus people orientated jobs. I

301

FACTOR L (Trusting vs. Suspicious)


1.1. Cross-National: - Despite the hypothesis of no significant

differences between the two nations, British adults had a

significantly higher mean than their American counterparts

for both male and female subjects.

1.2. Sex Differences: - A significant difference in favour of

males was found in this primary.

1.3. Social Class: - With regard to social class, a negative

relationship was found in females, whilst there was no

significant association in the male group.

1.4. Agee: - This primary source trait showed the strongest non-

linear association with age; males showing a slight decrease

in scores over age group, whilst females demonstrated a

decline in scores until about 50 years of age where they

began to increase again.

2. RELIABILITY

2.1. Group Values: - Cattell's Factor L was one of the less

reliable primary source traits with an approximate alternate

form reliability coefficient of . 31.

2.2. Social Class & Age as Moderators: - Age had no significant

effect as a moderator in the reliability of this trait. I

There were significant differences between the reliability

302

coefficients of different social classes. However this

effect was largely due to differences in variability of sums

across the various social groups.


3.1. Estimated Scale Direct Validities: - When Cattell's direct

Vc iid±tties were eS L' mate C1 by the composi'e i -err, scale

method, values of . 47 for males and . 51 for females

were obtained. This compared with alpha values of . 49

and . 53 respectively. When these data were compared

with the scale intercorrelation with other factors, the

items in Factor L appeared to measure the C. (negatively).,.

E (positively) and Q4 (positively) source traits almost

as well as they measure themselves.


Item Number Form A

113 If a person is selfish, I will show him up.

114 1 do not tend to make silly remarks in fun.

Form B

Adopting the criterion of item remainder first order

scale correlations of . 25 and over, there was not one

adequate item in Form B of the 16PF to measure this

variable. The best item with the value of . 203 was

No. 64 - People would like to see me in trouble

i

303

3.3. Factor Analyses: - As to be expected from the direct validity

estimates, this variable did not emerge with particularly

strong loadings on any of the oblique primary factors.

This trait appeared to be related to anxiety (positively),

assertiveness (positively) and tendermindedness (positively)

and as such seems an unreliable amalgam of different

variables within Cattell's personality model.

FACTOR M (Practical vs. -Imaginative)

1. GROUP DIFFERFACES.

1.1. Cross-National: - In both the male and female groups, British

adults attained a significantly higher mean score than that

of Cattell's American standardisation sample.

1.2. Sex Differences: - Contrary to the hypothesis,

males were higher on this factor than female subjects.

1.3. Social Class: _ A significant positive relationship was

found between this variable and social class for both the

male and female samples.

1.4. Age: - Results on this source trait indicated relatively

small differences across age groups. Product moment

coefficients were generally non-significant, as were

statistical tests for the significance of departures

from linearity. 3

304

2. RELIABILITY

2.1. Group Values: - This was another factor of low alternate

form reliability with an estimated value of . 27.

2.2. Social Class & Aqe as Moderators: - Reliability estimates

tended to be particularly low in the lower social classes

and in bigger ace groups,

although both of these trend-

appear to have their origin in differences in variability

of scores.



direct validities for this factor were . 37 for males and

. 45 for females. The corresponding alpha values were . 39

and . 46 for males and females respectively. It was also

noticeable that the items in this source trait measured

E, H and Ql almost as well as they measured their own scale.

3.2. Items Defining the Factor: - None of the items in either

Forms A or B reached the criterion of a composite item

scale correlation in excess of . 25. The best items,

however, were as follows: -

Item Number Form A

90 I dislike the way some people stare at

others. (r - . 22)

141 One should be careful about being with

strangers because of problems of disease, etc.

(r = . 22) }

305

Item Number Form B

15 1 dislike completing routine jobs (r = . 24).

65 It is more important to be concerned about

the meaning of life than making a good income.

(r = . 24)

3.3. Factor Analyses: - From the factor analyses of the male and

female intercorrelation matrices, this trait had negative

loadings on anxiety in females, positive loadings on

assertiveness for both sexes, positive loadings on

tendermindedness for both sexes and positive loadings

independence for males. The overall picture for this

factor, especially in view of its very low reliability,

is extremely complex. From these analyses, however, it

would seem to be a combination of emotional stability,

an interest in artistic matters and ideas and high

independence. We must question, therefore, whether this

is in fact an independent factor and whether the practical

versus imaginative label is the most appropriate term to be

applied to this collection of very heterogeneous items.

FACTOR N (Forthright vs. Shrewd)


1.1. Cross-National: - British females were significantly higher

than American females on this factor, whilst there were

no significant differences for the male samples.

I

306

1.2. Sex Differences: - Contrary to the hypothesis, females

exceeded males significantly in terms of their group

mean.

1.3. Social Class: - Despite Cattell's claim that this variable

is highly related to social class, no clear relationship

f12 S found .

1.4. Ace: - This factor showed a moderately strong relationship

with age with no significant departure from linearity.

This was in line with the hypothesis developed from

Cattell's model.

2. RELIABILITY

2.1. Group Values: - Along with Factor Ql, this primary source

trait was the most unreliable variable with an estimated


2.2. Social Class & Age as Moderators: - Some evidence was found

for Factor N to show lower reliability coefficients in the

unskilled working class but this effect was mainly as a

consequence of lower test score variance in these groups.

No relationship was discovered between age and the

reliability of the scale.


3.1. Estimated Scale Direct Validities: - The estimated direct

validities of the scales were . 37 for males and . 36 for

females. These values were in relatively close agreement

307

with alpha statistics of . 38 for both male and female groups.

When these internal consistency estimates were compared with

intercorrelations of Factor N items with other scales,

there was in fact a higher negative correlation with

Factors E, F and Q3.

3.2. Items Defining the Factor-- Only one item (No. 42 in Form A)

reached the criterion of an item scale correlation exceeding

. 25. Two other items did in fact exceed . 20, however, and

these are listed below.

Item Number Form A

41 I do not feel like taking part in tough

physical activity.

42 1 prefer polite people to rough individuals.

Form B

67 I get many unusual ideas which are too

many to put into use.

3.3. Factor Analyses: - The major loading of this trait is

negatively on a factor best defined by E (Assertiveness)

on the 16PF. In. females, however, there is also a negative

loading on a general social extraversion factor defined by

Cattell's F and H primary source traits. The results of

the estimated scale direct validities, factor analyses and

the reliability studies suggest that this is a variable

already contained within the extraversion/introversion domain.

I

308

FACTOR 0 (Self-Assured vs. Apprehensive)


1.1. Cross-National: - No significant differences were found

between the American and British female groups, although

in males, American adults were significantly higher.

1.2. Sex Differences: - As was hypothesised, British females

scored higher than the male respondents in this survey.

1.3. Social Class: - Slight negative association with social

class was found for this variable.

1.4. Age: - When Factor 0 was analysed by age cohorts, a relatively

flat profile of mean scores was found.

2. RELIABILITY

2.1. Group Values: - This primary was one of the more reliable

source traits with an estimated value of . 61.

2.2. Social Class & Age as Moderators: - No significant differences

in the reliability of this scale for different age or social

class groups were found.


3.1. Estimated Scale Direct Validities: - The internal consistency

of Factor 0 estimated by the composite item scale correlation

on items with the scale score was . 72 for males and . 71 for j

females. Alpha statistics achieved a value of . 75 and . 73

309

respectively. When these values were compared with the

intercorrelation of the factor with the remaining primaries,

they were found to be about as high as the scale's correlation

with C and Q4.


I ter Number Form A

18 I occasionally have a sense of danger

which I cannot. understand.

43 I feel very down when I am criticised in

a group.

44 I fear I have done something: wrong when

called in by my boss.

68 Often I do not feel in the right mood to

see anybody.

93 I feel down if acquaintances show they

dislike me.

118 I often feel that I will be punished even though I have done nothing wrong.

168. I am not regarded as a solid, undisturbed

person.

Form B

18 I can be overcome by thoughts of worthlessness.

43 My friends do not need me as much as I need them.

44 I worry if people think badly of me.

68 My spirits are not generally high.

69 I worry about an event at night.

93 1 feel overcome when small things go wrong.

I

310

94 I am disturbed by guilt feelings.

118 1 become afraid that my happiness will not last.

119 1 occasionally feel depressed and miserable.

143 When catching a train, I get tense and anxious.

3.3. Factor Analyses: - Cattell's 0 (Apprehensive) loaded highly

o ',

-be first, and largest t_ Factor from -ý '-he

prom ax sC,

This was clearly a general anxiety factor where Factor C

and Q4 also loaded highly. From the scale direct validities

and the factor analyses., there would seem no reason to

suppose that 0 is in any way different from the C and Q4

primary source traits in Cattell's model.

FACTOR Q1 (Conservative vs. Experimenting)


1.1. Cross-National: - As was predicted from Cattell's model of

personality, there were no significant differences between

the American and British groups on this factor.

1.2. Sex Differences: - Males achieved significantly higher scores

on this primary source trait. This result was in line with

the hypothesis developed in an early section of this thesis.

1.3. Social Class: - No significant relationship was found between

this variable and social class.

t

311.

1.4. Age: - Despite Cattell's hypothesis that this factor shows no

clear relationship with age, these results showed a very

strong negative correlation both in the male and female groups.

2. RELIABILITY

2.1. Group Values: - This particular primary source trait was

one of the most unreliable of Cattell's variables with an

estimated alternate form coefficient of . 25.

2.2. Social Class & Age as Moderators: - The reliability coefficients

of this factor showed significant differences between the

social groups, higher social grades showing more marked

consistency of scores across the two forms. It was

also noted that coefficients tend to be lower in the

older age groups. However, when account was taken of

differences in variability of scores by the use of the

SEM statistic, no significant trend was found.

3, INTERNAL CONSISTE2ICY


direct validities were found to be . 45 for males and . 42

for female subjects. The corresponding alpha statistics

were . 47 and . 44 respectively. These estimates were about

as high as the intercorrelation of this primary with

Factor E (Assertiveness).

3.2. Items Defining the Factor: - No items reached the criterion

of having a correlation of . 25 or more with the scale..

312

However, five its exceeded the value of . 20 and these

are listed below.

Item Number Form A

70 As a teenager, I kept to my own opinions

if I differed with my parents.

95 I would rather have a 'oh with a large salary

which meant my persuading people that I. was

worth it.

145 In a heated argument between people, I

like to see one side or the other win.

Form B

46 In a newspaper I like to see discussion.

of basic social issues.

95 I disagree with the idea that people should

get together in order to worship regularly.

3.3. Factor Analyses: - This source trait had loadings on the

factor which we have provisionally entitled 'Superego Strength'

(G+, Q3+). Sizeable loadings were also evident on Assertiveness

(E+) and negatively on Tendermindedness (I+). In all, this

seems a relatively unreliable factor which has variance in

common with a number of other source traits. It certainly is

poorly described by Cattell in terms of its relationship with

external variables such as age. Whilst it is difficult at

this stage to say whether Cattell's QI is in any way comparable

to Eysenck's social attitude of radicalism, we have seen from

these analyses that the trait as measured by 16PF items does

not appear to be independent of the remaining primaries.

313

Further, clarification and definition of this trait is

tequired, hopefully with far more reliable items than those

employed in Cattell's inventories.

FACTOR Q2 (Group-Dependent vs. Self Sufficient)

1. GROUP DIFFERE['JCES

1.1. Cross-National: - Despite the hypothesis of no difference

between the two national groups, both the male and female

British samples achieved significantly higher mean scores

on this factor than their American counterparts.

1.2. Sex Differences: - British males were significantly higher

than females on this source trait.

1.3. Social Class: - No significant relationship with social

class was found with this variable.

1.4. Acre: - The general trend is for scores to increase on

Self-Sufficiency after adolescence until about 20-25 years

where they level off until about 55 years. There was a

significant positive relationship between Q2 and age in

males but only a slight association for female subjects.

2. RELIABILITY

2.1. Group Values: - This primary source trait was only of

moderate reliability with an approximate alternate form


314

2.2. Social Class & Age as Moderators: - There was some evidence

that reliability coefficients tend to be lower in higher

age groups and in the lower social classes. This trend

was largely eradicated, however, when account was taken of

differences in variability of scores. The one exception

to this was in Form B of the 16PF where older subjects

demonstrated higher standard errors of measurement.

3. INTERNAL CONSISTENCY: -

3.1. Estimated Scale Direct Validities: - The direct validities

estimated by the composite item scale correlations were

. 57 for males and . 55 for females. These corresponded

to values of . 59 and . 47 respectively based on alpha

internal consistency values.


Item Number Form A

97 I do not like to take an active part in

social affairs and committee work.

122 I would rather work on my own than with a committee.

171 I learn better by reading a good book on the topic than by joining a group.

Form B

47 Books are more entertaining than companions.

t

315

3.3 Factor Analyses: - This source trait had high negative

loadings on a third factor extracted form the promax factor

analyses. The other traits with high loadings were

F (positively), H (positively) and in the case of males

A (positively). Unlike F and H, this source trait did

not load heavily on a factor provisionally called

'Assertiveness' and as such seems to be a purer measure

of group dependence or perhaps group adherence. From

our analyses it would seem that Q2 along with E measure

two sub-components of extraversion/introversion which show

only low intercorrelation.

FACTOR Q3 (Undisciplined Self-conflict vs. Controlled)


1.1. Cross-National: - No significant cross-national differences

were found on this factor.

1.2. Sex Differences: - The adult male sample achieved a

significant higher mean-score on this source trait.

1.3. Social Class: - This variable showed a significant non-linear

relationship with social class. Mean scores tended to be

high for the higher social classes, to decrease in the middle

class, skilled and unskilled working class and to increase

again in the unemployed and retired (Group E).

}

316

1.4. Age: - Results showed a relatively strong, positive linear

relationship with age but that little evidence of any

levelling off at 40 years as has been postulated by Cattell.

2. RELIABILITY

2.1. Group Values: - The estimated alternate form reliability

cf this factor vas . 5'_, it tending to be significantly

higher in male rather than female subjects.

2.2. Social Class & Age as. Moderators: - The reliability

coefficients for this factor were considerably higher

in higher social class groups but there was little

appreciable difference within the age cohorts. Little

difference between either the age or social class groups

was noticed however when the SEM statistic was used

to investigate reliability.



direct validities were . 66 for males and . 62 for females

with corresponding alpha values of . 69 and . 65. These

estimates were approximately as high as the intercorrelation

of the scale with Cattell's Factor G (Superego Strength).


Item Number Form A

23 When looking in the-mirror, I rarely get

confused as to which is right or left.

317

48 I keep my room well organised.

98 I am never satisfied until minor details

are given close attention.

123 I never have times when I feel sorry for

myself.

147 I do not allow my actions to get influenced

by jealousy.

Form B

23 I do not like to leave things to chance.

24 I go about my life in a deliberate,

methodical way.

73 When I need to make a quick decision, I

rely on calm logic.

147 I do not tend to get over-excited and rattled

easily.

148 I am not absent-minded.

173 1 cannot really be described as a dreamer.

A most interesting anomaly occurs with item 98, Form B of

the 16PF2which correlates -. 22 with the Q3 scale. Inspection

of the item (I make a fuss if I am left out of something by

others) suggests that the score of this scale has in some

way been reversed with respect to the underlying factor.

We are in the position, therefore, that removal of this item

will tend to increase the reliability of the scale.

3.3. Factor Analyses: - This source trait tended to have some

positive loading on the first oblique factor of anxiety

extracted from the male and female correlation matrices.

318

Its major loading, however, was on a factor provisionally

called 'Superego Strength' where Factor G was also highly

loaded. It would seem, therefore, that this variable is

largely Cattell's G with some anxiety variance. This conclusion

is very much supported by the list of items defining the

factor which is provided above.

FACTOR Q4 (Relaxed vs. Tense)


1.1. Cross-National: - British female subjects were significantly

higher on this factor than their American counterparts,

whilst there were no significant differences between the

male groups.

1.2. Sex Differences: - As was hypothesised, females were

significantly higher on this factor.

1.3. Social Class: - With regard to social class, no differences

were found for male respondents but there was a significant

negative relationship with social class in the female group.

1.4. Age: - The hypothesis of slight, negative relationship on.

Factor Q4 with age was in fact found in this data.

2. RELIABILITY

2.1. Group Values: - The alternate form reliabilities for this I

factor were relatively high with a value of approximately . 66.

319

2.2. Social Class & Age as Moderators: - Neither age nor social

class were seen as moderators on the reliability of this

scale.


3.1. Estimated Scale Direct Validities: - The estimated scale direct

validities were . 79 for males and . 75 for females. The

corresponding alpha statistics were . 82 and . 78 for males

and females respectively. These tended to be about as high

as the intercorrelation of this scale with Factors C (negatively)

and 0 (positively).


Item Number Form A

25 I find it difficult to calm down when

something makes me angry.

49 I sometimes feel tense when I think

of the day's happenings.

50 I sometimes feel that others are not interested in what i am saying.

74 I find that I am upset by criticism.

75 I find it difficult to keep my feelings

under control.

99 Small setbacks can irritate me too much.

124 I often get angry with people too quickly.

149 I get tense when I think of the things

in front of me.

174 Trivial things can get on my nerves.

I

320

Form B

25 I often feel restless without knowing why.

49 Certain sounds can set my nerves on edge.

50 I often feel tired when I get up in the

morning.

74 Useless thoughts tend to stray through

my mind.

99 I can be easily distracted from working.

125 If I am annoyed, I must let off steam.

149 A near accident or argument can leave

me shaky..

150 1 often feel that I am boiling up inside.

174 My emotions can run away with me.

175 I frequently feel so furious, I could break a window.

3.3. Factor Analyses: - This primary source trait had high loadings

on the general anxiety factor extracted from the correlation

matrices. As previously indicated, there would seem to be no

reason to suppose that this variable is in any way different

from C (emotionally unstable) and 0 (apprehensive).

I

321

CONCLUSIONS ON GROUP DIFFERENCES

In order to summarise results on group differences in personality,

Tablel5-7. shows the relationships between Cattell's primary source

traits and the sex, social class and age of the adults sampled in

this study. This information has been organised under the seven

factor solution originating from the promax factor analysis of

the female intercorrelation matrix. The factors have been

provisonally named from the Cattellian variables which have the.

highest factor loadings.

It is not suggested that these seven factors are in any real sense

the true definition of the personality domain. Factor analysis,

utilising a relatively conservative criterion of the number of

factors to rotate, was employed here to test the hypothesis which

we have subsequently rejected that as few as two or three variables

are required to explain the sixteen source traits. But the results

of the factor analyses can also serve as a convenient descriptive

and summary device under which to consider Cattell's primaries.

The first and largest factor, provisionally entitled Anxiety (C-,

0+, Q4+) shows group differences where females and the lower social

classes tend to score higher but where there is no clear relationship

with age. On Assertiveness (E+, L+, H+, N-), males achieve higher

scores as do younger subjects. Social class, on the other hand, is

not related here. The third factor of Group Adherence (F+, Q2-)

shows rather conflicting sex differences, no relationship with social

class but strong negative correlations with age. I

322

TABLE: 15.1

CONSISTENCY OF SEX, SOCIAL CLASS AND AGE DIFFERENCES ON CATTELL'S

SOURCE TRAITS FOR THE SEVEN FACTOR SOLUTION

FACTOR Sex Relationship Relationship Differences with Social with Aqe (Higher Class Scoring Sex)

I Tt:?:; ýi . ety't

C- F Negative Zero

0+ F Negative Zero

Q4+ F Negative Slight Negative

II "Assertiveness"

E+ M Zero Negative

L+ M? Negative

H+ M Zero Zero

N- M Zero Negative

III "Group Adherence"

F+ NS Zero Negative

Q2- F Zero Negative r

IV "Tendermindedness"

1+ F Positive Positive

M+ M Positive Zero

V "Superego Strength"

G+ NS Zero Positive

Q3+ M Positive

Qi- F Zero Positive }

323

TABLE 15.1 (Continued)

FACTOR Sex Relationship Differences with Social (Higher Class Scoring Sex)

VI "Intelligence"

B+ M Negative

VII "Outgoing"

A+ F Zero

NOTES:

Relationship with Aqe

Negative

Zero

(1) In order to maintain consistency with regard to the direction

of relationships in this table, group differences are reversed

where a primary trait has a negative loading on the factor

concerned.

(2) KEY:

F. Females scoring significantly higher on the negative

or positive pole of the primary trait.

M: Males scoring significantly higher.

NS: No significant differences.

?: Relationship unclear due to conflicting results.

I

.f

324

In terms of the fourth factor, Superego Strength (G+, 03+, Ql-),

the picture of sex differences is rather confused but the relationships

with age are consistently strong positive with no apparent differences

between the social classes. Tender-mindedness (I+, M+) again has

conflicting sex and age differences, perhaps because Cattell's M

tends also to be related to B (intelligence), but shows positive

rý-=at_ion.; h os v th social- class.

Finally, Intelligence (B+) has demonstrated a negative association

with age and social class, whilst Outgoingness (A+), unlike the

Assertiveness and Group Adherence factors, has no clear relationship

with either variable.

Thus these researches provide further support for the view that

Cattell's primaries are not functionally independent. When grouped

under the relatively co&irse headings used here, the source traits

acting as major contributors to these factors behave with considerable

consistency. And this is despite the fact that a number of the

primaries load on more than one of the oblique factors derived

from these analyses.

We have also seen that Cattell's age corrections had little appreciable

effect on some eight of his source traits (B, E, G, H, I, L, Q1 and Q3).

On six of the remaining factors (A, C, M, O, Q2 and Q4), they introduced

error which was far more extreme than the original scale-age correlations.

There was no clear benefit to be gained, moreover, in using non-linear

explanations of age differences in personality. Indeed, in some instances.,

Cattell'S quadratic expressions tended to compound rather than minimise

error.

325

CONCLUSIONS REGARDING THE RELIABILITY OF THE PRIMARY SOURCE TRAITS

The reliability of the primary source traits estimated by the

correlation between the scale scores on the corresponding factors

of Forms A and B of the 16PF were generally in line with those

provided by Cattell, although as a result of reduced variability

in scores, somewhat lower than those calculated on student groups.

The most reliable primaries were the factors. of E (assertive),

F (happy-go-lucky), H (socially bold), I (tenderminded),

O (apprehensive) and Q4 (tense), whilst four factors, L (suspicious),

M (imaginative), N (shrewd) and Q1 (experimenting), were of such low

reliability (coefficients ranged from . 31 to . 25) as to be of little

practical utility.

The variables of age and particularly social class were found to be

significant moderators on the reliability coefficients of the

questionnaire scale scores. These effects were largely eradicated,

however, when the SEM statistic was used to take account of differences

in the variability of scores within the various sub-groups.

Nevertheless, some tendency still remained for lower social class

respondents to show'lower reliability of scores.

The traits which cause most concern in terms of reliability appear

to be the primaries of L, M and N. which make up three of the last

four factors previously extracted from L data, and Q1, the first of

the four primaries only found in the questionnaire realm. We must

accept, therefore, our general hypothesis; reliabilities as low as .2

326

certainly suggest that these factors are untenable in terms of the

precision of measurement afforded by the model. Moreover, use of

the scales which supposedly measure these traits in the lower social

classes will mean that the reliability of measurement is lower still.

CONCLUSIONS ON THE INTERNAL CONSISTENCY OF THE MODEL

The internal consistency of the primary source traits was investigated

by way of composite item-scale correlations, being the multiple

correlation of sums of the items with the 16PF Forms A+B scale

scores corrected for spurious item scale-score overlap. This is

analogous in scale score terms with Cattell's concept of 'direct

validities' which he has described as the multiple correlation of

the items in a scale with the 'pure' factor abstracted from factor

analyses.

It was also shown that the multiple correlation of sums of the items

in one scale with another scale is directly equivalent to the

intercorrelation of the two scales. It was possible, therefore,

to compare composite item-scale coefficients with scale intercorrelations

not only as a measure of the estimated'direct validities' of. items on

their own scale but also the 'direct validities' for items on external

scales.

The primaries of high estimated scale direct validity were C, E, F, H, O,

Q3 and Q4. Factors L, M, N and 02, on the other hand, showed relatively

327

low composite item-scale correlations; precisely those traits which

showed low alternate form reliability values. In general, the

composite item correlations tended to be slightly below alpha values

and the alternate form reliabilities corrected for two form length

by the Spearman-Brown prophecy formula.

vTnen the scale direct -, validity estimates wer` corparad with the

'validities' of items with external scales (ie. scale intercorrelations),

it was found that the anxiety factors of C, 0 and Q4 tended to measure

themselves about as well as they measured one another. Similarly,,

Factor G (superego strength) items had direct validity estimates on

Factor Q3 (social precision) which were virtually as high as their

direct validities on their own scale. Other source traits which

appear to be made up of non-discreet item content were the low

reliability primaries of L (suspicious), M (imaginative), N (shrewd)

and Ql (experimenting).

We have noted that Cattell has advocated the use of suppressor action

to improve the multiple correlation of items with some estimate of the

pure personality factor. The presence of suppressor action within the

adult personality questionnaires was tested here by comparing the

composite item-scale correlations with conventional homogeneity

estimates provided by Cronbach's coefficient alpha. If the action

of suppressor items was marked within the scales, then according to

the general law of multiple correlation the composite item-scale

correlations would tend to be higher than alpha statistics calculated

on the same data.

328

Unfortunately, no evidence was found in this study for any pronounced

suppressor effects within the primary source traits as represented by

the 16PF scale scores. Indeed, alpha coefficients tended to be slightly

but consistently higher than the composite item-scale correlations.

We must conclude, therefore, that despite Cattell's claims,

conventional alternate form or homogeneity estimates do not

z derestirnate the reliability of his scales.

CONCLUDING REMARKS

There would seem a good deal of evidence here to suggest that Catteil. 's.

primary source traits cannot be described as the simplest, most

reliable definition of the personality domain. It is particularly

alarming that so many glaring inadequacies exist in such widely used

personality questionnaires. Most obvious of these are inaccurate

age corrections, the low reliability and the complete interdependence

of certain source traits.

What is the most likely explanation of these findings? Perhaps it

was Cattell's early work on ratings based on small student groups and

using relatively unsophisticated correlational techniques which was

in error. Considerable criticism has been made of Cattell's method

of rotating to a defined target. As factor analytic and computer

techniques improved, Cattell, perhaps misguidedly, has spent his

resources attempting to replicate the results of early researches,

rather than refine his initial strictures on the structure of

personality. In other words, Cattell's heavy investment in his

theory has blinded him to criticisms made of the primary source traits.

329

Indeed, if this theory has undergone any change it has been by the

addition of new traits rather than by the critical evaluation of

those already existing in the model.

This is not to detract from Cattell's pioneering work on the factor

analytic investigation of personality. His writings on the methodology

of factor analysis have mate ö considerable contrihu is; n to the field..

But, in the case of Cattell's primary source traits, our conclusion

must be that theory seems to have outstripped experimental verification.

AREAS OF FURTHER RESEARCH

There would appear to be little point in carrying out further cosmetic

research on Cattell's model of personality. Howarth and Browne (1971),

Eysenck (1972) and Saville and Blinkhorn (1976) have all shown that

clear inconsistencies exist in the model as represented in questionnaire

data.

Whereas anxiety as represented by factors C, 0 and Q4 in the 16PF

emerge as a unitary trait, the nature of introversion/extraversion as

a variable in the personality domain is far less clear-cut. The

evidence here points to possibly three sub-components of this dimension

and further research should concern itself with the problem of defining

these components by factor analytic investigations of the 16PF items,

which supposedly measure the A, E, F, G, H, L, N and Q2 '. source' traits.

In addition, the nature of the factors which have been provisionally

330

named Tendermindedness (I+, M+) and Superego Strength (G+, Q3+ and

Ql-) need further study in order to establish their psychological

significance and to ascertain their exact relationship with

variables in other personality theories.

331

16. REFERENCES

332

RF. FRRF. NC F:.,;

ABRAHAMS NM and ALF E (1972) Pratfalls in Moderator Research Journal of Applied Psychology, Vol. 56, No. 3,245-251.

ADCOCK CJ (1965) A Comparison of the Concepts of Cattell and Eysenck: Research Notes. Brit. J. Educ. Psychol., 35,90-97.

ADCOCK N (1973)

Testing the Test: How Adequate is the 16PF with a New Zealand Student Sample? Paper presented to New Zealand Psychological Society Conference.

ADCOCK N V, ADCOCK CJ and WALKEY EH (1974) Basic Dimensions of Personality Victoria University of Wellington, New Zealand.

ALEAMONI LM and SPENCER RE (1969)

A Comparison of Biserial Discrimination, Point-Biserial Discrimination and Difficulty Indices in Item Analysis Data. Educ. & Psychol. Meas., 29,353-358.

ALF E and ABRAHAMS N (1971) A Significance Test for Biserial r' Educ. & Psychol. Meas., vol. 31, Autumn.

ANASTASI A (1961) Psychological Testing MacMillan Co., New York, 2nd Edition.

ARNHOFF FN and LEON HV (1963)

Personality Factors related to Success and Failure in Sensory Deprivation Subjects. Perceptual and Motor Skills, 16,46.

BAKKER CB and LEVENSON RM (1967) Determinants of Angina Pectoris Psychosomatic Medicine, 29,621-633.

BALTES PB (1968) Longitudinal and Cross-sectional Sequences in the Study of Age and Generation Effects. Human Development, 11,145-171.

BAREFOOT JC (1969) Anxiety and Volunteering Psychonomic Science, 16(6), 283-284.

333

BARTON K and CATI'ELL RB (1972)

Personality hefore and after a Chronic Illness J. of Clinical Psychology, Vol. XXVIII, No. 4,464-467, October.

BARTON K, CATTELL RB and VAUGHAN GM (1973)

Changes in Personality as a Function of College Attendance or Work Experience Journal of Counseling Psychology, Vol. 20,2,162-5.

BASS AR and ROSEN H (1969)

Some Potential Moderator Variables in Attitude Research Educ. & Psychol. Meas., 29,331-348.

BEALE EML, KENDALL MG and MANN DW (1967) The Discarding of Variables in Multivariate Analysis Biometrika, 54,3 and 4, p. 357.

BECKER WC (1960) The Matching of Behavior Rating and Questionnaire Personality Factors. Psychol. Bulletin, Vol. 57, No. 3.

BECKER wC (1961) A Comparison of the Factor Structure and Other Properties of the 16PF and the Guilford-Martin Personality Inventories Educ. & Psychol. Meas., Vol. XXI, No. 2.

BOORER D and MURGATROYD S (1973) . Personality and Learning -A Select Annotated Bibliography MTM Publishing House (Wales), Caerphilly.

BOWERS J (1972) A Note on Comparing r-Biserial and r-Point Biserial Educ. & Psychol. Meas., 32,771-775. "

BROWNE MW (1972) Oblique Rotation to a Partially Specified Target Br. J. Math. Stat. Psychol., 25,207-212.

BROWNE MW (1972) Orthogonal Rotation to a Partially Specified Target Br. J. Math. Stat. Psychol., 25,115-120.

BROWNE MW and KRISTOF W (1969) On the Oblique Rotation of a Factor Matrix to a Specified Pattern Psychometrika, 34, No. 2, June.

334

BÜROS 0K (Editor) (1972) The Seventh Mental Measurements Yearbook (Vols 1 and 11) Institute of Mental Measurements Gryphon Press, Highland Park, New Jersey, USA.

BUZZARD RB (1971)

Obituary to Sir Frederick Bartlett Occupational Psychology, 45, Vol. 1,1-6.

CAMPBELL DT and FISKE DW (1959) Convergent and Discriminant Validation by the Multitrait-Multimethod Matrix Psychological Bulletin, 56, No. 2, March.

CARROLL JB (1961)

The Nature of the Data or How to Choose a Correlation Coefficient Psychometrika, Vol. 26, No. 4, December.

CATTELL RB (1933) Temperament Tests I. Temperament Brit. Journ. Psychol., 23.

CATTELL RB (1934)

Temperament Tests II. On the Nature of Preservation and the Tests devised to Measure it. Brit. Journ. Psychol., 24.

CATTELL RB (1943) The Description of Personality Basic Traits resolved into Clusters. J. Abnormal & Soc. Psychol., 38,476-506.

CATTELL RB (1943) The Description of Personality I: Foundations of Trait Measurement Psychol. Review, 50,6,559-592.

CATTELL RB (1944)

Interpretation of the Twelve Primary Personality Factors Character and Personality, 13,55-91.

CATTELL RB (1944)

Psychological Measurement: Normative, Ipsative, Interactive Psychol. Review, 51,292-303.

CATTELL RB (1945) The Description of Personality. III: Principles and Findings in a Factor Analysis Amer. J. Psychol., 58,69-90.

I

335

CATTELL RB (1946) The Description and Measurement of Personality World Book Co. (now HBJ), New York.

CATTELL RB (1946) A Shortened 'Basic English' Version (Form C) of the 16PF Questionnaire Journal of Social Psychology, Vol. XLIV, 257-258.

CATTELL RB (1946)

Simple Structure in relation to Alternative Factorization of Personality ST)he re

Journal of General Psychology, Vol. 35,225-238.

CATTELL RB (1947) Oblique, Second Order and Cooperative Factors in Personality Analysis Journal of General Psychology, 36,3-22.

CATTELL RB (1950) The Main Personality Factors in Questionnaire Self-Estimate Material Journal of Social Psychology, 31,3-38.

CATTELL RB (1956) Validation and Intensification of the 16PF Journal of Clinical Psychology, 12,205-214.

CATTELL RB (1956) Second-Order Personality Factors in the Questionnaire Realm Journal of Consulting Psychology, 20, No. 6.

CATTELL RB (1957)

Formulae and Table for obtaining Validities and Reliabilities of Extended Factor Scales Educ. & Psychol. Meas., Vol. IV, 491-498.

CATI*ELL RB (1957) Personality and Motivation Structure and Measurement World Book Co. (now HBJ), New York.

CATTELL RB (1959) Anxiety, Extraversion, and other Second Order Personality Factors in Children Journal of Personality, 27,464-476.

CATTELL RB (1959) The Dynamic Calculus: Concepts and Crucial 'Experiments Nebraska Symposium on Motivation (University of Nebraska Press)

336

CATTELL RB (1960) The Multiple Abstract Variance Analysis Equations and Solutions: For Nature-Nurture Research on Continuous Variables Psychological Review, 67, No. 6. November.

CATTELL RB (1962) The Basis of Recognition and Interpretation of Factors Educ. & Psychol. Meas., Vol. XXIi, No. 4.

CATTELL RB (1963) Theory of Fluid and Crystalized Intelligence: A Critical Experiment Journ. Educ. Psychol., Vol. 54, No. 1,1-22.

CATTELL RB (1963) Methodological Advances in Evaluating Hereditary and Env ironmental Influences and their Interaction Laboratory of Personality Assessment, University of Illinois Advance Publication No. 19, March.

CATTELL RB (1963) Teachers' Personality Description of Six Year Olds: A Check on Structure

Brit. Journ. Educ. Psychol., 33, Part 3,2=9-235.

CATTELL RB (1963) The Interaction of Hereditary and Environmental Influences Brit. J. Stat. Psychol., Vol. XVI, Part 2, November.

CATTELL RB (1964) Objective Personality Tests: A Reply to Dr. Eysenck Occupational Psychology, 38, No. 2, April.

CATTELL RB (1965) The Scientific Analysis of Personality Pelican Books, London.

CATTELL RB (1965) The Configurative Method for surer Identification of Personality Dimensions, notably in Child Study Psychol. Reports, 16,269-270.

CATTELL RB (1966) Handbook of Multivariate Experimental Psychology Rand McNally & Co., Chicago.

CATTELL RB (1966) The Scree Test for the Number of Factors Multivariate Behavioral Research, Vol. 1,245-276, April.

337

CATTELL RB (1966) Further Investigation of Objective Personality Test Factors in Early Childhood ETS Research Bulletin RB-66-42. August.

CATTELL RB (1968) Trait-View Theory of Pertubations in Ratings and Self Ratings (L(BR)-

and Q-Data): Its Application to obtaining Pure Trait Score Estimates in Questionnaires Psychological Review, 75, No. 2,96-113.

CATTELL RB (19019)

The Profile Similarity Coefficient, rp, in Vocational Guidance and Diagnostic Classification Brit. J. Educ. Psychol., 39, Part 2, June.

CATTELL RB (1970) Real Base Factor Analysis Laboratory of Personality Analysis, University of Illinois Advance Publication No. 15. Tentative November.

CATTELL RB (1970) Separating Endogenous, Exogenous, Ecogenic and Epogenic Component Curves in Development Data Dev. Psychol., Vol. 3, No. 2. APA Inc.

CATTELL RB (1972) Tabular Supplement (of Norms) No. 2 to the 16PF Handbook (for 1969 editions of Forms C and D) IPAT, Champaign, I1 linois, USA.

CATTELL RB (1972) The 16PF and Basic Personality Structure J. Behav. Science, 1, (4), 169-187.

CATTELL RB (1973) 1

Personality and Mood by Questionnaire Jossey-Bass, San Francisco.

CATTELL RB (1973) The Organisation of Independent Basic Research Institutes Symbiotic with Universities Higher Education, 2,1-14.

CATTELL RB (1974) How Good is the Modern Questionnaire? General Principles for Evaluation J. Personality Assessment, 38, No. 2.

338

CATTELL RB (1974)

A Large Sample Cross-Check on 16PF The Journal, Fall, 179-195.

CATTELL RB (1975)

Third Order Personality Structure in Q-Data - Evidence from 11 Experiments The Journal, Winter, 118-149.

CATTELL R B, BALCAR K, HORN JL and NESSELROADE JR (1969) Factor Matching Procedures: An Improvement of the s Index; with Tables

Educ. & Psychol. Meas., 29.

CATTELL R B, BARTON K and DIELMAN TE (1972)

Prediction of School Achievement from Motivation, Personality and Ability Measures Psychological Reports, 30,35-43.

CATTELL RB and BELOFF H (1953)

Research Origin and Construction of the IPAT Junior Personality Quiz

Journal of Consulting Psychology, 17, No. 6.

CATTELL R B, BLEWETT D and BELOFF JR (1955)

The Inheritance of Personality. A Multiple Variance Analysis Determination

of approximate Nature Nurture Ratios for Primary Personality Factors in

Q-data. American J. Human Genetics, Vol. 7, No. 2, June.

CATTELL RB and BUTCHER J (1968) The Prediction of Achievement and Creativity Bobbs-Merrill, Indianapolis.

CATTELL RB and COAN RW (1957) Child Personality Structures as revealed in Teachers' Behavior Ratings. Journal of Clinical Psychology, 13,315-327.

CATTELL RB and COAN RW (1957) Personality Factors in Middle Childhood as revealed in Parents' Ratings Child Development, 28, No. 4, December.

CATTELL RB and COAN RW (1959) Objective Test Assessment of the Primary Personality Dinensions. in Middle Childhood Brit. J. Psychol., 50,235-252.

CATTELL R B, COAN RW and BELOFF H (1958) A Re-examination of Personality Structure in Late Childhood, and Development of the High School Personality Questionnaire J. Exper. Education, 27, December.

339

CATTELL RB and DAMARIN F (1966)

Further Investigation of Objective Personality Tests Factors in Early Childhood ETS Bulletin RB-66-42. August.

CATTELL RB and EBER HW (1967-1970)

Sixteen Personality Factor Questionnaire (16PF) Forms A&B (1967-68 Edition) and C&D (1969-70 Edition) IPAT, Champaign.

CATTELL R B, EBER HW and DELJHEES K (1968)

A Large Sample Cross Validation of the Personality Trait Structure of the 16PF with some Clinical Implications Multivariate Behavioral Research, Special Issue, 107-132.

CATTELL R B, EBER HW and TATSUOKA M (1970) Handbook for the Sixteen Personality Factor Questionnaire (16PF) Institute for Personality and Ability Testing, Champaign, Illinois, USA.

CATTELL RB and GIBBONS BD (1968) Personality Factor Structure of the Combined'Guilford and Cattell Personality Questionnaires J. Personality and Social Psychology, Vol. 9, No. 1,107-120.

CATTELL RB and GRUEN W (1954) Primary Personality in the Questionnaire Medium for Children Eleven to Fourteen Years Old Educ. & Psychol. Meas., 14,50-76.

CATTELL RB and GRUEN W (1955) The Primary Personality Factors in 11-year-old Children by Objective Tests Journal of Personality, 23,460-478.

CATTELL RB and HOWARTH E (1964) Verification of Objective Test Personality Factor Patterns in Middle Childhood J. Genetic Psychology, 104,331-349.

CATTELL R B, KAWASH G and DEYOUNG GE (1972) Validation of Objective Measures of Ergic Tension: Response of the Sex Erg to Visual Stimulation Journal of Experimental Research in Personality, Vol. 6, No. 1, March.

CATTELL R 8, KONII OS E and TATRO D (1968)

Significant Differences of Affective, Paranoid and Non-Paranoid Schizophrenic Psychotics on Primary Source Traits in the 16PF.

Multivariate Behavioral Research, Special Issue, pp. 33-54.

340

CATTELL RB and MERLE JL (1960) The 'Maxplane' Program for Factor Rotation to Oblique Simple Structure Educ. & Psychol. Meas., Vol. XX, No. 3.

CATTELL RB and NICHOLS KE (1972) An Improved Definition from Ten Researchers, of Second Order Personality Factors in Q-data (with cross-cultural checks). Journal of Social Psychology, 86,187-203.

CATTELL RB and RADCLIFFE JA (1962) Reliabilities and Validities of Simple and Extended Weighted and Buffered Unifactor Scales. Brit. Journ. Stat. Psychol., Vol. XV, Part 2, November.

CATTELL RB and SCHIER IH (1961)

The Meaning and Measurement of Neuroticism and Anxiety Ronald Press, New York.

CATI'ELL R B, SEALY AP and SWENEY AB (1965)

What can Personality and Motivation Source Traits add to the Prediction

of School Achievement?

CATTELL R B, STICE EF and KRISTY NF (1957)

A First Approximation to Nature-Nurture Ratios for Eleven Primary

Personality Factors in Objective Tests

J. Ab. & Soc. Psychol., 54, No. 2, March.

CATTELL RB and TSi7TIOKA B (1964) The Importance of Factor-Trueness and Validity versus Homogeneity and Orthogonality in Test Scales Educat. & Psychol. Meas., Vol. XXIV, No. l.

CATTELL R B, WAGNER A and CATrELL MD (1970) Adolescent Personality Structures in Q-data, checked in the High School Personality Questionnaire. Brit. J. Psychol., 61,1,39-54.

CATIELL RB and WARBURTON FW (1961) A Cross-cultural Comparison of Patterns of Extraversion and Anxiety Brit. J. Psychol., 52,1,3-15.

CHANDLER RE (1961) Two Additional Formulae for use with Supressor Variables Educ. & Psychol. Meas., Vol. XXI, No. 4.

CHILD D (1966) Personality and Social Status Brit. Journ. Soc. & Clin. Psychol., 5,196-199.

341

CHILD D (1969)

A Comparative Study of Personality, Intelligence and Social Class in a Technological University Brit. Journ. Educ. Psychol., 39,40-46.

COAN RW and CATTELL RB (1958)

Reproducible Personality Factors in Middle Childhood Journal of Clinical Psychology, 14,339-45, October.

COHEN J (1968) Multiple Regression as a General Data-Analytic System Psych. Bulletin, 70, No. G, 426-443.

CONGER AJ (1969) An Analysis fo Ghiselli's Moderator Variable Psychological Reports, 25,519-527.

CONGER AJ and JACKSON DN (1972) Suppressor Variables, Prediction and the Interpretation of Psychological Relationships Educ. & Psychol. Meas., 32,579-599.

COOLEY WW and LOHNES PR (1971) Multivariate Data Analysis Wiley, New York.

CRONBACH LJ (1951) Coefficient Alpha and the Internal Structure of Tests

Psychometrika, 16, No. 3, September.

CRONBACH LJ (1970)

Essentials of Psychological Testing Harper & Row International, New YorkL Evanston London. Third Edition.

CRONBACH LJ and GLESER G (1953) Assessing Similarity between Profiles Psychol. Bulletin, Vol. 50, No. 6.

DARLINGTON RB (1968) Multiple Regression in Psychological Research and Practice Psychol. Bulletin, Vol. 69, No. 3,161-182.

DEYOUNG GE (1972) Standards of Decision regarding Personality Factors in Questionnaires Canadian Journal of Behavioral Psychology.

342

DICKEN CF (1959) Simulated Patterns on the Edwards Personal Preference Schedule J. Applied Psychology, 43,372-378, December.

DUNETI'E M (1972)

Comments on Abrahams and Alf's 'Pratfalls' in Moderator Research J. Applied Psychology, 56,3,252-56.

EDWARDS AL and ABBOTT RD (1973)

Relationships between the EPI Scales and the 16PF, CPI and EPPS Scales Educ. & Psychol. Meas., 33,231-38.

ENTWISTLE NJ (1972) Students and their Academic Performance in Different Types of Institution In: 'Contemporary Problems in Higher Education' (Ed. Butcher HJ and Rudd E) London, McGraw-Hill.

ENTWISTLE NJ and CUNNINGHAM S (1968) Neuroticism and School Attainment: A Linear Relationship Brit. Journ. Educ. Psychol., 38,123-132.

EYSEN KHJ (1947) Dimensions of Personality London, Routledge & Kegan Paul.

EYSEN KHJ (1954)

The Psychology of Politics London, Routledge & Kegan Paul.

EYSENCK HJ (1960) The Structure of Human Personality 2nd Edition- M ethuen & Co., London.

EYSENCK HJ (1964) Eysenck on Vernon's 'Personality Assessment: A Critical Survey' (London, Methuen, 1964) Occupational Psychology, 38,2,115-117.

EYSENCK HJ (1972) Primaries or Second-Order Factors: A Critical Consideration of Cattell's 16PF Battery

Brit. Journ. Soc. & Clin. Psychol., II, 265-269.

EYSENCK HJ and EYSENCK SBG (1964) Eysenck Personality Inventory University of London Press, London.

343

EYSENCK HJ and EYSENCK SBG (1969) Petsonality Structure and Measurement London, Routledge & Kegan Paul.

EYSENCK HJ and EYSENCK SBG (1975) Eysenck Personality Questionnaire Hodder & Stoughton, London.

EYSENCK SBG and EYSENCK HJ (1972)

The Questionnaire Measurement of Psychoticism Psychol. Medicine, 2, No. 1,50-55, February.

FLEISHM. z1N JJ and FINE BJ (1971) Note on Cognitive Factors related to Factor B of the 16PF Psychological Reports, 29,1075-1077.

FORBES A R, DEXTER WR and COMFREY A (1974) A Cross-Cultural Comparison of certain Personality Factors Multivariate Behavioral Research, Vol. 9, October.

GHISELLI E (1972) Comment on the Use of Moderator Variables Journ. Applied Psychology, 56,3,270.

GHISELLI EE and SANDERS ET (1967) Moderating Heterascedasticity Educ. & Psychol. Meas., 27,581-590.

GOLDBERG LR (1970) Why Measure that Trait? An Historical Analysis of Personality Scales and Inventories Tutorial Address to Western Psychological Association, Los Angeles, April.

GOODWIN W, JOYCE J, ABRAMS A, BIAGGIO A and BIAGGIO L (1965) Note: Relationships between Matching Scales of the Cattell 16PF and the HSPQ. Symposium at National Council on Measurement in Education, Chicago, Illinois. February.

GORMLY G and EDELBERG W (1974) Validity in Personality Trait Attribution American Psychologist, March, 189-193.

GORSUCH RL (1970) A Comparison of Biquartimin, Maxplane, Promax and Varimax. Educ. & Psychol. Meas., 30,861-872.

GUERTIN WH (1965) Updating the Internal Criterion of Item Analysis Psychological Reports, 16,376-378. Sourhern Universities Press.

t

344

GUILFORD JP (1954) Psychometric Methods McGraw-Hill Book Co., New York. 2nd Edition

GUILFORD JP (1959) Personality McGraw-Hill Book Co., New York.

GUILFORD JP and ZIMMERMAN WS (1949-55) The Guilford-Zimmerman Temperament Survey Sheridan Psychological Services Inc., Orange, California.

GULLIKSEN H (1968) Methods for Determining Equivalence of Measures Psychological Bulletin, 70, No. 6,534-544.

HARTMAN BJ (1966) Personality Factors of the Cattell 16PF Test and Hypnotic Susceptibility Psychological Reports, 19,1337-1338.

HENRYSSON S (1962) The Relation between Factor Loadings and Biserial Correlations in Item Analysis Psychometrika, 27, No. 4, December.

HENRYSSON S (1963) Correction of Item-Total Correlations in Item Analysis Psychometrika, 28, No. 2, June.

HOWARD KI and FOREHAND G (1962) A Method for Correcting Item-Total Correlations for the Effect of Relevant Item Inclusion Educ. & Psychol. Meas., 22, No. 4.

HOWARTH E and BROWNE JA (1971) An Item Factor Analysis of the 16PF Personality, Vol. 2, No. 2.

HOWARTH E and BROWNE J (1972) An Item Factbr Analysis of the Eysenck Personality Inventory Brit. Journ. Soc. Clin. Psychol., 11,162-174.

HOWARTH E, BROWNE JA and MARCEAU R (1972) An Item Analysis of Cattell's 16PF Canadian Journ. Behav. Sci. /Rev. Canadian Sci_ Comp., 4(l).

345

JAHODA G (1970) A Cross-Cultural Perspective in Psychology The Advancement of Science, 27,1970-71.

JOHNSGARD KW and OGILVIE BC (1968) The Competitive Racing Driver: A Preliminary Report. Journ. Sports Medicine and Physical Fitness, 8,87-95.

KELVIN P R, LUCAS CV and OJHA AB (1965) The Relationship between Personality, Mental Health and Academic PýrforTance in University Students Brit. Journ. Soc. & Clin. Psychol., 4,244.

KLINE P (1966) Extraversion, Neuroticism and Performance among Ghanaian University Students Brit. Journ. Educ. Psychol.., 36,92-94.

LEVONIAN E (1961) Personality Measurement with Items selected from the. 16PF Educ. & Psychol. Meas., 21, No. 4.

LEVONIAN E (1961) Statistical Analysis of the 16PF Educ. & Psychol. Meas., 21, No. 3.

LEVY P (1973)

On the Relation between Test Theory and Psychology In 'New Approaches in Psychological Measurement' (Wiley, London) Ed. Kline P.

LORD FM and NOVICK MR (1968) Statistical Theories of Mental Test Scores Addison-Wesley Publishing Co.

LYNN R (1959) Two Personality Characteristics related to Academic Achievement Brit. Journ. Educ. Psychol., 29,213-216.

McKENNELL AC (1967) The Choice of Correlation Coefficients for Attitude Scale Construction. Research Bureau Ltd.

MANDRYK TR and SCHUERGER J (1974) Cross-Validation of the HSPQ as a Predictor for High School Grades Educ. & Psychol. Meas., 34,449-454.

346

MEEHL P (1945)

The Dynamics of 'Structureä'Personality Tests Journ. Clin. Psychol., 1,296-303.

MOLLENKOPF WG (1949)

Variation of the Standard Error of Measurement Psychometrika, 14, No. 3, September.

MORRIS J and MARTIN J (1972) Computer Personnel Selection - 2: Programmers National Computing Centre, Manchester.

MURRAY HA (1938) Explorations in Personality Oxford University Press.

NIE N, BENT D and HULL C (1970) Statistical Package for the Social Sciences McGraw-Hill, New York.

NUNNALLY JC (1967) Psychometric Theory McGraw-Hill, New York.

OOSTERHOF AC (1976) Similarity of Various Item Discrimination Indices Journ. Educational Measurement, 13, No. 2.

OVERALL JG and KLETT CJ (1972) Applied Multivariate Analysis McGraw-Hill, New York.

ROSENBERG M (1970) The Logical Status of Suppressor Variables

SAVAGE RD (1972) Personality Factors and Academic Performance In 'Personality and Educational Achievement' (Ed. Naylor F D) John Wiley & Sons Australasia Pty. Ltd., Sydney.

SAVILLE P (1972) The British Standardisation of the 16PF: Supplement of Norms, Forms A and B NFER Publishing Co., Windsor.

347

SAVILLE P (1973)

The Standardisation of an Adult Personality Inventory on the British Population Bulletin British Psychological Society, 26,25-29.

SAVILLE P and BLINKHORN S (1976) British Undergraduate Norms to the 16PF (Forms A and B) NFER Publishing Co., Windsor

SAVILLE P and BLINKHORN S (1976) British Undergraduate Norms to the 16PF (Forms C and D) NFER Publishing Co., Windsor.

SAVILLE P and BLINKHORN S (1976)

British Undergraduate Norms to the IPAT Anxiety Scale, Neuroticism Scale Questionnaire and the Eysenck Personality Inventory (Form A) NFER Publishing Co., Windsor.

SAVILLE P and FINLAYSON L (1973)

British Supplement to the High School Personality Questionnaire (Form A) Anglicised 1968/69 Edition. NFER Publishing Co., Windsor.

SCHAIE KW (1962) On the Equivalence of Questionnaire and Rating Data Psychological Reports, 10,521-522.

SCHEIER IH and CATTELL RB (1961) Neuroticism Scale Questionnaire IPAT, Champaign, Illinois.

SEALY AP and CATTELL RB (1966) Adolescent Personality Trends in Primary Factors measured on the 16PF and HSBQ Questionnaires through Ages 11 to 23 Brit. Journ. Soc. & Clin. Psychol., 5,172-184.

SIXTEEN PERSONALITY FACTOR QUESTIONNAIRE (16PF) Review: Buros 5th Mental Measurements Yearbook, Entry 112.

SIXTEEN PERSONALITY FACTOR QUESTIONNAIRE (16PF)

Review: Buros 6th Mental Measurements Yearbook, Entry 174.

SPIELBERGER C D, OORSUCH RL and LUSHENE R (1970) State-Trait Anxiety Inventory Consulting Psychologists Press, Palo Alto, California.

348

SUI VW (1953) The Cattell 16PF as a Prognosticator of Accident Susceptibility Proceedings of the Iowa Academy of Science, 60,558-561.

TATSUOKA MM (1972) Selected Topics in Advanced Statistics IPAT, Champaign, Illinois.

TATSUOKA MM and CATTELL RB (1970) Linear Equations for Estimating a Person's Occupational Adjustment, based on Information on Occupational Profiles Brit. Journ. Educ. Psychol., 40, Part 3, November.

TIMM V (1968) Reliabität und Faktonenstniktur von Cattell's 16PF - Test bei einer Deutschen Stichprobe Zeiteschrift fur experimentelle und angewandte psychologie, 15,354-373.

UNIVERSITY GRANTS COMMITTEE (UGC) Statistics in Education: Universities, Vol. 6.

H. M. S. O.

VAUGHAN D (1973) The Relative Methodological Soundness of Several Major Personality Factor Analyses Journ. Of Behav. Sci., No. 5.

VELICER WF (1972)

Comment on the General Inapplicability of Ghiselli's Moderator System for Two Predictors Journ. Applied Psychol., 56, No. 3,262-265.

VELICER WF (1972) The Moderator Variable viewed as Heterogeneous Regression Journ. Applied Psychol., 56, No. 3,266-269.

VERNON PE (1964) Personality Assessment Methuen & Co., London.

VON SIEGFRIED GREIF (1970) Untersuchungen zur deutschen Übersetzung des 16PF - Fragebagens Investigation of the German Translation of the 16PF Questionnaire. Psychologische Beitrage, 12,186-213 (Berlin)

349

WARBURTON F W, BUTCHER HJ and FORREST GM (1963)

Predicting Student Performance in a University Department of Education Brit. Journ. Educ. Psychol., 33,68-79.

WILSON G (1975) Wilson-Patterson Attitude Inventory, Manual. NFER Publishing Co., Windsor.

WOLF R (1967) Evaluation of Several Formulae for Correction of Item-Total Correlations in Item Analysis Journ. Educational Measurement, 4, No. 1, Spring.

ZWEIG F (1964) The Student in the Age of Anxiety Free Press (MacMillan Co. ), New York.

350

SDCTION 7:

APPENDICES

APPENDIX CONTENTS

APPENDIX 6: N-STEH NORM TABLES

Index of Norm Tables Tables 1- 81

APPENDIX 7: DESCRIPTIVE STATISTICS: PRIMARY SOURCE TRAITS BY SCCIAL CLASS

Tables 82 - 90

APPENDIX 8: DESCRIPTIVE STATISTICS: PRIMARY SOURCE TRAITS BY AGE

Tables 91- 103

APPENDIX 10: ALTERNATE FORM RELIABILITIES OF THE PRIMARY SOURCE TRAITS BY SEX AGE AND SOCIAL CLASS

Tables 104 - 119

APPENDIX 12: ITEM ANALYSIS DATA ON THE HSPQ

Tables 120 - 134

Pa es

3

4 5-85

86

87-95

96

97-109

110

111-126

127

128-142

I

2

APPENDIX 6'

N- STENN0RMT ABLE S

Please note that Appendix numbering cross references with the main text.

3

z 010 1 HÖ

a a

LO Cý4 CO N M tO l- CA

N Co [- CO La le C) N V-1 0 U) U V-1 CV Cl d' U) to l- Co U)

w N M Cf) N U d ' t to

p n

U)

ä (D Ln N rli (IN c-4 M d' Ln 4.0 tto N

o t ßf1 d' M CV c4. C) CF% ýp to c-I N M c' U) in lD N

dt N

to v CV M ý

to h

141

[14 M CM Ir-i CD (n Q

Uj vi CV c') M LI) to N

(ý M 7 C

Ch

M N A N

Lj v-+ CV (NI Il) tO N

d' N

ý N le vi r -I M Ln h

H

W W Ctl ti, Q OD + Q CO t Q fA f "0 0 ý+ w a Q Q

U)

)

r4

C-14

Ln 10 N It

N

j m4

I

11 CA

O Ol

* N

lD ON

ý-1

-4 Q1

M

H N

.4

00 lfl 4

C7 d1

M

M d'

in

-I N

M

N lo

M

m fit'

M

o 01

N

00 -'

mot'

ka _4

m

00 W

M

to M

M

in . -1

tf)

Lý N C'

CA

.0 M

I-

lo ri

In

01 d'

m

0 O

l0

N I

O

d' l

d'

r4 . --1

m

N 7

Ol

-tv r4

N

m 71

a%

co Ln

O

in N

r-4

N Ul

O

-4 r1

-4

00 c

N

U

Q'i W U W W U` x. H a ý" Z. Q 1-1 a

N a

M Oý

cN Ot W

O . -I

O fV

1 I'D r-1

m r-4 1 N H

to N I

M N

ko N

1 (V (V

W N

1 in N

Co tV I

61

lD N

I LA (V

O tV

I CO . -a

O N 1

N . -1

110 N I

O N

O N I

lß -i

l0 N 1

ý-4 N

O N 1

-i

O N 1

CO r-1

O N I

CO . -I

lfl N I

N N

O H

oý

LA . -1 I

dr f-I

-4 . -1

(V (V I

1-1 - N

. -4 (4

ý: p N

co 11 I

N -1

-W N

I M N

CN r. 4

I

1-I

l0 -4 I

in --1

Ol r-{ I

CO r-4

in e-i

O N I

co -

co r-. 4 I

N . -I

N 1-1 I

l0 ri

N i-1

r-1 N 1

O N

a1

co O . -4

O N

O N

OÖ rl

ß'V1 N

r1 i N

l0 r-i

to 1-1

cV N

r4 N

LA . -4 M

. -4

. -4 l-. -4

ýo r-4

d' . -4 M

.4

N r-I

tÖ r-i

ko ri*

in -4

l0 --i 1

to º1

al , -{

CO i-4

00

N

rl ri

Ch r-1

CO I

r-f

iN r-1 I

l0 r-4

O (N

I O1 r -l

et' rl 1

N r-4

O N 1

CO r-I

N r-1

rn . -! I

N r-1

in r-1 I

. -1

N '-1 1

-I rl

in '-4 I

M r-4

in 1

i qll 1-4

T . -I

II

e--1

er ri I

m . -d

N _4 I

l0 -1

h

O r-I co

N ri I

in ri V'

CO . --1

r- I

r-1 -4

C- r-1 L1

ý-4 r"4

ä1

ý-4 r-i

Ö

M r -q

O r-4

N r-I

ri

M -

N

N r-4

r-4

N r-1

Ln -1 I

N to

I V

fM

I N

tG

1 d4

O I

Q1

c4'

I N r-I

I N C6

fV

1 . -1

O

1 01

-4 -4

O I

C1

. -4 1

O

N

I H

U)

Ul t+

'd' co 1

N

ýO tß'1 -4 1

r-i r4 I

OO

M r-4 1

CO 1

N

.-4 ri I

01

ko IN I

lfl I C1

co 1 N

00 I

N

'

1 01

QO 1

tN

ON 1

CO C1O d'

M ko LA

H i

O

ON

00

H .4

m

l0

LA l0

to

-, T

in CO if

in

tß I

co \p

Ln

;

lD

i M

N in

vl'

-w 01

C6

IN

to

CO C_

d' N

to d'

M

l")

N

W, <f' M

c

l-

lÖ

' Ln M N N

" M

(DI M

ý N

ö Ln

b tD

a H

ö CA

Al N

ö N

cl, to

Cb C

CID . -t

C, ý to

0 N

0 V

0 r-I

0 a

w a tja U w w 0 W H a z o ä Öº Cl Oý

a

w

}

w

44

wý ä2

z

o, Un Un OD (n co 0 ' Ir w c, m v' M co o A (NJ O IN VI 00 M tf) t. () u) (N a) Q\ H co A

" C7

" M

" N

" M

" d'

" a'

" M

" In

" M

" M

" M

" N

" M

" C. )

" M M"

" d'

" U)

U) '. o U) 0 M N V' VI ' d' t0 CYl VI O tt' ýi' Z c+ LO C N VI 0 C O S tO 0) M '. o t O N z 0

pl N N -I

M VI

M VA

N -I

(V -4

OA co N VI

ON O I

O i -1

N r

N i c c c C r c v

U 4 co U W ka 0 r H a Fam. " C7 Q

W G.

O M ko to tD O 'O O O %0 O '. o O O O '. D N VI N N N N N N N CV CV CV N CV (11 N 1 N 4 N A

Ol 4

cl, [" d 1

ý ! 1 t N

1 4 i N N N cI N ci 1 N

tD 4

9 p i (P

9 N s c r c c

c') (V c'1 lD kD 0) OD VI 01

-4 V

N I VI

(V U) (0 VI

I -I t[ -I 9 c C N

c N C'4 VI

U) VI VI

r c VI

c 0 N

V' T-4 O 0 N V ý' d' t0 to O'. VI N N N VI N c1 ri VI %-1 T-I

O t A CA ý

Azr L o 'ID 1

IQ 4

00 a

VI cý VI A VI c c t _ r

c

+ 1 V-ý c VI ß

VI VI

tý

VI '

c

tD 4

c

r+

ri

ci to

V N

VI A

r4 -1 c 4

VI!

,4

ci n

ri VI ti VI Ti ri I VI tI ci cI ci %-1 VI L VI

(sý to N M to O O CN r+ N (V rM [t' W rl VI cI VI VI VI ri rI c-1 c4 1-1 c-I

i t cm 4 r r T-4 r4 r4 r4 ri Tq V-1 ri r4 T ri U)

Vý Ln M CV N N O O ri cm N C1 r4 ca ri VI co VI ?I ci ri VI ca

ý 4 qqrAi It (* r! A C', l A A J

1 r4 1 rq ' ri a l C V X-4 T-4

Tq O 1,

OO OO

f

OO

I ý W 0 o0 Cp

m 1 40 N Ö' Oi o Cl) O N O

a VI

VI co I- V) U) N O) 0 N M In tf'1 { Go i co I '. 4 Ol I Ui lp I M

CO W

t- Ln r- q to r- U) N d' d' 01 1 1 j M M rt U) v sf1 N ( N

in tn ý ä ä I ä ä ä ä ä ä ä ä ä ä ý a a

w 4 0o U w w cD x H a z o cr ä ör U

w

en w

H

I

s

ri

z

O A ri O tV Cr) r c') -' vi d' N Ul N LO M ri A

" [11

" N

" N

" d'

" d'

" '

" Cr)

" Wi

" 1

" Cr)

" Cr) Cr)

" '1

" N"

" Cr) N"

" In

" VI

N c-I tN O\ to Cr) Cr) (V to N M N [- In co [- Z c+) c+ 0 O to Cr) cl co PN ri tr CD V N z QQQ

ON N Ln N V Cr) Cr) 01 coo N O O al %- M VA

R o w u w w 0 x H a x z O CY är Or ä

O c r) to tO 'O W O O '. O O '0 O O ko (11 c-4 " f-V N N N N CV (V N (V N N

P A N + 4 N ' ON Q' L-,

o c4 L-, ßl

t ' N N L c-I ri N N CV N r4

t c-I (V ri rI

k c-I

a % c-I N

0 cl c 0 N N N -I t cl c-l

ý (V

i ci 1 1 1 I Cr) N 1 Ln

d' 4

i " ol 1 4 1 1 I r

N cl c

c-1 C vi

C ýp -1 N c

r4 c ýý C\l (V

;; r l

CY) -4 4 i -I

-4

N

d N c

W cl d r

A T

N c

N c

cl cl c c

1 T '9 (\I -I t-l CV X-4 -1 i i

0C ý -ý c c r c c

N (11 lD tf 0) CV ci l0 d' d' N

E- 0 1 o ý N

i d 1 f

i l` c-I

CQ C+

t c-I c-I 0

c c-+ c-i

u c-+ c-1

C ) vi

ý k vi

N t4' f") t1) l0 '3' C11 (V d' to c' 14 H c c-ý + c c-ý cý c-ý c+ rº c+ W ö lD O W L N N t

I ll ci 4 N 10

t ' ' c 0 U ri 0 ri ri %-I C-I ri ri ri c-i t-I

Q

U

V] m c-I Cl cl j ri CO ri ý

r, cl C'l '

ý 1 1 T-4 tn OD 0 co a c- ý 41

d' r 'O tq C), % ON cl

f L- w

CO N op

1

i C0

d' OD CO O o

j C r4

p

I d N 11 co 1 1 1 , ID I- l

ci 1 o l

N 1 Q'

N N L O m cm .

N Cn Ný lf1 C") tt1 f') (V c-I ko d' ca <r d' tN N

`ý d d d d d d d d d d d d d d d d `ý

Q u w w 0 x H a z z O ör c` cl ä u w

z U H e-!

O a

Xa HE

Hö d, ýw

$! a

" Q') N W t- co ON M N c-L -t; f' O '4 t u) ' " A 0 N co a N r. ' Ln N M w 0) a) co N O {- A M N M M V' M U) 1 M M N Q' N (") C4 4 Cl)

0 tN r S H to tN c+ Ln W u-) 0 c1 O tV co ä '; r 0 ON ýO co 1o N O O O t- d' k0 %Zv -zz3' Ci 0 co N ci O 0) c'4 M N

Z

G 7 c-t c4 r4 c-4 vl rý ý

.

a a u < m u w w co x H a .: Z 0 oar a cy a

0 M (. o %D W W O O "O O ' to O O O 10 N ri N N N N N N N N N N N N N 0

tÖ N N d N Ö fl d { + W ý

c-q c-4 (V N CV N t 11

l c-1 (V

pp r N -I

pp ß -1

O 1 c c c N

°' " -I

,I { ýp

om, N -4

N d °'

C \l ýp sa cý N T

ci i i ýa ci c

c c N

0 ß 4 h T Vi c -4 T4 N c 4 cc'4 c i -1 T co 1 p 1

i ' A % I co Cl CA I,

IH c N ( IH

(N W c-4 -1

T-4 T-i ,. . V C

(V rI

u-i ¶4

LI) ¶4

N 0

Ol ¶4

N r4

¶4 ¶4

I c4

Ul ¶4

N r-)

In ¶4

l0 ¶4

N ¶4

¶4 .ý ci ci O -+ h + i n

c-4 ci ¶4 c-1 ¶4 ci c4 . c i C"4 M ¶4 c

r1 ¶4 u ri i "

co N . -4

Cl ¶4

N ¶4

u) ¶4

N 'r-+

O r+ 0) ¶4

N ¶4

Cl ¶4

Nr t-+

w

r-, I

Cý 11, 14 J,

Y ci

r-4 r4 TA r4 ri co ( )

IH T"I

ON Tq ri IH -4

co tO ö (A

Ö °D r- I I ' o o r a o c ri +

d' cý vl N T-1 ON co

chi t" I 1 O j °1 tl-

' t- 0 N T-1 4

O d'

ci u co

e) 10 C

O

! ýi

tc 9 trZ CO {

co tr) '. D

c, OO

1 t") l0 Ct C

In Ln Ln 00 it) co N 1 M QO 1 1 I 1 M 1.0 l0 d' N to [ tV

N N N M N l0 N fV N U) t(1 Cl to to N

d d d d d d d d d d d d d d d d

a a w u W w o x H z Cr cf o

w ý,

ýý d

Hý

0 qw N

. -1 c N v N O I- I- 0 tO to al N r-1 l0 01 N ri CO . -i ( N Ul O N LO Co Co O M CO o- (

M N M c}+ cyv m to l'7 M M N M N mT: cN : Y' (j

r- D N -' I- M qr 1- O W I- CO CO (0 to cI' C' N l0 I- m bt M tf) N tI1 l0 N M N H f

0i O Ul 0 Ol CO N 0) Co N H 0) 00 N cf' r-4 r-1 ri ri H H H r-1 H H

ýC cn U W C=+ C7 W H ý7 H 2 O Oº ä ä Oi

O CO tD lD LO O tO O O tO O tp O O O to o N . -I N N N N N N N N N N N N N N O 4 I I 1 1 I 1 I 1 I I I I 1 I . -i Iý q M Co Cl '-' tO l. 0 O co m to M --I H r-1 N N N N ---4 H N H H . -i 1-1 N

lO O N 01 ON 01 M 10 10 On r- co to co Cl O H r-I N H H . -i N H r-H H r-1 H . -1 H H N

O1 n I I I v ö d rn L N ý co to t ý . -H ri .i N H H ri r-1 . -I r-H

It: N -q to to 00 O d' M I- Lo i-n CO I- co co A 01 N H r -I H N '-1 H - 1-4 H H H r-i H

Co I I I I I I I I I I I I Co CO 0 10 C- o co N l0 I; T N l0 Iý r-I N 14 r-1 - H ri H r1 -1 H i-1

N ON V' CO I- 00 N r-1 lf1 -j M -I In r- to 1 00 H r-1 H r-I 1-4 r-H H ri H H r-i . -H . -i H

n 1 M - M N p ý 0 0 N %. D 110 l ' H '-i 1-1 r-1 H H H H H H 8

-4 I- N H tO 10 O O ' N r-i o d' to ch '-1 N H H r-1 . -1 H H r-1 H H r-4 H r-1 H , -i

I I I 1 I I O l0 -4 O N ON m O 0) M N r-4 H ri H r-1 1.4 '--1 H H

Ln a C rI N H IT 01 to H r-4 Oa r-4 .1 01 CO 1 . --I ON co H r-1 "-1

i-n I 1 I I I , I I 1 6 1 I f , t. f) O1 00 H M CO [ H C - 1 M H

CO H -4 N CO t. f) N N H co N tO [N IN

ý -i CO

I I 1'' I I I I 1 I I I I d' N N IN tp r-1 N o1 O l0 to 0. lfl

H O H to -'}' r-i lO tO "--4 to to Y C co U1 10 oo -4 ü7

M I I I I I I I I I M to O t0 ('1 M ßf2 10 " I- "^

0 ,* CO (A d' N co d, C M to r- CO 'd' CO

N I to ( 1 N M d) C 1 %D N l0 a1 N

N N I, N H N M M N 'V Ln -I CO tn CO H H 1 I 1 I. i 1 1 1 I I I I I H O O O o o O o O o 0 o O O o

r-q N M o ý pq 0 W w t, x H º-a 2 0 C 01 a a a

i

01 M N d' Ol 01 to O' M to N N I, c'1 t, m "ý O O cYý . -1 a1 to C l11 '-I N N O Lf) to N Ä

M N -zr T Ln M LO M M M M M M CO 'n Cl)

. -1 - 'V 00 N r-1 II) cal ICT O lD l0 OD u1 00 cM O N tD to 'r r-i O to (N 61 Lfl O d1 r-1 N

r) O CO to N .1 M '7' Ol O M r-I -4 . -4 I-I r4 r-1 -4

O M Ol _4 M -4 r- i 4 -4

m u w w 0 x H a Z z 0 1Y äN c+ ä w w

o M to to to C) o CD to o to o 0 0 to O N . -I N (N N N N N N N (N (N N N N N O -4 1 I I I I I I I I 1 I 1 1 -4 N N M N d' LO Co to . -. 4 Co d1 tN m N

r. 4 r-1 N N N N "-1 r-t N '-1 H -a -I N

to (N -4 M 01 lqfl N 1; 31 O N CO to CO M fN . -1 N N N r-1 N - 4 N r-1 r-I N

I 1 M 0 Ol Ln O " to t to Ö

N 'N N N

.' r-I ri 01 O CO N l0 m ßl Lf) LI) Vr N Co 01 r1 r1 N r1 N rl N i-1 r-i . --1 ri i4 p-1 r-I .4 . -I co I I I I I I I I I I 1 I QO O Co N . -1 IV CO d' V' c'ß'1 to tý N

(N . -1 r-I N r-I r-4 . --I . -1 r-I r1 . -I . --I

M O ß1 1- to N O M N I- M co N u1 to to . -1 r-I r-1 . -4 -4 N ri r-1

N I 1 1 1 I 1 I I 1 I 1 I i N N CO tf') to CO f N O to .1 . -1 Cý ý ri r-I r-i -4 r-4 r-1 r-I r-4 ri r-I ri

M U) 1- H v) N O o M to ON rl ý4 e-1 -I r-1 rl 0) r--I r-l . -I i-4 r-1 rl O I I 1 1 1 1 I I I ! 1 1 64 q. Ln q CO

1

U) O Lff N r-i M d' M O N M . -a . -i 00 . -i i1 r-I rl . -ß Ol N -4 r-i CAD 01 r-1 . -s . -i in I I 1 I 1 I 1 I I I I I u1 ý N 01 tý C O CO -4 r -I -4 - r -I

- U)

1., -1 -4 ý4 C14 CD CO NN -a CO IN tO l CO tO N I

CO N Q N CO 1 I N 1 [

Ö k. 0 L ) to 01 Co r-

V-4 r-1 1-i H

^ý O tO to '--I CO to 01 I- u1 d' ' -4 N mot' if co 01 l0 CO I M

If) O l0 tf) C'- to r 17 N 00 ":

v

N d' 1 0 Ln W to L'1 M M C) Ul M 111 tD N M N 1 Oý

N "r Jr N 1 N N l

M N er N M M N r-1 N W N ey" qw , -f 1 I t 1 O

I O

1 O

1 O

I t 1 1 H O O O O O O O O O O

a a r-f N M ý tý0 Cý W w C7 pC H z O E a a a o a w w

10

x O

z

U d1

3

" c-4 o% ci C) cI to d' 'H ýp N N 01 Ol 01 ri cv) " (+') r+ CO L! ) LO N 00 LI) d to c-+ p I tT LO to 0 0 «) N M ;, '' 111 C') to M c4 M M Q' M C') C) LI) U)

N (n C) C' ß1 N M -l (V 00 N C-) (V O\ r N tD d' c-i co c14 c-i N U) N O 00 N w tQ 00 ý a%

" [N

" LI)

" fV

" N

" N

" M

" O

" co

" m

" Q

" o%

" 0

" ci

" (4

" ¶4

q W

¶4 c'1 ¶4 ¶ ¶4 ci ¶4 ci ¶4 C'4 ¶4

V

Q ¢ m cý w w o x H a z o `iu- ü ä

C

w w

o « L0 /o w O O o 0 w o to o o to N 1-4 N (V CV N N CV N (V N CV N N lV O O ý

i

ý

N N ¶4 <V

¶ ¶4 N H 9

1 9 all N O

r 1 1 r

cn V-4 41 ; 1, W t-D N ¶ C- CA (V N' LV CV r-i i c t -4 ¶4 ¶4

ON Ln ¶4 ri -4 N d N ýp cn N `

d' Ol o 'Pi A m O N a% c

CV CV (V c

N c4 ci c-4 t c-1

' 4

I ¶4

ri tV %-1 CV N c4 c9 0 ¶ IH ¶

cý d

S J

a O c i 9 " o 4 t + 1 co r-4 CV ¶4

\ ¶4 c (V c-l

c t-'l c t

c-4 ¶4 ' c pQ c c

I D ) r-I

4 d ¶4

N A "i

4 n )

IH V4 .) c'{

l ¶4

l C4 c'i

c %-I ¶4 T"4 ¶4

c ¶4

I c4

f ¶4

W N v Ln dq 010 to ca N NN d' W a ¶ ¶4 ¶ r+ ¶ ¶ ¶ ¶+ ¶ ¶

-4 A A 4 4 f o X W ' ' 4 1 ¶ % q ¶ -I ¶ r % ¶ ¶ L

E-f E-4 U) N N N 0 O N N N (f]

0) ¶4 ci c4 r4 CO T-i %-I 01 ¶4 ¶ c4 c-4 N Gp t- .4 OD t. 1 I C\l GI cc) C1 m O p

n ý- 1 t0 ý d,

; Oi e i '

CO l 0 co CI 0k t- t0 to

8 [ý lp N

1 O ¶4 N ý O LU CO

r- rl- %0 Ln C\ tn 'D r- (71 kD tý to ttl IH

5 ä'

-4 O ' n iD p ) M D U a u

O N

C d' M m m l'ß'1 IN u'1 M d' 4 1 A N

tf

N M lt) N . 41 N N (V to ' N M N) v N

`i d d d d d d d d d d d d d d d

cm 4 as v w w x º-ý a z o Or (Y en ä orr w w

11

rH-t Ö

aý w ä cý

z

O v 0, O N c-I N in rý N ch

Ö ui O rn "

Ct) "

(fit "

N "

Cl "

d "

C "

tý ) "

to "

M "

c ý) "

M "

M "

d' "

M M "

C') "

d' "

1;

In Un Ln un (. c-i in 0 c-+ q O cn c1 to q co 0 C1 M m to L! 1 N CC) r- iý l1) in C') U) N C'

ßý l0 c 0 cI c1 cI cý c c'I c -l t- c-ý cý

O Q E-i o 4 Co U W G+ O x H ý7 ? 0 C7 C7 0 Cl U

O cn tO ýD 'O O O in O t0 O O O w cm V4 N N N N N N N N N N N N N A N N N M ON M

N t 0 I i d 1 1 Q j 0

TA "1 N N c N N c-

. c-i ri

[ I N

ci i

ý ( N r c c c

IH C\j (3) w r- VI N

{ i I N l I

N Lo -i -1 ' d

c T C c-{

O N

pp ca c-I

T c-{

t l0 -{

c s c N

(IN c-I "V d' N U) O\ N

co m i ý d I t-

i d

ci m I .4 in

+ I 4

-i ö CO

c N c-I

a0 cI

c N c-4 c-I c c U) cI vi

c c-4

9

N Q1 l0 N l0 a% N N to fit' C') to N Oý c c+ c+ c1 ci c c c-I VA c-I c-I

h h h h -ý ýý lý

i c p

i -q H c i c i I c

ri I i c

c % T T r

W Q N d' In d' lD O 0 M N N c'') d' [N W 1% 't. 4 c- c-I c4 ci t'i %-4 ci c-I VI c-I ri i7ý O t0 I N N N 4 ß1

41 N

'A A

-It N O l0

U V!

l i H 4 i -1 a 4

% H

t q q q c c r c c T I c r r C

E-4 E-4 C') O 0 O T-4 T-4 vl

ca ri N-4 c4 T14 co ca T-4 ri T-4 x-4 LO co tD '

O1 d N

i c4

O , ) O c

() ? O cý l ol (A r' a% d' t- Ln cV Oý CC) c0 OA 00 CD llýr

CD 00

I

`~' A "Zil ö ri IL

I ýý .4 N

r in ' cý ý I. co

e ()N l0 Cl'. U) It W to d' 1ý N N I

tni N M

A 'A t1 alt

N N

cc) N

N ch h d' m d' m IV N d' 'd' N d' tM l0 r4

r' d d d d d d d d d d d d d d d d `i

ü < m cý w w o x H a z z o a cN1 c CY U

w w

12

a

in O O O O OD O\ lqt' lD lD to O in U) 00 M C O O O ( 'O (N CV 1 CO --I . -i . -I , --i Q

N N d' C -V M Ln m M M N mot' m r') M 'r En

CO N CO ý. O Ql 00 O Ln in cr (N N f'ß'1 C1 dý Q. Ln ßl M e1 tý dý m Lri 01 M . -i v l0 N

Ol ýQ d' N N N ON CO -t O O "-i N N

< W U W r%4 Ci x H a z z O a ä ä a b.

O cn I'D 'o lo O '0 O O l0 O w CD O CD %Q O N . --I N N N N N N (N N N N N N N N O r4 I I I I I I I I I I 1 I 1 1 1

. l M N m 01 Qý N l0 CO N O N 00 01 . -1 r1 _4 N N N . -t N '-1 --I r-I rl N . -1 r-I -1 N

t. l) O N -4 N co Cfl lO in N 10 01 o N OD 0 "-4 N N N . -1 N r-I -q rl ri ri r--I ri N

01 I i i ; r ß

1 '- N -I

N N -4 . r -1

M O N O N H d' M lO in N in lO N m "-i d1 N i-i N -i N . --I e-I r-I 1.1 1-4 . --I r-1 . -i r-1 I O , M ! 1 00

l C M t1 lÖ J'

in CO . -i r-1 -1 r-I -4 rl . --I -4 rd

N co %o N "o co N N d' -V In M "p r-- i--4 co r-I ri r-I rl rI r-1 r-I r-4 r-1 r-I -4 . --I r-4 -1 N 1 I i I I I 1 1 I 1 1 1 I 1 1 N '-1 N ýi' in in lfl ri r. 4 M M -V N M in l4 r-1 r-1 r4 .-I ri .1 rl ri .. 4 . -1 p- I r-4 .ý r-1 '-1

O l0 m -'r d' Hfl O O N N m -4 -V to rl N ri ý-I r-1 , --I 114 r-1 r-1 ri . r1 ri r-i r1 ri

1 1 1 I I I I I I I /o In N M f7 M p> - N M M

Ct] rl r-1 ri r1 rl H ý-i . -I e-1 [zý

8 ß8y

eT 1 N N N r-1 ý4 a -4 C, 4 z 4 r-4 -q ý4 -4 m 00 O 1

z m u 1 1 I I I J

I 1 l

I in Cý7

CO ()) O O r4 C. O o m O U) r-4 r-4 ýl r-4 r-4

C4 C) -4 N ºn H 01 ON O\ 1ý N ON m 01 co -1 r-4 M ý ä O

. 1 1 C w OD N N Ö1

Cl N . -1 r-i

1 l0 Q' OD N CO N %. D to 00 N t0 lD CO 61 l0

("" I I I I 1 1 I 1 1 1 1 I M 61 N tD N %O in ý" N Ln r- CO in

to en 0c) ko to t0 Ln 41 M %D t0 V in t0 N N d . 4 J ' ' N

' [ r L M

t11 ý1

en N to i M d' M ; N in in c4 V' C' l0 N r-I b b b b b b

O Cl) l l l b O O O C ) C ) O C )

a O N M

ý as v w w cý x 1--1 a z o a Cl a C

UU

13

J

0

H

y

I I IV M N z

N tf1

I

lJ M --! ' N 00 01 M C' N C' tip CV "-i M u lD CO N O -: 3' O lo cM N CO 01 co N -i O Q M r-I

4 4 4 ('n1 LA M N M N

4 M M Cpl Lf1

O CD d' O lO M "Y" eM m Ln I- " . -i LA M lp 1ý M N N O N 01 . -1 fß'1 CO N M LA O 00 1, Ln M lD _4 'r l- m -i Ol m Ol 00 O

-4 rI ri rý ri r- I r-1 rl

N+ ýi

4 W U fý C7 Z. H ºa Z Z O OI ä Izz, a

Lý W

C) M to lD 'O O I'D O O C) l0 C) O O l0 O N -1 N N N (N N N N N N N N N N N O -4 1 1 1 I I 1 1 I I I I 1 1 I I r1 t0 N N Ln N to ', q LA 4 1 Ln cl> N Ul C O M

r-I r-4 N N N - N r-E r-t N r-4 N N

LA -I M -i d' lfl -v t! 1 O mot' C LO N N r'4 1-4 N N N N r-4 r1 (14 r-I r-4 r-I 1-4 1-1 N

I M ! 1 CO 00 L) -4 N N . -1 1-4 -4 -4 N

. -T O N m M LA N M C1 N M N M l0 O r-1 -4 N -4 N rl N ri r-1 . --I -4 . -i r4 r-i r-4 (N

CO 1 I I 1 I I I I I 1 CO M O CO N N tf) to N LA O r-1 N r-4 N N ri -4 r-I -4 r1 r-{

N O%% N r-I d' O -4 N N "W M -4 -P 00 , -i Ol ri -4 (V rl N r-1 ri . -i r-1 r4 .i ri '-4 . -I

r 1 I 1 I I 1 1 I I I I I 1 1 N 1 '- CO l0 C D U) 03 O I - . M M N O M lo

r-i -4 r-I N -4 -4 . -i -4 -4 r-i -1 _I r-4 ri

O N LA 01 N N 0 N r-I N -1 N L) Ui rl CO ri r-4 r-I r4 r4 C\ . -1 rl '-. i r-i r-4 Ol r4 -4

ýO I I I I I I I I I I 1 to 01 Ln CA N Ln CO . -t O '-i O r-4 d'

. -I . -4 ri 1-1 r-i -4 . -I . -1 . -1 r- (

ß N %D r-4 -4 CD co N r1 . -4 -4 rI t- 01 01

-4 tT CO

ý r-4

ý I I 1 to r

ý CO CO

CO N a r4 r-I r-1 . -1 r-4 r. 4

Kr CD C) ýO .4

; '-4 m l0 N O N tN l0 00 i

CM if1 O M 01 LA

CO Ln tO N

CO

ri . -1

. -i N ,w LA r-i ßl CO Co C LO N t) ý, - Lp LA tO r-

co 1 ! I i { { t i E I O\ O3 O 1ý t0 l0 LA M LA V+ to

r4

M co N 01 tG LA M lA tn N -v M N 1 i I N

C C A N

C. 4 d' %D i d' M i d' M M M N M M

b b ; O ö

0 o O O a a a ' O O I a

D O

N M "Cr 0 RC m U W W C7 x H 0 H 2 O C Cl d Of O Oi

w

14

0

N

d z

cn = c12 N

Ha O

a ätw

{ý+ cr-i

l

91 4

" c{ t! ) l0 c-1 N l0 co M ul m m l0 c- M ON " R o0 O O o% h CO %t' N OD oa h to o C CY) k. 0

. . . . " . . . . . . . . " . " U) M N d' m T N to M N M N Q' N M M 4 U)

ON N a0 c-+ co d' M c-1 0 tN q in 0 o C t0 Z M h h ; f' l0 '. O M c-+ N lD O co N N N Q FZ

[ý 00 h to in 4

r+ in co C\ O O C' 0'. 0) r1 N ci r-i c- T-q c1 ci cI ri c-I

a 8 d

U ¢ M U W ý, cý x i ,a z z 0 ä a w w

0 in LD to 1.0 O '-O O O 4-D 0 ' O O O N t-1 N N " (N N N CV N N N CV N N N

O 1 4 1 4 I h

4 1 tp

1 U1

I A [i

I Lý t I 0 0

-4 N rI N

c CV N H -H N c-1 %-i N T-1 T-4 r4 ri

pp ci

N

N ri ri c-i r1 CV cr% U)

- TA M N ! M

N C' -1

I LO -1

' 1 I h i

' 1 r ri N

N c OQ t

A Ln r a

V

N Q) cl w N in h C" [' 01 c-ý N c-I CV Ti N T-i cI c-I r-4 N 0 1 ) ) I

i l N c4 N A WA I

N co

r-I O pý O t C v U c T vi CV c-I N ci c-I c-I c{ rI c-I cI

C\l a% r- ý N ý c-I Cl

c %-I T-4 TA c -I c -I r-i

h 1 A

t`

A i I T

T-i X-4 T-4 c-1 ri cI T c T

0 h in lS in tf) N t-I O M ct'

-H c-i c cI tl 01 c-+ c-I c cý c-I a

-I -4 Vi T T

r EA E-4 (n Un c+) d' M O O U)

OD ci rq IH c c-I - h

ON ci c-i ON O-, I co

c'l in N

V' N (fit c -I c CO 0) 0) N co 0

c-I c-I r4 ri cI

en r4 T4 '.

O cl c 0 co %D h o )

I W i ' t O 0a W ui

t- h L ID 01

i CO u) c

d o n

(31 C- m Go M Ij IS) t l I

U N kr,

W M

U1 4

I U) ch

r4 co CO

I'D N N ii' ) 1

4 j h ( M IV U) I') CV d' M

LA d' N

CD

c-i ('n h U') to to N (V M V d' r-4 M lV d' M

`ý d d d d d d d d d d d d d d d d `ý

a ap a co ü w w O x H a z o

ä Or c 0

4

w w

15

z Ly w

LA HZ c)

C\l wD 0

ä cw,

Cl) 94

H°4ä

" iD L' of c+ tN O to m in N N in G' O O c-+ " n (O o co 0) t-o o Co co m M cs, M r1 N a " a " " " " " " " " " e " " " " s "

lh c-4 Q' M d' N to M N M N v C 1 M tf) U)

M N In M l0 ; i, 0 CO M 00 tf : Cr, c-) C') N M C ýn ýn ýý N rn M O N c rý L in C: -N N `c

I (+, ] C) [N Le) M M N C') CO 61 C O C1 CO CO CV 14

o < m U. w w u x H 0 (i CY Ci ä v

w

o M CO 0 Co 0 0 0 to 0 0 o w N cý N N (V CV N [V N (V N CV N N N (V 0 c-)

i to s-I

I N H N

1 c N CV

I Ol ci CV

1 CO c-1

I CO ý

1 00

1 Vý c-)

I Q CV

I LO c-i

I CO c-)

I Cv c1

I N N

O Ni

d N [N tD O. ) t. () ci cý N 0 T-i c-i c-I N

m i 1+ s-I

c-+ ca

() N

O N

1 14 CV

co H

M CV

v' vi

' c-q

I lý c-

I LO T-i

I tO N-4

d' c-I

I d' v4

I 1 J

(V N

(n c-I

0 CV

N c-I

N N

cr) N-1

C') c-

if) c-1

(n c-I

0) ci

co N V-4

O c')

1 ci (V

I a ci

1 O) ci

I to c-I

1 s-1 N

1 N ; -I

1 N c-I

In c')

d' c-I

I 4,

T-i

I N 1H

M c')

N c-i

1 p) c-i

CO

cý [-i

I O c-+

C

0 CV

I CO c+

N cl

1 )n c-I

CO c-I

+ kD c-1

to c-+

0 N

I CO 174

c :1

I O c-I

ca c-+

c-) I

C') c-+

M c-)

C+') c-)

1 N c)

c- ci

I O s-1

(V T--'

I c-I c-I

l0 -q--i

I ul C-1

N

c c-I

W cN '' to -'T N N c+ O n) . ̂'. iD G1 W

ri c-i A

cq 4 4

VA A

c-1 0

c-i

5

c-{

8

cI

Oil c N-1

TA

C\l

OU

U1 lH T-4 c) Irl cf c-ß

E-i v] ýn N cý ýY 0 0

N CO

1 c-t IH

1 c

ci Ij

N ri

74 I

c1

CO t

N

0) I

a)

V-4 I

00

cI I

01

M I

CO

CO 1

N CO 10

1 N

vi 1

m in

c-t cI

M c-I c-ý Q T-i

IT I m to 1 CV O

ci 1 O 1 Cý I CO

tO N r C) U, j

r- l0 CO

T-1 c-ý c'{ c-i

c-)

M C

M

i

O '-+

Oh

cc

m

C)

N

'L

+n 4

CO

u1 tO N

1 ID

ýt uý in 0)

00

1 rý

N fV d' (7I

co tý I

Ln

rI

N I

N cl) d' d' V)

ýI

N d' I

N 1

IL M N

H

d M

d N

d lO

d c1'

d tO

d c-+

d N

d M

d M

d ýY

d c+

d en

d cH

d to

d N

d `ý

ä Q

w 4 co U w w t7 x H a ý. z o

Vi or

N a

M (Y

". p) Q

w

16

Cl) V4

H

z 0 H H

C)

U) H H H CW3

W

c-ý

z

14 In t En ýc

d

W W

" CA o U) a' 0 CO H co In CO Cl a' 7-4 in In 111 A is O' co r CO tN c+ in ul O In N ci Co c-+ A tl1 M IT m 14 in

4 to M N CC) M 'r N M c') to v}

N Co LI) N (n -t' 01 0 O N M co Lo m z co C`) c-I d' 1. O M c-4 l0 Ol N O N CO CO C- CO 2 N Cý

in C N C') Q co U ci Q'ý N N N c'I

c-+ c c-+ c':

Q. Q ¢ O w w 0 w H a Z z o ä O O u w

O M kO CO 1. o 0 1. o 0 0 0 W 0 0 0 N N ý-+ N N N N Chi N N N (V CV N N (N (V O 1 I j ! t 2 t 1 I I I I 1 . 0 k1l N c-I C, i --t CJ -i (N

U) M a' Co Ul IN t- a' c2' N N s-I N ci 'ci H c'I '1 C%

CY) ci 1-1 N

co ci tc'{ N Co

I 4 11, d G'

c-i CV c-I N c-I V

c+ c-I 00 c-I

c c-I

c-I C\!

cI 0 N tj' IV) in N N N Cr. C \l CO M O

Cp I t

C\ 1 I

I 1 I 1 00 (V ci C I c-4 ci c"I

r -i

ý a ' ti Cý

i c c c

O ca C-4 T-1 ca cI c ci c4 c{ m co

'4 LA Co ID

'I A

i 1', 8 " A A O O 1 C`

c c N C) 1 ' , ) Lf. T-i r4 t-i NA

cý C\l W

CYi Oi ri c-1 c'i c-i c"I ri a' 01 %-I O t ca I 10

1 Cl)

1 A

1 1 1 11", -4 '74 f 1 I I ý' LO U

c- ý c-i c-ý cci v 61 CO co ri

c ul

H 0

Cl) Ul) (N N O cat r+ CO ri co in C- l- I (

O { ( co 1 0 N N N ( in

V ri c-I ci

c cc-II 0 c4

OT-4

M O CV Q ri

d ai ri 7-4 co j

1 1 O) .W

1 I Q)

%D [- Oý cD t0 A r-I ý! '

L O

c+ O

CO M U) (S 1 Cý h

t ýO

10 l0 1 1 1 `

4 d' I I CO

T-4 in V co

(y) zp N It l! tfI

CO in V 4 d` !

Co CO 1

I N Il) N to

CO lD V' N L' -' C') Co CO cj` CV N N C` N

° d d d d d d d d d d d d d d d `4

cEq

a CA U iw Cý x H I R". O V4 CJ a

Cy Cl) Cl or Üi

is.

17

H

qH

0

cu

baW 4

" rr C)

O lV

N N

d G*N

N m

Co CO

d CN

m N

rn qr If)

O co

to N

m C)

N U")

co c+

0 Cl "

+ Cr) N M Cr) N if) Cr) N Cr) N c}+ N c*) Cr) d+ + Ef)

in Ol 1,0 t- C) r C) in O N ca N C) C) d Cr) Z 0 N in c-I C) cc- 0 V' Il) C) N N 00 -. J c1 N z

q C) Ul ¶ O M C" C) C) O c-ý C) N Q1 M 0 c+ c+ ci r c+ c-1 c{ c-ý : z:

E-4 U < rQ U W G-+ 0 x H .4 X. Z 0 TA d

C\j O

m Cl 1: 31

Cl E-4 u

w w

0 o (V ch

c1 41o N to N (o N 0 N (o N o N 0 N (n N 0 N tO

(14 O N o N 0 N to N 0

ci cý c-I

4

N ci N

GI c-i

oll c4

11) N

Iý e'"1

4

r-i G1 c-i

OÖ Ö)

ri

4

ri 1i c-I

Q) c-i

Ö

N

110 rl T-4 4

O c

N N

N co C) N

I c a o

to r-,

IH I- ý

! IH N

%D V-1

C) T-4

c-I I Q\

t N c-I c C \j

t c-4 c-I

l cI c4

co

C l- IN cm d' N to C) C) U3 [N

N N to 1 i

ci ci 1

'C M 1 to %-, 4 71 00

ri c-4 r4 N ci cr ci ri ri ci 1H

ri O in d' to O M M M w in ci CV T-4 T4 T-4 N N O VI d' 1H 0 c TA vi

c+ CL)

I

r4 -4 c-I

I

c-4

I c-4 CI

V"'ý t

Iri

VI C\1 cNi

VI I

vi

I

vi ci

N

N Cr) N Z~ ci ri ca cI CV W

("I Co

1 c{

r4

c-+ 1

r4

ci 1

vi

'' 11-1

V4 a

r4

r+ 1 D

T-4

6r vi

1 C) T. 4

m s1

vi I

IH

01 1

c+ 1

T4

d' c+ I

V4

C w 0 U

z E-4 En OD Ln 1-' vi

0 VI -+ cM H Ql c-1 ) r4 c+

0 H

ul t l0 ! N N 1 CO O\ c1 N C) 1 tf) C) c-

cI cl ý

I N

c-1 CO

t0

(N q N

ri c-+ C- CD c0-i t0 `O

co rl,

cýIý

c0-1

tN d'

VI C

c'ý G

V' 1 O

` (D

to

to

-1

O) u'1 ý1 to tJ

l0 O ý4 4

u1ý ý+

c-I

N N In (n

IL

In 4

d' A

(0 d' A

d+' 'I

tr) A I-

cM N (11

CY

'A

CO

9 N N

T4 VI

a N

a LR1

ö Cr)

8 ä! 1

8 LO

o N

ö fV

ý M

ý N

ý lD

a c-I

c N

A N

8 t0

A cI

a V Ö Ö

E-4 ¢ w v w w cý x H a ý. z 0 ä ä Cr) ä u

to .

CO CO 'r 00 Ln N to N In N l0 I, ( W O M CO l0 ýi' O r1 M a Ln CO N N r 1l CC) l0 l0

m -1 C cp Lf M LI M N M ýM lý M M tf) CO C7

. -i --i C) 61 Cl) t. O 14' LO . -i I'D O -4 00 O I'D r- N M . -4 O 1ý 1 O N LL) c ll ý Y) Co lO -4

co 00 l0 M (V N d m m . -4 O Ol C1 00 N -4 r-1 rýl rl rl r'1 r--1 r-ýI . -1 eý-{

FC4 m U w w 0 H a z z O O+ of a s w

O M W LD l0 C) I'D O O W O LO O O O W O N -4 N N N N N N N (N N N N N N N O

1 1 1 I I I I t 1 I I I 1 1 L I 1 r, l0 N -W P-4 dý Co to N lD -i co O N r M 4 rl N N N 4 N -4 -4 N . -1 N r-1 r-1 -I N

tl "-I M O M 1N v lU ß. t1 O IN 01 %D l0 CO N r-I 4 N N N N -4 '-1 N 1-1 1-1 r-4 -1 1-4 N

(14 m 00 LO Ln C) r1 N r-I -4 1-1 . -I I--i N

C1) (N 61 -4 W M LI M Co lD N -: It N Cl

1-1 N -4 N 1-1 N I -q --1 . -4 o--1 -1 -1 .4 r-I '-I Co I 1 1 M 1 CO

r -4 CO Ö1

1 4 N

lp t1) dr M M

0 0 N r-1 r'1 N -i -1 1-1 r-I 1-1 ri . -i r -I

N O O N 00 t. I O N -i LI M N N W r , -i r-1 N ýr-1 r1 r -i N -4 1-I r-I r-1 r-1 -4 r-I r-I r -j

tý I I I I I I I I 1 I I I I l- '-1 Ol lD LO V' Ql IV M N -I Ln In

r-l r-1 rl ri r-I r- I r-q ri -1 . -4 r-1 -4

O co l. f) to ('M CO O O M N "-i C) d' qr r-i Ql r-i . -ý ri ri r1 ri ri . -i rl r. 1

1 ! I I I I I I I I I '. D a% I- V' N 40 Ch N . -I co O U r-q

ri

8 8 p -4 M C) co 00 :4 p4 ý OD O, 1-1 ° U' m U) ° Ln I I 1 I I I I I I I ! I I L1

N V -4 -4 Co O N Q) N CO (fl -I

p ° -I a 1-4

M ° t0 1ý r -I

° 01

° N N a1 01 lfl N l4

4 N

d' i 1 1 ! I I I I 1 1 1 1 1 1 -º' eM N 01 CO dl CO lD U) OD LI lD u) O) l0

-4 M . -4 CO N CO N L lD ßf1 --1- CO to

Cl , o i ;

M N lfl d' \o l0 M N C` 1

-4 In a% WO lC Ln co L Ln U) N mot' Cl l0 N N "

; J 1 " N

! I r

N .4 1 . r N LI

d' tD LI Cl to -4 N M Cl dl 1-1 M . -1 ei0 p-1 .i i . . I ! I I 1

O I

O b I

b 1 1 ! f ý. i O O O O O O O O CD O

PG ý M W P4 !' x H a z o

ä ä ä ä

19

w

w

H

z 0

c: a 0

x EP) H

w

N

z Irl U U,

HU

" C tp d' c+ u) ýt ý+ O O c) cý cý "

a) co c-H (N Lil N "

tD v) Lf) H C') t: CV d t! ) GJ q C)

" (r)

" C'S

" 4i

" Q'

" Il) Cr)

" LI)

" Ci

" N

" r' , ')

" 'v

" M

" c*l

" tr

" "

H N N LI) O Ln Ln N N r-i ,v G\ ('V M " " " " " " " e " f " " " 7 " " "

q

H H H H H co Gl H QQ 01 (J

vN-' cv-i [

H H cN M '' EI

.4 fc U iiý t0 H a Z, 2, Q 0 C Or O U

O M 1, W LO 0 W O O to O 0 O O O W N c'H N N ('\J CV N N N N N N N N N CV O

-i I -

l 14

I 1 1 t

I N

I fit)

I I 1 I I 1 1 O T

i c-i

N i T-4 N

H N

in N

a) c-1

) t (V H-1 H

H CV

OJ H

(T c-H

N H

In rH

01 c-i

CV N

c-i

flo d' d' to O [N co ur c+ ci N N c-{ N H H ci C'J i i I l 1 CO

Lo N N c H C u H co ', 0 N ul 7-1 p

H tV N H H H IH H N

fV ß1 N CV 3' C') tN in tp cr '' N c1 cý N v-I N N c'H H r H rl c-i H c-H c'H

W M 1 4 c10 -0

A 1 CV

I I 1 7 Ö co 1 N N

H I N

7 c-i H

II c-1 c-i

1 1 ci

c c 7

l ý-. 1

N H

N O N 0 in 0 N in d' N tO c N H N H N H 1 r H H c-1 r- 'A

H O1

, 1 0 rä

IH c

1 d'

Cl) (V -1

rl -I

4 N

i c- c-I Ci co H cI

0 H t-I H c c

H c H

0 O in N N 0 Cr) N c`I ci 0 CY) 'c'ý ci H s-I c-ý H H H IH H r1 c-I

410 ý*1 :

( Ol

t 0

t G

M ! i 01

Oi -

1 ý

1 d I I I tW U O

1 c-i c-I c-I

c In c+

c N H

t c+

O H cý

Cl C n c-I

N H O

Ei Go vi r-4 Ti H m M

- 0) 0) 1H -H

H vi to

rt 4 ý N H I

rl l I I 00 I I ul Gi

c cc) 0) a N H ci H H H H

T-4 H T4 0 l0 c{ IH H l0 tO N N H a) '' i to

n 1 t ( 1 l 1 I q OD I 1 1 I (N 0) C a , O'. u) u) 1 t0 rn ti

H c-+ CD CO co r- [ to I CC) to M d' t9 C N

! G N 10

to - O i)

' " ' M

. l t lJ M C N L 1

A N

N I t0 1 M u)

vI u) cf W j N

O

H d! N d' N in M N d' M M tV U) N

`A d d d d d d d d d d d ° d d d d

E5 CO V CO W c7 x H a z o

ä Cf C Q

w w

20

4-

U cý a 00 en

a aý u

i7 N xwu HE VI

ý f4Sý7

H AEU)

O Cl) Cl) O) co N co r+ N' In p In O '1 O c " : LO Ol ON co N Ol Ll r+ N UI o, In ai H-+ (Y) ao n

. " " " " " " " " " " " " " " . " U) In c-4 In In V' CV In M N M N G' CV In C") ý'i'

C`') W

lD c'

Lf) N

d' Cý

0 O

N ýi'

c7 01

N Gl Ql

N N

CT) (V

l0 fn

c' M

L'1 ;o

Q' H

Cl N

. -ý

N N il T4

N 'l-1

V' H'4

N r4

M C-i

N CO C''" c-I

CS C; ) CO N C'H

c4 c-I c-.

O C-i cý N c") G"

O E-i

< 'a U W G C) X H i-=1 ý. 0 0 0 Lý a ,' Q 0-i

0 In G 1. c 1. o O '. o O O O '. D O O O CV H (V CV CV (V N N N N N N CV CV N N

O ca1 c-I

1I t--1 N

NI N

("I)

N 00I c

All N

I lC

i

1 to ýý

1 I H

I

s-1

1

cý

1

c-i

1 H

i N

U

cý O c-I

M CV

N I

Cs

CV (V

N H

In N

c-{ A

c"i

TA

tý d, rH J)

c-{

c-I

N

cI

C))

c-1 4

c-+

N f

p

C') ci

N CV t-I

N-4 N

CV N

N H

M H

L(1 c-H

"o c-i

CV ci

M ri

N c°i

(3) 0

N

c

1 d

CV li H

d

N IH 1ý CV 'A c4

CV H .4 VA

1 X-A c-I H

fV c-I cI i-I

TA cý

1 p

G\ H

ýI c"1

W H

InI

0) H

1h In c

0 CV

I a) c"'ý

O -H cI Iý

cH I

(N

H 1

T-4 ci O ca v)

q-4

W a: Oi

1-0 M Ch

co CO

N c1I

16 H

d' c-+

11) a1

to H 4

cý

d' c-I

A cý

N c-+

in c1

m co

0 ý-1

ON I IH

c+ T-1 rI

ci r1 d

cý

01

0 H

(A

d' H

cý

d' c"+ N

c1

w lam' O

iO

Cl) V)

Ln N

r- 0 4

c-H

c-I

c-I I

tV

H ý

H

H r- 00

H ý

H

GS ß1

I,

CO

[ i CO H

H

H d

c-I c-I c-H c-4 H 7-4 c-I

en

N .O

01

T-1

pO 0

0

00 I N co

il) kA

0

Oll i1 -41

In I M in c-i

r-I Cc) o, % tN C: 1

ý

10 d, I

LI1 U)

h I ý I

.? ' 1

N I

tf) C"')

N N ýi

O ý t- 1.0 N Ln 14

C', Ln 4 in cm (1

M ko CV

`' H

ö N

d [N

d LI

d d'

d l0

d CV

d (Si

d (I)

d C')

d d'

cS c-i

d CV

d N

d to

d N

d `'

o a

w M U w rz, C) H a z 0 T-4 ä cm ä m ä ä U

21

ýi m

a *lo

q

rn M N i!

14

uJ

to -4 "D h Co Ol 0 ßl co ßl N 0 l0 r1 r-i p1 O O [`- to N M c in 1 i`N in to -i M p

M N d' M -t' N Ln N N M N -P N M M m c

kO o- 01 M ß1 N l0 -I O N in N in N N d' M in M t`N 1. N d1 CO in llý . -l N 'P. O 4f1 C'

00 w r-I M N N 00 . CO a1 p O co 00 r-P N

--l clq Cl) Itr ý tY. l U W W C7 t-1 a ý Z O Ot OI Ot ot W W

O M LO LO ýp O sO O C) l0 CD \O O O O %p O N . -i N N N (N N N (N N N N N N N N O 4 1 1 1 1 1 1 I I 1 1 1 i 1 I 1 1 "-+ l: ) -I - d' CO l0 t! ) W L O cM lfl CO N

N N N -+ fN --i -i -1 --4 N -1 "-i -i (N

c } c

in q M O Cfl N M U1 -'T in .D a% M Ln N 4 r1 ri N N N r"i N r-1 r1 c-1 rI -4 r-4 r-t r-4 N

r 1 I I I n I I mot' i 0 0 i N

Ö

r-t (N rl N N r-i r-i N

M C) NN O t0! "-i m M Vf' d' l0 N N CO C1 r-i O N rl N r{ N r p-4 ,. q -4 +-4 p4 . -1 r-4 r--4 ß n O 00

Ö1 lD l Ö1 N M i r-4 Ui C r-4 ci -4 r-I A -1 rl r-I . -1 . -1

N CO In CO in CO N . -i N M Q -4 r- r-I co r-1 r$ rl r-1 r-1 .i r-q ri ri r-1 ri -f r-i 4

1N I I I I 1 I I 1. ! 1 tl- 1-1 N d' N +I` lD . -1 N m in '-I -i r -I ri r -i 1-4 -4 r-1 r-4 P-4

(o f+") l0 M in Q -1 O M ß b N i r-1 .i ri . -I rý .1 0' . -I r

Q0 -A 1 1 1 1 1 1 l l ! ! 1 " ß M rn ON C) O N ' Ql -f-1 r -4 r -1 I . -1 . -4 S

ý r4 ° r4 CNI

CO X) 4 r-4 -1 00 CO 0. C 00 CO in I 1 I I I I 1 1 I I I it) N M O r1 4 00 00 N O r4

-4 <n

- O O I'D in - m i N [`N Q l t ko M

r

t ' i

CO in ; CO lfl CO 0 4

cM 'tt' r-I CO 0" Ql N U) LO lo N to in in CO N M I I n N (r1

I N LÖ i it) 'ä' --w i

M te) oD -, O NN CO in v -T l 0 M M d' N 1 Ö V n

ý, + Y N N N

) 1 t N C ) in in ( )

, --4 N (0 d' in -, O "-i M M N C' N M N d' Chi

+ "- 1 a 1 1 1 c I 1 1 t 1 l 1 a -4 O O O O 0 C D O 0 O 0

O C a

O O

U W w cý x H a z o Or c ä, " °C)l

rz

22

N W 1ý

lJ

i;

MG

. -1

" Q at- 1D

.

"-i O

N M Cl

. -t V

N ri

ON .ý

C) 1

14 M

N co

N kD

1-1 M

N O

cN 00

CO O

N N Gi

(! ) in m N tN co l0 O

ý- 4 l0 in ul ct' M in 'n ý, O C ilk

M O

co

(N O

SST

CO '7'

O

i- N

l0

N 1ý9

N

tN O

. -1

%D L l!

co

m

l-

Gl CO

co

N M

M

CA N

co

N mot'

O

iý U)

-4

to CO

co

N in

. -i

.v in

Ln -i r-1 M N M N N r-I . -i N r-1 N N '4 N N c

p Q co u W W 0. x H 0 Z O 1-1 0 OI

CY) OI -zv O+

9

O --I

O mot' 1

114 N

1

N lt) I

N Ln

I

(N to

I

O d'

I

N in 1

O F

I

O C

I

N LO i

O -V

I

N to I

C) d' I

CD -7' I

O d' I

N it)

1 C) -4 O

M N N

Ln ý'

N ''

C) ý'

M m

N d'

. -a M O M

CO M

CO N

N M

N M . -1 M

v' M

M -ä

0) ON N I

r-1 N

I "t' N2. I

. -i ýN I

CO -W I

N C1

I

W d'

I

O M I

0) N

I N M I

f- N

W M I

-1 M I

Co M I

m M I

N cM

I 01 I- N

0 CV

N ; 7'

0) M

LD ýP

4 M

M d'

CO N

N N

M M

-r M

O M

CO N

N M

C) "V

co

1o CV I

Ol . -1

1-1 "7' I

CO M I

' U) w I

C) m I

N "4'

I

N N I

lD N

I

N M I

tD N I

CM M I

Ol N I

N N I 1

, -i M I

Ol M

I co 111 (V

co M

W M

1-4 14T

0) N

O "t'

-v7' N

to N

O M

; 7' N

01 N

CO N

L1 ) N

CO (N

lG M

lý d' N I

CO . -1 I

tN M

1

N M 1

O 1t; M I

N N 1

0) (')

1

M N I

el' N

1 0) N 1

M N 1

00 (V I

N N I

c' N I

tý N 1

Ln M 1 c-

'-1 N

1ý "4

U) M

r'1 M M

U) N

W M

C) N

N N

'D N

-4 N

Lfl N

t! ) N

N N

it) N

ri M

'O

O N 1

lW r-4 1

d' m

1 O m 1

1- M 1

d' N 1

lt1 (Y)

I

O1 . -i 1

r-1 N 1

Ul N l

c) N t:

ýM N 1

'I' N 1

-4 N I

dr N 1

O M

1 lO CO . -i

in r1

-I s C")

N N

M M

N N

ßl N

N -

C) -4

C ) N

Ol -4

. - 4 N

N N

C D N

N N

W N

ix7 to W

I- .4 I %0

'V' . -I

O M

1 IN

%0 CV

I M

N M 1

0)

i-1 N I

0

00 N

I "t'

l0 "-4 I -: r

CO -4 I t0

N ( I

ri

CO . -I I N

0 N 1

rl N I

CA -i I

s-i N I

U1 N

I NJ

8 . -1 N N N . -1 N . -I r-i N # -I N . -i

0 ri

N r-1

01 . -1

M N Q

st'

tn . -I

ý1i

-I r-I

N

I d'

N

I O

N CO

i Ln

CO

F' in

N

I 0)

s--i I

.4

' , -I I

V'

O ýI co

'D

I \D

1

co

1 lo

i 00

1

N II

d' C) . -i N N N r-1 rl -4 .i --4 r-i

M .q

S . -i

1@J ý4

kO r--1

WO 1

ý1

N . -I 1

N r-1 1

M N

1

Cº .i 1

IT N f

'a' r-1 1

l- r-1 I

O -4 1

M r-1

I

N . -i 1 "

M . -i I

N "-1 1

to . -I I

N '-4 I

U) r-I I

N r-1 I C') 0

'-1 u-C ri

O N

N . -i

0 ('1

M r-4

N . --i

01 N . -i

a -1

N r-1

U) to . -i

O ri

M e-I

. -i rý

N rn c0 I

c

. -i i It

0

. -4 1

110

. -4 1

(71

.ý I

N

u4 i O

r-1

. -1 1 CO

CO 1 N

. -1

-4 1 O

in

' . -i 1 -1

.4 I MI

s I oo

NV

s-1 I O

C)

r4 I tý

N r-4 r-4 ýq ý r-4 r-i r-4 r-1

"-+ N

CI

co I

O

Ln u--I I O

v .i I 0

m -I I

O

Ol I

O

N I

CD

l0 1

O

0) 1

O

m . -1 -k- O

O . -1 1 O

M 1

O

N i 1

O

5 I

O

Ol I

O

W I

O ý--4

a pa

a rý u q w cý x H a 2 o r-i a N of

C') Of

' Of

w

23

N N

M

cm C4

13

C4

O . -i

d' M !m

C) OO

'D

co Gl

lo

d` 1-9

co

Co N

t! 1

O "-1 O

lD 00

to

co -1 LO

[- l-

in

C) LO

O Ul

a' O

in

a' %. o

t11

Ln t

to

-'T r-

CO

CO in

r- r1

N Cl

Ln .i

co N

. 14 M

1-

I'D N

N CO

r- N

O in

N

N N

0 N

O -4

rý r-4

in O

010 r-i

01 ()

M N

a' a'

C1 r7

co in

0l -4

ý. o CC)

01 'A

in in

O N

M 1-

M N

l0 N

C N

Z

a Qs4

oa v w w o x +-ý a z o o a N

a M

a mot'

a w

O "-+

O st' 1

O M

.o N 1

N N

N to

1 U1 cr

N to I

.-4 d1

N in i

Ul Itr

C) W !

it) M

N Ln I

l1) lqN

0 zr I i

M

O W 1

a0 N

N in

I CO M

O -! '

1 d1 N

N tt) 1

LO I

M

O 'd'

I

M

O P

I

M

O w i

X11 M

N to

"I

O

ON 01 Cy I

N N

r-1 N

d' -tv I

N

O -W

co I

M

d' -V I

fV 'cr

eM M I

N ("1

d0 1*1 I

M 14:: p

O M I

010 N

N N I

l0 N

N M I

(V M

co N I

I- N

-4' M I

. -4 M

Q1 N I

CO N

M M I

O M

-N M I

M M

'-1 IýT I

co M

O'

GO

tC N

I U') N

O N I

dl -4

-4 Nr I

C3) <q)

N M I

In M

-4 IV I

CO M

-I M I

01 N

(V `w I

N CM

Ul CV

I f'ý1 CV

it) N

IV

N

-4 ß't1 I

O M

to (V

I 1n CV

c) M I

00 N

N N

I l0 N

ß) N I

N N

N M 1

O M

I- M

-Jr M

oO

N

d' N I

.i N

CO _4

CO M

I if) M

'd' M I

O lM

N M

I m M

00 N

I N N

l0 M

I N M

fV N I

O N

M N I

-l N

C1 N

1 in N

-.! ' (V 1

M N

N (V 1

M N

to N

1 ch N

LO N 1

M N

a' CV I

00 N

M m 1

a' N

1-

W O N I

CO r-1

1- -I I l0 1.4

'-' M I

N M

01 N I

N N

N M

ýI N

%D N 1d '

N

M II

CV

ON -I I N -1

O N

I dl r-1

to N I

C') N

N N I

. -i N

(V N

I O N

N N I

O N

N N

I -1 N

N fV I

'ä N

00 N I

In N

to

W u'1 iz

8

r. . -1 i

in -1

U() .i 1

et' . -1

I M

I Cl N

kD N

i M N

N ("1 I

d' N

tM fV

1 . -1 N

N (V

I N N

%D -4 I

"Ir r-1

0) . -i i

ýG r-1

fV N I

.i (V

O N

1 co "-1

(n . -4 I

l0 .. i

0) -i i

00 -4

O N I

. -i

f) N

1 N (V

q' (V

1 a' uI

ý

ýi

Vý d' ,4 I r-I

1 T-. 4

co

I N

N C-4

I N

In N I

ý

N

1 . r-1

N ý4 1

. ý-I

1 u-

ILO

u- 4

C-4 ýO

r-4

l4

I . -.

uC

I rN-1

[, 4

I rß-1

-4 [-

1 . Ln-4

'4 (-4 I

-4

u-C co

M .i "-1 1 0)

. -4 r-I

Ln N 1N

(14

Gl r-1 1N

OD . -i 1

CO rl 1

Ln r -q 1

'"1

. -I

I O

N . -1 I

u- i

1-

1lD u')

IM . -I

I O . -C

'd'

1 M -i

NN

I M . -.

CO

1 tp -t

to

1N -4

M

N OD i

l0

p V-4 I

01

r4 N

rLI r-1

to . -1 I

. -4

it) r-I I -1

u-I

tf1 . -i I N

. -C

O ri 1

N

a) I

N

O r4 1

C)

lfl . -I 1

C ri

N . -I 1

e-1 . -1

Ol 1

N

N r4 I

. -1

.i

N 1.1 IO

r-1

in rI I

N ri

1-4 . -1 11

C N

In a

Co 1

O

r-4 1

O

C%j

1 O

O

0 "u-4 I

O

w I

O

to 1

O

co I

O

M-I

1 Cl

-4 1

O

l0 1

O

C)

I O

C1 1

O

'u-1 1

O

co 1

O

,

kr-,

w z o 01

ww

24

0 H

ä ýr

a 90 -e J)

N si

Ö W

bi w Q C, F 14 4

" In IZN

ý. p C'

to c-4

to 01

ON Cl

to ct)

to ,

('M N c1

M l0

N to

Ol Ql

p Ct)

O b

O M

0'+ U)

f: i " " " " " " " " " " " " s " " " to Cl) N lD N to In Ci lf1 Q' N ý1' If) to Cpl 0

0 m 0 %x. ) N N (D O c-i c-+ c-4 O N W if) z N O c-4 M N 0 t) N CV (�Si rI M N Q1 N c-I Z

z N ci

to c-+

r+ M

to N

to N

to N

l N

r. tN c-I

m N

c-i N

01 c-I

co t-+

o N

lD CV

M N

O Q

Q q U W is+ C7 x H ca z 0 C7 C7 CJ Y O Q

O 0 d'

tO N

N If)

tV to

N [C)

0 d'

N In

0 v

(D -f'

N u-I

O Q

N C

O %J-

O !' 0 '

N to 0

I OD N

I 1-, N

4 d'

I d'

A d'

16 M

1 d'

I M

1 N

1 Ct)

(ý pý

1 M

U1 N

M M

U) M

Ct) d'

cI

t- C\'

O N

M d

O d '

N d'

to (I)

tO \1'

O M

00 (V

d+ c)

co N

U) M

N N

N CO

to M

N

cri N i 1b 00

4 4 W In

M N l0

Ö 4 MI

O

N d' M Ct) CO mot' CV N CO Cf) N CO Cl) CO

CO N

Q1 c-4

c'i d'

Ln M

N CO

M M

M d

N N

4' N

(V ct)

N N

ca M

to N

Cc N

m CC)

CO M

O NA (A cl)

CO

C,, lj CO of V

(N

C\11 CV cl, N (V N cot

A

CM (ý4

4 co

M (V to CC) M M CV N

N N CV N N (C) M

Ö N LA M

Ö"

N N Oi N

4

co

'

N

Ö

N CV

4 N

4

N ci N

4

N Oli N

OÖ N

ß1 r4

L- T-1

dt CC!

O N

O N

00 N

CO M

O N

M c'I

[C) N

(y) tV

M N

O CV

CO N

cc) N

r- N

1

, CD tU

t-i

1ý

c0

A

N IL N

A

N co CV

CIO c'+

r-, r1

A

N (V N

pl N

d%

c-i

N

N [! N

CO N

W tN c4

CO c-I

O rt

d' N

11) N

W N

N N

tN t-4

Up N

N N

c-I (V

0 ý

qý ý-1

N LO N

C\l N

w

N N o i N

y

tf1 c tý

Ca N

N

N N N N LA c-!

t e CV

ä1

c-4 lÄ ri

1 c-i

ck -H

.4 N

00 c-4 v7

Ei

d' V4

ci

CO c-I

V-1

N N

N

c-4 N

Tq

c4 tV

ei

cO! N

(v

c4 N

4

d' r4

t'1

d' ci

c-4

O N

c1

CO c-I

r4

ei' T-1

V-4

U) 4

H

a) c-I

ch lV

N

N

rf

VI

(C)

cl

Cis I

c-I 1

O cl

N

1 cI N

ýp c-i

1 N r1

r- c4

I C d

O N

! Gc1 c-ý

U)

1M -I

ci r4

t4

N d

c'+ c'

N ßc-4

t,! 4 c-i

LO ca

1 In c-I

O d

p> !

CO c-j

N c't

!N c-1

T-1 N

I cl

N c-I

I

N co

Ct

0) 0 N N

to ci 4

CO ti ý

CU ci A

N VI

CIO

ON 0 cl

lfl d A

d'

ý O

tD

N c'" Ö

l %-4

C

Lý 'ri 'A

C) J) N

t0 3

co 8

c-1 c 0,

cI c-

8

ýHo 8

-q: r I-

8 N 8

co 8

Ol

0

ý

8 cN-I 8

U-) 8

Q1 8

01 8

in

8 d' 3

x aO

.4 an U w rL. (D x ý-! ] z o ä ä ä 4

w w

25

E H

N Ir

z

Lfl

N N

w ý

I 0) I I n lD N N

CO O 111 O

N O

i i w n u r w ko u) O I -v In co cri (-i Ln a) G7

In - "7' O co C) CO m lD M ! -- Cr) l0 ßcr ul t1 N 1 0 0l w l0 co N N N r -1 M

. '

. z : n d' O 'G' r" 1 1ý0 r- ! kp N N O [ pý ýp cr "-4 -4 M N N N N ý-4 r-4 N N N . -4 . -i N N

0 N M

a W U W L C7 x ! -ý a X 'Z O O! (DI O{ a C)

O O N N N O N Cl O N O N O O O N C) cr N If) ºn to 'd' 1f) v tT LO 'c7' In c' C' '! ' IC) O 1-1 1 1 1 1 I I i I I I I I I I I 1 -1 O -4 Cr) C) -4 in t- O 0) Cr) N CO n N Q-0 in

N) N l! ' "Cr d' M "t' Cr) N M Cr) m N Cr) Cr) .; S+

41 O CV 61 O d' "0 O) co N ri ! - lD . -! U) , i' N N dt M '4' M " N N c") Cr) CI) cV Cr) Cr) v

of 1 I 1 1 "I. 1 I I 1 1 I I 1 C- 00 ý--4 'V' co M n 1n . -t O U) C) 0) 0) N d' M co 'f' N N (1) M M N N M

! -- a) C) Cl) n M N %D ll;! ' O 0 'T' N co -W CO N -1 -f' M Cr) Cr) mot' N N C) N M N N (Y) M CO I 1 1 1 I { I I 1 I I I 1 , I co co . -i N ON 't' N OO n Ol . -i In M U) N M M re) (') N N N N N N N Cr) (-)

Cr) CO n '

O r- q N 00 M . -I n W QC) O '! ' N d' N - r 1 el Cr) CO M N N N N N N N M M n 1 I I I I I i I I I I 1 1-1 N I.

' co CO O c! ' O Ol Ln in t. f)

1 N

1 O

I al

n N ri fß 1 N N Cr) Cr) N -1 N N N N (') N

ýcJ O l0 M n n d1 f'') M 00 cP v d' M , -1 al 0) 4 N . -1 M N N N M -4 r-1 N N N , -1 N cV N G] 8 %0 i

00 i I I

' I I 1 1 1 I I I I I I I -O üi

U) '-! in '! C O Q1 CO i- C) M O co O n V' Cl) f4 . -4 M CN N N N -4 -1 N N (N -I N N N

U) rý IV C) qv m r- 00 r- t. 0 N N n O\ Cr) - 4 r-1 N N N N '-i -1-1 ! ! -4 ' ! N N N Cl) Lt) I I I I I I I I I I I I

ý ý 11) M n . -1 O ý N L) In O O t! ) Ii) ! - d' ri

rI /-! N N N N (V . -1 '-! N N r-1 i-f r-4 N N

c1' N l0 O d1 d' ---1 w V' Q% 0) tP -tT lD M CD

- (N N YI Ci C C 11 P 1 1 I 1 I I 1 I I I I d' N r-1 sý Q1 l0 M n C'-) A CO C0 M u) . -1 U)

ri ri N ri .i CV -4 i4 r-i , -1 -4 r-4 N '-!

-+ c M co In cam: v) N '-, I N- r, cv m a .-4 N i r-1 (N r-+ . -i . -i r-1 "-1 r-( . -i (N "-! M ! ! i ! ! ! i i t ! t { { ' I m 0) O In N O C' . -! --i Lo O N N ! - N

N . -1 . -4 N -4 . -! . -1 . -i -4 r-1 . -i r-I . -1

a1 d0 -1 0) (4 O O er U) , -t -4 Co O) . -4 . -1 . -i 4 -4 ---t ri -1 -1 C1 r-I -I i-f . -4 cV I I 1 1 1 I 1 1 I 1 I 1 1 1 I N 40 iß rl Cr- l0 CO (n CO N -gr n O m C') 00

LO CD Ln -4 N CC) r- -4 ko ON 00 t i t 1 I t t 1 ! I 1 i i

ý-! Co O C) O O O O O O O O O O O O O

CO U W W C7 x H º- Z 0 -4 Oi

N 0

M O! i

cya O!

2

0

26

cn II

N

z

in Ulf

j Ql Cl

tf)

-4 N

fM

%D O

N

U)

[-

O N

L

Ol 01

:: M

-1 '! '

(5

LO to

to

N v

mot'

'o U)

Ul

. -i C'

NP

C' i

I-

l0 --

-'S'

Co tf

l. [1

Lt) C

eP

L CO

co N

c '!

i-

Ln M

M

fn 0

O

L) to

1

cV M

O

M O

t-

to C

to

61 iý

CO

CD 'I

l-

N ý1

N

61 N

r1

00 fl0

O

O 4.0

l4

to r!

N

O tý

r-

'r d'

-q

ý2',

.1 r-1 M N N N N -4 1-4 N N r-4 '--1 N N (V

a a CA U w w 0 x H a z o -4 ä ä ä 0

w w CD to r4 N N O N C) C) N O O O O O N

O 1

O N1

N I

. -1 CV

1n 1

d' d'

U) I

01 M

Ln 1

to M

v I

to M

In I

Co d'

M 1

N M

W 1

N N

LI) I

ýM M

-:: v I

lV M

mot' i

Ln M

V1 I

Co N

y I

IV M

-'ZY' I

N M

Ln I

O O '-4

0 O r O 1d) U1 N r-4 to m . -1 W N M lU 01 N CV 61' c') M M 'V' M N cI) M M N M M M

(71 1 %D N

1 01 . -1

1 N V'

I 'a' N'1

1 N f'r1

1 M d'

1 N CV

1 mot' fV

1 N M

I O M

1 O (")

I M N

I O M

I tf) m

1 Ln M

crn _

co Ul N I

V' N

co . -1

ri -' I

ON M

M M I

0) N

-4 (Y) 1

co N

c' M

I N M

N r

OI M

to N

N

M CA

1 N N

. -i cM I

Ol N

C) N

1 a N

Ql N r I

N

N N I

r-1 N

CA N

I a) N

.' (I)

Q' M I ý

f+')

Co

N M N I

r-4 N

N r1 I %b

CO c"') 1

U1 M

CO N

I L) N

IN N i

' N

r4 M

N M I

M M

I' N I

N N

"-i N I

O N

ao N

I U') N

N N I

to N

to N I

M N

c) N I

ON r-4

N N 1

tO N

cr1 M 1

. -1 M

O M 1

N N

N

tO

O N I

CO . -1

In . -1 I sT r1

d' M I

. -1 ()

C N I

N N

Cý7 CV I

O N

O M

OI N

C%' M I

N (N

-4 N 1

Ol r-1

ßl

1 N r-I

er N I

M N

LO N Jr N

N N I1

0 r-4

CO -i I N

-4

to N 1

c'-1 (V

c) M 100

N

to (V

(IV N

to

19 8 U) N I-

I

M 9-4 I

C -4

O M I

N

r-i N

I

v--4

Oh

. -I

IN N I

N

to N

I N

Co r-i I 4

to 1

I

LO r-i

N N

I C N

M N I

i N

CO . -1 I.

to v4

%D .i 1

rv-i

N N I

N N

1

N

-4 (V 1

N Ln

W

-4 I

cl, . -1

. -1 I

cý . -I

N

4r V'

(V

CO

t! 1 Ln . -I

ý-1

I -W . -1

N I

A N

C)

I I r-4

-

t A1

-4

rv-1 M

'-1

. v-1

l- -4

N 1

ý v--1 1

rN-1

. -w-1 1

-4

1

. ý-1

N

1 1 C-4

- I 1

/N-1

mot' -4-

U]

M r-I Ch

Q1

C6

N

., 4

-w ri N

M r-4 Ö

N CV Ö1

11) v-I

fV

N

r1i

N

. I1

tD ü1

CO

N

ý--I G1

N lD -I

N N I

O

. -I

. -i I

d1 M

N N 1

to

N 1

to

0 N I

C o rl

. -1 I c 0

C1 1

to

00 . -1 1

N . -1

.4

. -I I

00

O . -4 I

01

O . -I I (in

'cr r -I I

N

to -1 I

Ln

00 I

l0

C)

I a%

.1 .M I

ý--1

v--4 I

G10

CO I

ýN N

ý-I U) I

O

W 1

O'

r- . -1 I

O

tN 1

O

In I

O

rv-I I

O

N I

O

00 1

O

00 1

O rl 1

O . -1 I

O

Ln I

O

a) I

O

CD

r-f I O

ri

O1

M 1

O -4

a

a w w w 0 w H a z o ý--1 a

N a

M of

C' a

w w

27

H

M

. -1

Ný

r-i c) N N 01 r-1 "-a k. D N co -'r M en N O N N tf) ýY' Ln 00 [- -4 CO O) A N N Eý N

. . . . . lp C') N N

rn 'o

ä LO Ln in ": 4- CO :.

7 lD .C

ö

u1 ý r+ cn O N (14 N co r-f '( O M M Il- Iýv ON 00 m ko N C! ) lD CO N N M l- U) 4 mot' hfl CO l, N O z

t OO U;

4 1! )

4 to 00 00 N to .i m CO O LO C")

0

r-1 fh N N N N -4 . -I N N r-I -1 N N N

ö '-i N M G

W U W W 0 x H F-7 Z Z Ja of CI cv a G, W

O lC (V N N Cl N O O N O N C) O O N O C' N t! ) to to ct+ t. 0 rM Q' to [ -,: P ;r 'i` to C) 1-4 1 1 1 1 1 I I I I I I I I I I I . --4

01 N 'Y -I 1ý '. 0 CO cr CA 01 N S -A V N LO N N -' d' Kt' M d' M (N (') c"1 M M M () ýN

CO . -4 (l O l0 Lf1 N M CO 00 . --1 l0 C) Cl) l0 N N eM 1141 llw M d' M N M M M M M m

O1 I I I I I I I I I I I I I 0) N CO N w . '-4 ko if) O rl 00 N LO a1 . Cl d' d' M N rM co m N C') M M

l- O -q t- rl d' [0 O Ln -W Q) C) l- . -1 V' cu N N '4' (') d' C') 'd' M N (0 <'4 C') N M M ("l

co I 1 1 I 1 I I I 1 1 L I ! i ' CO lC C! :r [0 C7 O ýp M N lý r 00 M 0 N d CM C) fn d N N (0 N N N N M 4 ,

to ON 01 t0 er CV 01 t0 N r-4 lD %D CV l- N " N M M M c"'1 M N N C`') N N CV N CO C')

r- I I I 1 I 1 1 1 I 1 1 I I 1 I rý N l- 11-1 co C) ll1 (V -4 d1 ! f1 (0 r--4 ' CD 01 N M M N N M N N Cl N N N N M CV

. -4 CO ID O N CO -W 4 O CO "w N O (n O1 00 N -4 M M N N M N N N N N N N N N

%D I 1 I 1 I l 1 1 1 1 I l I1 I I 'o 0) (0 [0 M [0 d1 l 0) CO U) CV

-CO a l 4 1 N CO 0-1 M N N N N r-4 T-4 N N r4 r-4 N N N

8 00 N N LI) N t1) Co co N d' . -1 l- 00 O kO (N i r1 M N C4 (V N N N -4 r-1 N (V N

V] I V M 1 1 J M 0 1 8

ýD Ö1

C O c+ 7 1 1 t ) ) V ' N CO d.

CO '-1 ý4 N N N N N 1 N -4 -1 r-4 -4 (N1 -1 R

tO In a . -4 C1 N CV C' "M N 00 M t- M s .I p4 N N r. 4 C4 N r4 r'4 N ^I r-4 r-1 -4 N -4

d' 1 t ! ' t 1 " 1 I I I I I I I "a' [V lw d' O N Cl O1 N c , M . -1 5 r- 1 Ln O cT . -1 -4 N r-1 -4 .i . -1 . -1 r4 CV r-i -t -4 1-t lV 1-4

. -i C') C') co Ln co 'D rl N O [D O (n -M C) c' P4 r-1 N '-1 11-1 r-t .4 -4 p4

M 1 1 I i l ! 1 I 1 1 I S 1 1 1 1 M r-1 P4 l0 (V tf) N Q "-! tf) 0O N M --: f' N. - 4 N -4 . -1 . -i -4 . -t r-i . -i

O O ul f-i v' p-I O r' d' r-I N M co . -1 N . -i rl .4 .1 0% -4 N i -4 P4 [O

N 1 1 1 I I I 1 I I I 1 1 1 1 N [0 "-1 c r-1 co n ß 44 [0 G1 O 0 IV

r-I -1 -1 . -1 1 . 4

. tI as " -l " 44 r

° N N lo "-+ "-i Ln CO rn ON (0

ý I ý -ý a (b ( D 0 0 ý ý ý ý ý ý a a

P4 N M Q W U W W C7 x H 2 O OI Ol OI a

W W

28

Li

(1 Ný

H

N U) O c') fß"1 N '. D m a en . -4 N to rl CO -i M r- O C' .4 CC) -i O O IV w %D 1-I VV O

910 (fl l0 co O l0 O t0 ul lU L1 1'- 1'1 1j; Lt C Ü) ri r1 (I)

M N fM N U) Ln to CO -4 O N Lo N O u) -4 O 41 !n If) C N CO N N l0 to r- N lo Cý N

CO lfl r-4 Ln rý :r f- r- r- 4 O 06 0l O ci' M

. -i r4 M N N N N 4 . -4 N N ri .i N (N N

LY.

FC W U W W C7 w H º 4 O Or C9 Ol Or O

W W

O l0 N (V N C) N Cl O N C) N O O O N O V' N to U) U) r U) it 'd' in y U C d' V' U1 O H 1 1 1 I I I I I 1 1 I 1 I 1 1 "-4

r-1 N d' O a I C) l C . -4 O N . - 1 to O ' N l0 N M N . Cr ' 1' ;P M -V M M to m M M M M cr

O M 01 N --Zr 00 O ßl l0 O -V ßl M in -1 Co N V M . 40 M d' M N M M M N M M "V

( 1 I I 1 1 I I 1 1 1 I ON N N «t' e d' 0) l0 1* . -4 O 'V' 00 N cl' . V. M d' N N M M M Cr) M

"o O --4 CO M N cM Imo- Ul CO 0) O co 0) M r,, N N -W M qtN Co d' N N M N ( N N M M

CO I I I I I I I I I 1 I 1 I I CO I to 01 ßf1 Q1 ri O U1 d' . -4 N N w N N v N M M M e mot' N N M N N N N M M

d' 01 00 IY co O a% d' M O to W Lf) %0 r-4 M N .i M M M CM lt) N N M N N (V N M Co

N I I I I I I I I I 1 1 I I I I I N fV CO %D O V' CO d' '-4 O CO 4' M st' d' Ol CO N -I M M M N Co N N N N N N N N N

r1 N [n d1 to r- M O 0) 1- M N M en CO N N M N M N M N -4 N N N N N N N

%D J I 1 1 1 I 1 I 1 1 I 1 1 1 1 tD m N ý CO U1 0 co CO U) rl Ch Co rl to mot' r4 M N N N CO P-4 H N N . -1 N N (N N

S CO %D . -i in t- -' al N N ýP O 00 O in en Q

U) M N CV fV N -4 ; N N i i ('1 N N Cý i i N M

Ln to dl r.

f N

U) lfl N CO If) N CO M Co . -a '. 4 N N N N N H -4 N ""4 -1 r-4 -4 N r4

0)

MC' C' CO O 4 -4 N cP to -4 N mot' lO N N I- . -i i CV N N lV (V -4 . -1 N r"r -4 -4 N -4

q: ' I 1 1 1 1 I 1 ! I 1 I I I 1 d' t - th n 0) N ' V N M C O L tl (V

4r ' cr C O d+ .

r-1 -4 i CV -I . -4 . -1 . -1 r4 . -1 .1 . -1 "-i r-4 ri '-1 r-1

O N .0 I- l0 r- tf) r-4 fV N d' -1 M M N CCM ri . -4 N -4 . -1 -4 -4 r4 - -4 . -I -4 ri r-I r-1

M i s ! r f w r I ! t 1 r r s M 01 . -t N d' N t' "-a C) r-I ßf1 c'M 0) (N ri Lr) r-t

. -i CV . -i -4 -4 p-i -4 11 1-4 . -;

O r4 M -1 M O O Ln tV . -1 O `V C) CO r1 N rl r4 r-1 ' -4 00 rl . -1 . -i N r-1 '-1 . -1 . -I

N I I I I i I I I I I I 1 I N %D rn rn oo . -+ N N ýr O rn C6

.4 H r-4 -4

co C) 0 M r-1 tC CO 1-4 -4 N 1-4 \O l0 ON "i 1-1 M 61 CC)

-4 N

-' I 1 I I I I 1 I I I I I 1 I 1 H O O O O O O O O :) O O O O O O

Q', Cll U W W a 4 ,' Za O a a a

W

29

C

GL

aM

ý+ U

V) ax

U HO

NHWU

" ýi u'

d' c-I

N N

0 O

O ýD

Co lO

(') O

O1 M

M O

pQ ci

N CO

N lD

M to

CO to

Co in

M C% " o . " . " " " " " " " " " " " " " " to

" Ln M tO N Co In O to to to d N d' In uI 0) " L: cf UZ

0 d' to s, n d' 00 W O CO N tD 0 0 -i 2 Cý ci f1 O l0 CO M l'- N O N to lD N Cl N Z Q " " " " " " " " " " " " , " " " iil r

10 lI

d C1

- + c M

In N

to N

d' N

b N

tD c-ý

N c-4

CV N

O N

Ol OD c4

O N

C N

M N Z

o Ö

o ¢ m o w w o x H a Z Z O O ä 0 0 0

O w N N N O (11 O O (V O N O O O cV O C' N ul U) in d' to d FC tt) C Ln [1' c d u) O

Cl c_ Ul 4

'0' to 01 01 C) O

iN Cl

cV to N r-i

N (V d' C Cf4 M d' N lV M M M N f'-l el

CD N

O N C'

0,, M

M c}'

q

M d' d

Co N

Co N

n! M

01 N

0 M

00 N

vi M

in M

c+ cJ'

'0 Oi N

tN "i 14 cf1 if lfl ci CDI N 9

011 4

011 N c-4 -e M C1 m e fV N Cn (V M (V CV M M

U1 CO to dl M N 'V' Ln O N N W CD M Co cV c-I d' M c1 M d' N N cM N M N CV M M

O en Oi

1 M ýÖ ci 0'.

N v' OD tD W .4 ci

C, M M M M M N CV CV (V N N N M M

N N Co cm N O Co M IN to t- M 10 O M N c- 4 M M cn M m N N N N N N N M M

c1 N

1 tp %--1

u to

CO N (41

Ö, N

4 M

Ö N

Q N

to N N

� N

1 N

N N co, N

op N

m c-1

to c-4

[t' M

N N

O M

W N

M M

ol X

O% H

d' N

M N

N N

O N

N N

LN N

t N

cW p

zW

H t! 1

tO 1 N ci

CD vi

cý

d' s ! r4

1 c-4 M

o ('1 m

N

I u1 N

b' N N

N

1 Cl- N

lD N

N

N

. tn N

(V

1 0) N

N N N

N

I N c4

to %

c-4

I Co c"4

tN vi

ul c4

I N N

c- N 0

n1

I N N

Ö

N

011 c"I

I O N

al c-i

to' c'_

1 0' c'-4

CO c-4

N c"i

1 ci N

0 N

CC) ci

N

In N

f1 lV

N

M (V 0 N

£4

O

H Ui

c+1 t"'I

N s-1

S N

In N

c-I N

UN vi

N N

CO c't

N (N

N

N [t c-i

M c+ N

c"ý

d' c(+

7 s-I

0,1

pý c't

W

N ýr"t

L7 c" N

ci

ID c4

In c'1

N c-_

111 -4

N N

0r (V

cl Vi

uý vi

s-I c'4 c' 0) tN 01 W c-4 N r to vi d' d' CN d'

cr t Cl Q c-'

1 In

! c. 4'

t U

i (N

1 C

!, "'

1 : u)

t ý

! Cl

t ll

t (-"i

1 r

1 ci

CO O N

V' r l

M r'

U1 %-'

c'4 -I M

O c

CP Z

M vi aQ

N c'1

c, vi

l0 c'{

O ri

N Co

c+

1 M vi

N

c1

N

t-t tp

M

C"t ci c

lý ci vi

Ol C! N

to CN

O1

cq vi O1

0)

O1

r4 VA OI

iN

OI

N

O1

CO

OI

c4

O1 vi O1

U) I

O co-t

I O

(p i

O vi

I O

lD 1

O

ä 4 w v W w cý x H a z o öt ä ä oar o

30

Ný

IEIINI

01 N to CO 01 N N N IRV 'a' c" r-4 00 00 N to Q M c") 01 co to "qr r- O l0 l0 M tD ti' O l c) Q

In ff; t! ) lfl 00 to Ol tf) d' ýf' d' N d' tf1 tIl 01 U)

Cl co t- lfl M r- N m Cl O -;! ' co u O Cl U) N CD O Q Cl N 0) O CO H M Qý d' (14 O 0)

1- M CA .

M In C LO co [- 1 ri ri CO O C' V' ri . -I N N (V N N H N N N . -i N N N

m u w w 0 w ý+ a z o Of ä Of öl w w

O l9 N N N O N O O N co N O O O N O -V N in U) U) -V UI -' C' U) to d' -d' wN UI O H 1 I 1 1 1 1 1 1 1 I i I I 1

O 0 d' r-I i! 1 U) N Ol Cl *"i O N a1 O tO fV M N d' w -N P1 -' N N M P M N M M eN

01 al M 0 d' <N to co co O Cl l0 co C1 If) . -1 N +--1 'V' ICN M d' N N M N M N (V (n "W

O1 I I I I I I I I I I I I I I I a1 Iý CO r-4 to O M N to U) C1 dq l0 CO r ß1 N r4: M w m d' N N N C7 N N M M

to Qn N ri to d' CO co M U) 1- C1) co W N O M M lh d' N N N N M N (V M M

a0 N V-1 Nr I I I 1 1 I I I I I I I co ' N Ln 44 M N N c, Ln

( 1 [ M M M N C13 N N N M M N M

O ,M O to M N to to ob N to m mot' N to tO M M M M N N N cV N N N N M

en I I j I I I I I I I I I I 1'- ý r-4 r N f-4 fn N M

.I lÄ M

N M N M

N N N N N fV N . -i M

to 01 1. 01 O O N M Ul O (V to a1 O Q' N N N N N N CV N N N N N N N

to N . -1 M I I I I I I I I I I I I I to I ( C ON N N N r N N

C D Cl N N N

( 4 H -1 . N ý-1 N

M N er lf1 I- CO O '-1 r-1 ca O M UI IZ tý r - M GC N CV N N ý-1 4 C

1 N 1

N . -1 (V t'0 N [s] t11 , r N 1 I I I I I I 1 I 1 I to

8 1 1 1 "i ri N "-i l0 1

l0 C1

r"1 01 �'i

a0 .I

N .1

Cl r-1

N N

rl CV 8 Cl)

to r-1

Cl .i

to N

N N N CV - U)

O O . --I O UI u1 CO co N to 00 H

. N

O CCCJJJ]]]

IV N . -i

H 0-1

!n cV

N 1

N I

N 1

N 1 1 1 1 1

N . 1: d'

A M Go O 4 , r W in ß O1 '-1 . -4 .4 r. 4 . -1 r-I '4 r -1 r -I 4

r-1 N

1- N co to N i+ý1 tO N M. Cr U) CO V4 N C) N 4 -4 -1 r-1 rt ri ri r-4 ri r-1 ri ri r-1

t"1 +-i ý--4 N i i 1 I { 1 I 1 t I I 1 I 1 1 I If) Ifl tp M ' -4 N mot' UI O N M N ý-1 O CO 01 ri , -1 -4 ri '1 -4 . --t . -i . -i ri . -1 r4 ri

G9 d' .0 U) N O r-1 M w . -1 (N l0 O C1 I, - r-1 u-1 . -1 .i r-4 r-1 r-1 . -1 -i ON r1

N N M ý Ö N N N to I C1

N ic 1 N Cl Cl N r-1 . -I r-4 1-4 r4

ý V-4 C%j . -1 M C5 C14 \O It) -1 . r-4 CO f-4 tO CO CO r-1 tp f CO r-4 to

-' in O

I 1 O

1 O

I O

1 O

I O

1 O

1 O

I O I

O 1 i "-+ Cl O o 0

P4

M CJ w w 0 x H a z 0 ä ä ä 'S

31

II

Ca In Ny Rzj

-i P-) Pq to C5 t- C> i 0 Ln w C) Ln C) CO O CO lO l0 d' lý t t) N 61 -4 w N '' Q N N M M M Lo N M M M co M M M cI, U

Ol O M C) N O Cl ýO M cf+ O N to M c O to to C' -4

I [- M O in O N M co 1 O M 01 O O L1 -4 ri -4 r-1 --I r-4 ri -4 rl ri ri . -1 Iq

a ca v W r%4 0 x H a z O ä ä Ol ä

a

w

O M to lQ l0 O to O O to O t O O P to C) N "-1 N N fV N N N N N N N N N 414 N O I I I I I 1 I I I I I I I I I I --i N N M 01 in 00 M 01 to m co N N 00 N '

. --4 -1 N 1-4 N ri N . -1 rl r-4 r-1 N '-4 r-I N

tO '-1 N Co . ýr N N co u) co N -4 l0 N lO (') 1-4 r-I N r-1 (N -1 N e-i -i '-1 ý4 N -4 ! -1 r-4 N

ON I I I I I i I I I I I al 4 to N w -4 N tO O Ul to N N -I N --1 N . -'I . -4 N -i . -1 N

Ul O O LO +--4 LO O N d' to Ln 01 to 10 .4 - -4 N -i N . -1 N r-I r-1 . -4 -4 1-1 ri r4 1-4 N 1 I I 1 I I a0 Vi co ýN CO fM LO QO C) m4m O

r-4 1-1 r-1

('') N M C) M N tO N IV N N -'r m ßl r-1 01 -4 .1 N r1 r- rl r-I . -1 r-4 r-1 r-4 r-I . -1 i-4 N I 1 I t t 1 i I i N 1ý Ol l! 1 ul "-i M L n (Y) co

[Ia

8 rd -4 ri 1q r-1 -4 1-1 --i -1 f9

Cl) it) (N co O M 'c! ' i--1 N N (-

( r-q r-4 0 c 1 I I

I I I to C34 -4 lfl -4 N m N M c 1-1 . -i tQo

. -I M O U) O . M r-4 O O rl . -1 r-4 . - . -i -4 CO r-4 -4 '- ON

L1 1 I 1 1 1 1 1 1 1 1 Ln O C'j a% C1 O t- > O C; m M O -4

. -1 N O ON lD -1 CO CO 01 - to 0) d1 . -4 CN 00 co (

I

O a ý 0 CO ') -4 1- 110 1-0 L" N N- 61 to to N

M d 0 o Ö M N

ä1 lÖ

0) ' ) t l CO in to

Ö1

'D C 00 Ln 00 to m 00 M LO iO 1, v to LO 00 N

V1' N I N

1 d'

-J N

T .0 r N

r- t'n to e r--4 N N W hfl M M c'M to Ö Ö I I I I I Ö Ö Ö Ö l

O Ö Ö Ö

C ) Ö

a' LO U W W C7 `; 'i H I-a X Z O +-{ O!

N Of

M a

IM a

32

I--

OM 0% N

J M

WZ Z W

M

I- u

01 NmN

W JUQ

co %0 "u'-

lY co Li'% I- CO O% N V O Co N Co - CO O OD 0 OD 0 %, O - - 1- M %, o O CO - v c - O

Cn N N M ýt ýt M ttl N M M M M M M N d u;

v oo N ao %0 - v N %z Co Co orn r- - %o r Z M O N N 0 N 0 M Ln %0 1'1 ml N M Co M 1 Q Q W - N pý t') - N M I- - - N CD - - W

fY O O

U Q Q] U W LL CD Z - -ý Z 0 d d O' d U Q Q LL. Li

O M %0 %0 %0 0 %0 O O %0 0 '. O O O O co N - N N N N N N N (N N N (N (N N N

O 1 I t 1 I I 1 i 1 I i I 1 1 I 1 O - N N v O' M OD -t Oý U\ 0 Co O u OD 0% M -

- - N - N - N - - N - N - - - N

- N OD N N1 ct Oý 1ý Oh 00 N - N - N N - - - - - N

O% 1 1 1 ' Iý I ' I\ C\ 0 0 t OD e Co - N - N - N N - - - - N

O 0% cO O N M co O N - - - N - - - N

CO u1 I 1 1 1 I I I- N 1 ul r- 1 '. 0 I 1 CO - Oý M 00 u1 CO - - % - - N - tcl pý

-; t Co N t` 1N CO - Ul V - Ul 00

N 1 O' 1 I 1 cf I 1 1 1 ý- i I I - I N M tý - %0 - to

- u1 -

O -

ct -

- f! 1 -

O - -

- N - - - ö

O N

ýD O ul M M M M M %0 c/)

Z ýp N CO 1 1 1 1 I 1 1 1 I Oý 1 1 t ýD Z

- to 0' ýt N N - N N K) N N t1l W

cn N

9-1 rj

u1 1- Co 1 - - 1 N 1 tc _ M 0 '. 0 O O 0 - M

O N O O (N O O N _ _ _ ýf 1 ýD 1

1 1 1 1 CT 1 %0 1 t I.; Y

0% - '0 0% O\ N -- ON 00 O\ -

0 Ul Ul 00 a7 Z V Co CO Co ü1

M 00 1 t I 1 I 1 0 f 1. I I t N- CO h fit' O' 'ý7 ýp N U\ - M 1- N N- C- CO

1N Ln d 0% '. O \o I- N 1 n CO t% I %0 1 1 N cc %0 1 M 1 1N 1 N

%0 4 M Co U1 U'% cc

Ul N 1N N M Ul N N - n n N ý7 %0 u

0 0 0 0 0 0 0 0 0 O

Ö 0 0

< M U W LL- (D M Z O O" O" G7 3

LL- W

33

Fr

H E-+

0

d'

O Lf -Ir .O 0 c m co O N (n c- 0 N m 0 " :n Ol O N CD ri u) LN cl m c4 ci t- Ln (; ) O "

" N v-I tý' d d' M L) N M M Cl Ct' N Cl N t! 1 "

c- to O q% c d co (V 'V' Ln N lU O1% co 2 r+ O %-1 c+ M %-I N LI to v- co 40 V' '0 Q1 c-A 2

.ý k7 " "

N "

d "

O- "

" "

N "

c-1 "

M "

C-- "

N "

-, 0.

N "

N "

" " 4 "

V

Ei

Q ¢ m U w w C) x a : Fl z O cl C\j Cr c C Q w w

0 M k9 tp %D 0 W O O 1 O to O O O W 0 N N-4 N (V N N N (V N N N N t4 N N N O T-1 t 1 ! 1 l 1

` i I 1

O1 "I

M 01 t ý ri Tq N

c N

ýp T-4

t J N

ßl cI ca

O N c-4 fV ri ri

Cp T-1 N

tC rI N c-l N N W LO O\ co O N a) N Cl c- N c-4 N cc- N c'4 ; c-I i C\1 c-i i r-4

(TI 11,

t 'p a1

A N N N

N c cd ri ["i ri N

in r4

O T-i

O N

to ri

CO to c-I

O C\j

[N `.,

N c i

t0 r-I

01 c-I c+

lD vi

l0 rl

TA c

03 ; 1

-i

l

0

I

I ýpI -I

i N

rq ci c c t

c-I ' W O W

c-I ý

%-ý N O U')

T-A Ln IC-4

ON r-

N N A I II

In Ol 4 4

lI N

r4 c1 ri cI c-I cl N ri c-i

(V %D ci d' v Cli Cl) d' M c+

ci' a0

M cl

th Tq

(0

tD Ok1

1 N

A j A

.4 C\l A A 1

in ýD

vi r-ý W

z ci c-1 N

Cl e-I C4

c-1 N ci

c-I '

Cl c-4 t-

cl rl

N vi

N ci tN

T-4 c-I

N ci

d c-I

(rj E-4

A N N O1 c11 O'.

Ö

vlý c- 8

tä N

H

CI) cý c-I ri c-'1 c-1 c1 t-1 vI IC-4 Vi

ON tD %-I t0 CO 0

0) C\j

U) O O

ý ý O'H

'

Ill N (1 ts ca 01 Ik Ok c-ý

CO to O c-t d' l0 00 Ln

O v 11 CO 0) Co. d' W CO co

M t r t-I I

cý M

N d' N cn d' t- r) 01 !V l0 l0 c"') U) N lC N 1

l"

to N N

1

In Cn to N N N ca d' to Cl) N d' CO dý

ö ö of ö i

ol ;

O d d Cl) O O d ö of 1

O . 7.4

v d Co O w W U' X H º7 X Z O ör

or ör C ü

w w

34

)

M

W

"'l

2 H

C,

H

4

w

4 w

0 In N

Ln

w 4

d' CO

CO c-+

rl O

N c-{

O N

N [-

N N

co at

CO : fl

CO M

CO T-.

O ON

N 4.0

114: ' c'

co N

(V 0 "

C; " N N d' '' c' N U) N C4 C4 M M N M N [t " tI) Ui

V4' CO d' at M O CO M CO C N N at CV LI) pl 2 Cr Gý Ul N CO O U) M N r- in CO l0 CO ON N z

O O to M OD O ý4 cI M N -1 N (N CN "I CI , 14 ts %-4 IH c-I %H cI cl CI c-I c-I r4 t"1 c

O O a Ü

w S W U w ; _., 0 x H 4 z 0 cr O '

o a Ü

w

o cn i) CO kD o tO o o tO o 9 10 0 C tD 0 N cl CV N N N N N N N N N N N N (V O N-4 I

CLi (N a 4 8

all 0) A I

al A

CV 4

all oll 4 V,

r c-I N N c-1 c-i N cl c'i 0 N t-4 ci c-1 N

N ci

c-l c4

O N

0) cl

CO cl

N N

0 c1

In t-4 cI

00 cI

c'ý N

M r

CO rI

CO t-i

C) N

I I kD ci

CO 0 (V

M c-4 c-l

Oi ca

Cv T-1

N T-1

(V N

to cI

O cI

01 cl

CO Cl

N c4

N c-I

01 c-I

co c-I

N IH

t0 V-4

N r4

Op c-I

c-I %-I

tO c-I

N c<

c- N

OD .4 Is r-, C-1 1-11 L11) 1ý c

ý

t-1 c-1 IH t1 c4 cl c-I N

M cl 01

Cp c-'ý

cm c-I

U) c- I

to c-I

to t0 vi

O c- I

ý4' cl

CO tl

to c-I

O %-i

U) c-I

U) vi

0) c-I

tÖ 1- J

M LÄ In In J

.4 <` N

cl c-I C'l ti 0 c-+ c-I r"I

p

c\j CO

!!

ý

' T-4

c0-+

ö' T-1

T4

cc-I 'rI: r -4

c

cl co

r"

M r+

c\11 r-I

d' c-i

cl

d' cI CO

M r+

r+

d' c-+

V4

tD IH

1 wa O En

p en 0 M c-I Cl) c"ý N N cI N Cl) 6

EI U)

fl c-I tý T-4

1 (V c-I

CO 1

N

T-4 1

Ql

Y+ 1

(V cl

c-I 1

Ol

c+ 1

cV c-I

tO c-1 c-I 1

Cl

c-I !

cA c-I

C- c-I 1

O t-I

c-I T-4 1

cI c-I

H Ui

ON U) rl to

u; CO

ý T, OD

kb cl U7

4 c-I

011 T-4 -

a1 l0

ý Ol

CIO c4

O

0)

co v. 01 V' tO -H to C: > T-4 M co at co N

0 cl co

N to en cc

ll N1 N d,

cc l-

Cl) N

ON [i

N to In

N! 1 to d' to CO r

N

%-4 d'

a N

ý is

d a eI

a to

ö c-l

ý lO

a O

a d'

a 14)

ý ul

ö M

ý to

ý r-

8 M

a ý Wo 4

G4 4 GO U 67 Cr U' . 7C H º-7 F. 2; O Cr

cy cm m cr a 2 Q Gy

35

Q Ln r

N

l0 -i

N

! - ý

M

Lf1 ['ý

m

r- C0

m

O . -I

m

M O

tf)

co lý

fV

U1 '-t

M

"I er

M

61 CO

N

L N 4

O 01

N

C1 N

M

fV lfl

N

M U1

C! 2 U)

)

H

N m

tV

tn

o -r . -4 m r- Ln c) co rn r O rn 'n r a0 1ý O C1 1ý 00 CO I--i O 1C1 9 r-I O M c

Ln C' 00 m M O M 00 '--1 N N [-- N M C )

. --I -4 - 1-4 ri

M . 4v a'i M U W (X4 (9 x' H ýl x ' O ä ä ä ä

oI

rn

Co 1

11

10 1

M

N

O (h 'D kp %0 O to c) O l0 O 1.0 OO C) l0 N . --I NNNNNNNNNNNNNN

M, -14

M M, a1

N N' -i N 00 mM (\

Ü') I

r1 .iN . -1 r-I N "-i . -1 N r-I N r-i . --I r-1 N

%D O. -I IN co (71 Na C1 00 ON 00 00 . --I . -I -I CV ri -$ ri N r1 r-4 -i -i N r-4 . -4 -4 N

r-I CD C) -I-I 010

Ir ON

lqM C1 --Zr' Lt) 00 '. O MN lG co . -1 r- [-- O

-1 a% rl r--1 'i . -1 -i P--4 . -i r-I ri .qN 111I11III1II

00 00 MmNNt! 1 N l0 %D Ol . -1 r-1 -4r.. ý. -4 r-1 r-1 . -i rl r4

M 1, NN lD Lf1 Lf) '-I -, r If, ýo c, t. [) In 00

MÖMJ Cl lÖ

ri rl ri r-I rl -4 r-1

N to O . -I Lfl N d' 'd' mot' Lfl '--I CO -1 "1 . -1

rl 1 ä1 Öu

u-I tý

M VI'

ri r4

O qv e' 0 C') r-I MNNMM

r-i to ý--t CO ON u-I rl ri 00 . -4 u-1 ri l0 -1 r-1 r-I IN

NNIM O1 NNON u-I i-4 N 1-4 - P-4 r4 r-1 rji r-I

. -I N rl r-1 OO '-i rl Ch U) rl l0 N r-1 co r-I ' d1 ri ; Ln i . -I +

'D C-4 tÖ to co mM

C) C> OOO co M r-I LI) to Lf) . -i d' 00 P--I N 'cr CO -I 00

lo (N CO C') -tv C% Co MM l0 00 L r) MN 01

t, N

Ln rl 00

O

m

00

N

Ln

qw

M

N

to r', Iý N f9 l0 N t0 "-' "" \O "4, ri "O t- -r 1 1 1 1 1 1 1 I 1 1 1 1 1 1 1

O O O O O O O C D O O O O O O O

a' 6ý U W W C9 º't'. 1"'ý ý"a ý+ 'ýZ. a O a a a w

36

H H

Ü a

M Co ý-1

z

H ýä

M tO CO if'1 CO -4 pn CO O O O N Q " C h N l0 to 61 CO Ol 1-I ßl ýN Ql CO

N M to m In N M M M M M M M 4 Q

to

al CO LO N -'M - O 00 O Lr) N to (0 CO N 00 (Y ßl r-1 r-1 ,+ N CO -i to ! . to

cf' m N ro M [- M . -t 4

N- ý4 N M

rý v w w x H a z o ä Cl Cl öl w

Co rn lo to to O to C) C) lD O tO O O O l0 O N r-I N N N N N N N N N N (N N N N O ý-1 1 1 I l 1 1 t I i1 I ! 1 I I ' co N v o l d' 1 Q N --T c to o 4 - C A O tO O M

-l . -i N N N r-1 N r-i r-i N -4 N '. -! ý--1 N

l- lM r1 PM 00 M DO LI) O CO o 00 0) N N N N - Cpl -4 N . -I 4 r-! r4 '4 N

CA I I I 1 1 I l I I I I I a1 l0 -1 00 -1 CO c ;l Ol N C tý CO . --I N -1 N -4 -4 r4 r-1 -4 -4 N

LA -1 0 N O N- N N N co to h CO ýp N O -1 N -4 N -I . -1 -4 -4 -I -4 r-4 -1 N

O I I 1 I t 1 1 I 1 1 00 dl to ON O r1 In l0 CV to c r-1 r-q r-I N -4 r-i '-1 ri -4 ri

dý O Co CO to 01 to CD N LO r-I tf1 to OO . --I i-1 r-1 r-1 ri r-i . -"I r-f r-1 r-4 r-4 r-1 ' -I tý I I I I I I 1 1 1 I I I I I tl- M c t- N to tf1 I- Ol to CO lqv O r-

4 r-1 r-1 r-I ri . -1 r-4 . --I r-1 '-1 r-1 -4

N to 1 to -ý: f' to to In N CO M ý: N to . -1 cc .1 - . -"I . --1 . -i . -d 00 r-1 .j _4 61 -i rl -1 lO I I I I I I I I I I to

ß 6 M M ý CO N M ß -1 r r -1 r 1 r t r I r

8 U) r-i d' N N N d' CO r-I O r-1 N tit

r-q ON ri '--I r-I r-1 1, - 1-4 h r-I -i r--I "Z Ln I I I I I 1 I I 1 1 I 1 1 1 *ý E

Cl) O CO Co O ý-I I

O M to N C) 01 tO 0 -4 r-i E { r-1 . -4 .-4 -4 r--I -I .-I C/)

(V O (N ' O O m l0 '--1 1- 01 1-1 CN r-! to .1 Q1 CO to a- r-I r-4 a' 1 I I I I I I I I I I mot' Co to cc (7, 00 r-q d' ON 00 o) al

r-1 O I'- t! 1 r-1 u1 N CO h -1 c'M 00 00 N 'V N 00 00 C'`) I I I I I I I I I I M O to t, tO N to ýp 11

to V' C; ) M ºn lD d1 N w N 1 1

L) N M CO N

113' CO 00 1.4 c! ' ul N N F-I U) Ln V' r-I CO t ' '-1 I 1 I I I I I I I I I I

o V O O 0 0 O O O O O O O

J> I O

I O

I O

I O . -1

V tý U W W C7 x H i-o z z o Cy O I ä Of

2

U w

37

C

0 w rn tn aN

4Ü N U)

wa ä

H°w ýn

In N N c-4 M d in l0 0 O CN M N M c+ co " C) aý O Q) (l u') N as a, N N 0 0 to C'4 t0 " CV T-4 ' M to M Lý CV 1V [*) M 1; [h M C1 c' "

a)

z N M as c- O O

07 a)

O T-

O lo

c-I M

N c-1

M M

h Q)

N N T 0

U'i 0 2

" " " " " " " " " " " " " " " " N '$ al M c-I N M N N ci c N c N

cý c-{ cý c-{ c-{ C4 C'4 c-I cý c-4 c-1

O

Q CO U w C:. O x H Z z 0 Cl ä 0 ä

Q

c: a w

0 O N

() c-4

(-D fV

to CV

0 (V

O fV

O N

O N

O (V

0D N

O N

to CV

O (V

O (V

O CV

to N 0

TA r1 X-4

+1 1

N

1 c-I

) fV

i oil r-1 (11

j -H "c4 o1 [U 0:

1) c-i r+ CV LI n c4

I a, ri

I rn c-I CV

0

lA c"ý

m N

kp ý

M (V

Ua c4

M N

co c'i

CU c'{

a1 vi

S c' {

O N c-I

CO 10

CO c"i

N N

i -I N V j M ý ý c CV N ( la al L C CV N s-i N V-4 c'l cI ri c'" c-I

U) t'l

0 74

N Lr) ci

%-4 N

W c'l

vi (V

E- vi

T-4 c"f c-ý N-4 ri %-I vi N

co of V-4

d' %-I

al c-i

a1 -I

to c-i c-i

1 c-i

Oi (V

co

d' cl

cý Ol c-1

r M

.1

CO t. n C'1

co t"'l

"1 to Tl

O vi

I u1 c-l

d' c'l

Ln s-4

.1

O c-i

Ln cl

4 Ln c-I

4 C) cI

1 r- T-1 T-1 T-4 cl ri T-4 cl

W 08

O

N c"1 CO

W ri

<ý

O c"I

In Vi

tý-I

.1 c'i

t-I

d' c-l

ci

t1) r4i

l

co '' cl

ci

en c-{

ri

M V--,

c-I

(71 M c4

T-4

M H

l0 c-4

O

19 Ei

H ü)

ut

T4 cl 3

cl

[N d' cl A

cý

O.,

CIO

m c-I

'A cl

N c-i

c s-l Ö

c-i

M t"l N

1

(V

c,, Vi

c-{ c-1 Ö

ci

l -H Ö

\-

N 1

s-I T-1 Ö

c-I

N H

cý l

d' H N

ci

U UZ

Et Ui

Oº tý c. j ci

t

N1

lA

cl

CIO Ö\

a%

l- cN-4

ci cý

U) ri O)

CIO

O1 LO O'

ap

0

A -H Ö

T-4

10

0o Ln 4

Ln N w LO 0 cl C" aO N CD ci' :s rý o,

Cl) I M N a; in Lö

c ö cý

N S

l0 d' W M I U) ( 00 N 0 l0 u') M la to a) N

NI fV U') LA N LO N

TA u1

I O

CU I

O

W I

O

N I

O

CU I

O

41 T4 1 O

EN 1

O t O1

d' 1

O

d' 1 O

c1' 8 IO u1

1O d' I

O In 1

O

u a ao v w w 0 x H a z z o cr c cr Cl H u

38

u)

m

L) H H

O a

4 i Ö

W

W rý

[w

N ON

N U ul N4

v

8

[N 0 L O tN c-I N d' 00 V1 01 c-i l0 U) to c-i s N O CC 5% U cV ci tN Ln M TA O'. 0 M 01 to " "

4. CV M M C; Cn U) N [h (*) !f M M f'') CV vv "

L: M rn ci O N 0 00 t0 0'. N U') N co d' G 2 CO N N ("J : +) CU C') O ()' c-4 Ol cc) c-i N l0 : t' Z

O UD M O', 1V 4

c-{ cM N ri CV C) c'! c'i C' . e. -{ c-I c+ c-+ c1 c-4 V-4 c-I r c-I c-+ c-t

ix P-i

w Q U) U w w cD x H a z O

0 O

N a

(fl of

d of

C4

0 w

O O CV

!

c-I

M c-I

I

c-i

UD N

1

(V

"0 (V

1 r

i0 N

I

N

O CV

I

1D CV N

CV

O (V

0'. rý

O cV

t'i

l0 (V

0k ci

O fV

ß? ri

i0 CV

(V

O "

u1 c-1

O (V

CO c1

O (V

00 c-t

(D CV

N

0

10 ci

I

0 c-4

c+ N

1 ý

S c-I

I.

H N I

O

cli

l- c-I

c-! N

I O

co rq

Ul t-- 4

00 ci

I

c

N ca

O N

I

T-4

d` c-i

t- ri

t, ri

N CV

T-1 01 a'. c'i

pp c-+

u) c'"1 4

lý

rn ri I

Tli

ko ri m c-A

t- ri

r- VA

U) r-i

M c-4

CV c'i

ko c-1

Ul ci

17 0 U)

9 0

rn r!

CV c-I

c-a

t( %-4

c-I

to 0

c-+ CV

0'. c-I

CO

M c-! CO

N -H

IH

C') cI

T-4

N ci

U) ri

ID cI

t! ) c-

ri c

,I

d' c-!

r-I

In T-4

T-4

UJ ri

cI c"!

d' IH

d' ca

W 'H

W S O

(V ci

ci ca

S tf) c't 4

c'! ci

O c+

c-i

cM c+

N

ct

C'i

ci

c{ Ol

"

CV t

M c--!

! 11 c-1

c+

Ol

CO

r") c-t N

cý

Cfl

N

S s -t

Lin c+ W

Ö

U

H

Ui

N

0

I

c-I CI OI

cl

1

t*.

N

c-I

rI

Ö

cI

N

ri

M

r"t

ct

I

c4

T-1

c-1

ý H (J)

ON U1 c'i ID ON

CO t0'+

O; co

10,

V-4 c-t l0

In 0'.

co O ri

O; O c-!

(31 l0 C,

ý O Vq

0) Cl c-t

ol

1 r. 11

t 1 C. - zz,

i cl

3 M

t C-

t .

I i

V^ !

OD

ID (+1 OD M 11) U) C°1 CO N U) l0 tD (f) N 5 i N

t) In In N

ti) N U) N M ü) N N c't c In U ' !

O i

O i

O 1

O 1

O 1

O I

O I

O I

O 1

O 1

O 1

O I

O I

O I

O I

O

4N U 4 to U 14 54 Ch x H a . -. Z O

0 Öt Cl Off'

O U

39

x

cý 1.0

RE,

N

z

ul

O M q o0 t- -w l0 mot' CT N Co tf) T O , t- f- N [- Ln w N p CO 00 Ln m N

" " N N M M IV M 'ci' N M M M M N M N zi'

"O -, zr N Co l0 LI) M i m "o r) CO /D O q. N O M N /D ri O O N do -4 L) . 4zf' V co a0

\O M C1 1- N r4 M O6 O N M CO 14 ý4 4

0 U ýa w cý x H a z o ä ä Of öc

w

O m 'o lp lO O l0 O O l4 O l0 O O O l0 O N -4 N N N N N N N N N N fN N CV N O 1-1 1 1 I I I I 1 I I 1 I I 1 I I I "--4 o .i . -i Ol m al ri 00 lJ OD CO N hfl 01 ý M

r'4 r-i tV -1 (V rl N '4 -1 . -i r-1 N r-1 _4 N

0 O Co N 00 O I- 1- N -1 co I, N r-i N -4 CV r-1 {V v1 ''-1 . -1 N i-1 ri r)

aý 1 1 I 1 I 1 I 1 1 1 M (7N %o c) -4 a% qzzp .

N r-1 N -4

ul co L) "- ß1 tO Co l0 CI) Ln to 01 N 1- to ri -1 01 -4 . -1 -4 ri - --I ri r-1 r-1 r-I ri r-I r-I N

I CO Co dý [I-

Q' N lÖ to

Ö

v-I r-I ri rf r-1 rl r4 N

M t0 M %D LI) to Ln N V' LO N . -1 to to al 1-1 Co - fq v-4 1-4 '-I r-i ri 1--I 1-4 - r-1 . --1 r-1 rl

tý i i i I I I i i I 1 1 1 N in N Q' v CI) 1i' tO O 1. Z14 di N

r4 r-i r-1 r-i 1-1 ri -4 1-1 r--I r-1

N V' ý-1 c'') d' tM --: r O N M in m m tD ri IN r-I r-i '-1 e-I l r-' '-i '-i r-1 '-i d1 vI v-i r-II

to 1 I I I I I I I I I tO O N m 1.4 dl . -i N (V Ul . -i r-i -4 r- 1-4 ri rI r-1 rl

r1 M rl N O M O N (I) v1 r-1 cf' r-I tp -I m 4 . -i r-I ri CO vi v-i ri 00 r-4 r-1 r-I

LO I 1 1 1 I 1 I I I 1 in N co O m N C- r-1 N l- O M

jz] rl r-I . -I ri 1-4 r-1 r1 ri [ý] IZ

i O

In -4 ° ý ° 1 O

in -4 . N a1 co -i r k. 0 in -1 . lc 01 1-4 d' I I $ I 1 C 1' I 1 I "'

++ tt ON CO tO 00 O CO 01 ý

CD C) 00 d' tO N 00 ßf1 -4-4 t1) N ßl 01 Lfl N Co r-i M M o Jr d 6 i I M ßl

t to o C p p

N ý M Ö N

f' N N f 1 l

co to to l 1 N

ko r-I l0 N N Ln N N -4 -ýQm Lf1 tf1 co d' lo t. p 1--1 1 I 1 I I 1 I 1 1 I 1 I I I I I -I

0 O O C D O 0 O O O O O O C D O O

a a o Q 4 w 0 w w 0 x H a Z z o öi ä of ä

w

40

)

_)

r- cr)

6

ýI ýý

ui N II z N

ßl in o lc' In (- M N ß1 l0 Co Co d' co Co CO N CO m N Co Ol 00 N lD N Co N O Ol in r-1

M . -1 M M N In N (4 Co N cY M N Co lcM U:

M O lfl d' /O N lo N l0 N LO N N 00 ßl Co ! - l0 00 N [ý cM r1 CO u0 O lm r, 1 l0 d' lD t! 1 Z

O N Co O t! 1 r1 O Co O ßl M 00 00 m in

4 W U W W C9 x H F7 2 O 0 OI O+ Of W

W

O Co \p lD lO O ' O O lO 0 %D O O O l0 O N -4 tV N N N N N N N N N N N N N O

I I I 1 I I I 1 1 1 1 1 I 1 1 1 H t, CV N O d' CO : t- to a\ to N "o in l- dr -1 r1 N CV N . '1 CV r"1 r-1 r-I r-1 N r-I . -1 r -I N

l0 s4 r-1 4G M N Co l0 CO iÖ 4 In d' '. 0 Co rl , -1 tV N r-4 (V N . -I r4 r-1 cV

I I I 1 1 I C1 O N 'p N in Co "o O mot' Co N N - "-I fN -4 -1 N r-4 -4 N

In a 01 %D N If) r-4 :' N If) -i' o Co N Ln -1 -4 . -I r-1 '-4 N -I N r-I r-4 ri . -I r-I .4 rl '-I N

co 1 I r I N d Ö O CO 1 . CO l C O p r-1 r--1 N ri r-I r-I ri 14 . -I CV

N Co O d' [N M r-4 Co N N -4 rl Co d1 rt 01 r1 -4 N r-I r-4 rl r-4 r4 ri r-I -1 1-1 -1 . -1

N 1 1 I 1 1 I I 1 I 1 1 N M N CO Co N %D O O N CO r-I rl . -4 r-I r. 4 r-1 . -1 r. l rý r-i rl

(V l0 . -4 N N Co N O . -1 r-1 In r-I [N r-1 CO '-1 r4 r-1 r-I r-I . -1 r-I r1 p. 4 r-1 Ol Ol '-1 r-1

(x] tD 1 1 1 1 1 1 1 1 1 t 1 P4 r+ ýr %D r-+ H Olt O O 'r Cl to Q . -I . -1 ri r-i -4 '-1 r-1 "I l1

m CUJi 8 M ° ° M °

N -1 . °

ý4 r-4 00 M a1 " -4 00 CO Oll r-q I I 1 N a d' ä

0 0 -Jr r-i

l r-1 rl rl U)

r-1 Co O 01 l0 r -I co rl . -i N 0) 1'- CO CO . -1

d' 1 1 1 1 I I 1 I I 1 I 1 1 O N N 01 l0 tD [ý r 01 l0 W l0 N

N in 01 lC "-F CO to CO in W I'D CO in N in -q Co I - M I n I M

\Ö (V

to I N I N (

in d' d' 0'. l0 N '. 0 'N <M in .o d' M d' CO N

d' I CO n I N

w i N Co t-

R) r) in Co t- ul r-i Ln co M d" N .4 Ö t Ö Cl)

Ö Ö Ö Ö Ö p p ' p

N Ö

tC) Ö

FO U W W C7 x H ý7 z m ä a ce)

w w

41

O a4 Ln r Cv z

qM 1

NO CKII 1-4

9i H

" ,^

" U)

! c

" 7

!

Lý

ci

Co C

" ('()

Co Oý

" M

in d

" c"

Co GO

" CV

O

" LO

Co O

e M,

CO

" N

Cl Lý

" M

Cl if)

" CV

M Co

" d'

CT C7'

" N

' c1

" M

N Cpl

i 14

in V

f (ý

V

<;

X

N C'

O c-+

ý' H C')

rt

[ in

i! 1 Lo

.' c-

Co N

N c-ý

0 WO

ci c-ý

V' lD

c1 ci

cH O

0;

0 co

(31

O u)

O s

in to

N ca

c-{ N

CO

co C')

C)

Co Ql

O c+

0 Co

' ci

o E-+ o w

Co t,, w e, LD z H a Z , a, 0 0 CI

N CY

M Or

ýS Cr

p

Ü w

O r'

O N

1 co rA

Cl) cI

1 di ri

Co N

1

N N

Co N

1

J c-+

Co N

N

0 N

CIO ci

CO N

ý,; N

O N

oll T-I

O (V

M T-i

Co (N

C) H

O N

Co H

Co N N

N

0 N

ci

0 tV

ui c-I

0 N

ap T-i

Co N 4

CU

O

C)

N c"H

I CO ci

H H

ci N

I O N

or) ci

I LD c-i

Cß'1 fV

I N N

N c-{

N N

1 ri N

co T -I

I 1ý %-I

(V ci

N ci

1 Co T-i

in c-i

ci N

I 0 N

I! ) t-I 4

e-i

NT T-i

IN cH

i tD H

M N

i N N

ON

00

tf) T-1

O T-i

rn ci

co H

In c-i

VI' ci

c+ N

Cil c-I

t0 T-i

O cm

Ol cI

CD

if ci

V" H

Ln c-

H

4' ri

0) VA

CO ci

C') H

H

C') ca

(V 0 H

U) a-4

ci N

N

co

c3'

11, Irf

N M

N

c-I N ri

in q

t) H

d' O M

N

N

M

N

1-4

[N

CH

H ci

Ö

H

N \-4

p1

I

T-4

CO

p

N H N

CO c-H

I in H

H c-4

I O I

Co 7-4

I to H

d' ci

I C') ci

'' H I

CV c-H

M H

I N c-+

0)

I 0)

c-4 H

I O cl

c-H H

if) c-H

I cý) c+

O tý

I Ol

01 CV t--1

t vi

I CD r+

CO W Ö

Z

EWi V)

H I

c-I

cN r+ c+i c-H

0)

CO

c

A

cI

C\j

cI H r-, Oi

H r C)) H H Ö

rI

CO N

9

%

OI vi

H N H

ý}

6 in

H

U]

0 Co N c-i

ci H

N

CO

N H

rL c-i

r'I c-I Ö

c{

00( I

O H Ö:

CO co [!

Ql IO

c

0)

o

t-

ýj

W Ol I

N c-H

I

T-i 0

R

N LP, 0 H

c

in 0 ci ) lt)

ý"

C) : '1 \P CN 5

COQ

t1)

c

N

10

C)

c, II

CM (D %T OD Cl)

I N cl

N M t1ý C-

N

ýý C" 1 o

Cýl I

O r-

1 o

CIJ 1

0 ' ö

C

0 ö 1 0

1 0

1 0

I ol

I 1

o cV

O (y) ö

to

0

oHý

Q

w Q m u W W t7 x H a z o

ä a C\j en

0 ä Q

ý,

42

rn c)

w

H

0 H

äo

z

L, Ea m H

H ofu

wä ý3

Ir) II? 0 C' a) d O O Q to v4 lD c+7 N co " c4 a? O N U) to Ol 10 l0 l0 N- d ! - M l0 l0 " " M c- U) N N M (V d' N M M c}'

f

S O c'+ Ci 7-1 M t; M (T CV to l0 C) lD QA G*' lQ V' N N [S N -i CO L- t-I h In Q\ N lD d' `ý.

1 " " "

' " " "

C .

(v 4" "

N " " r "

h " s

4 "

.4 I Z

O t

h C c-I

0) M c-+ cý c{

O cl

O ci

(N ci

O) ri cI

K

Ü Ü

o M U w , c9 H a Z z O OJ O a a u w w

0 C k0 tD t. O Q 0 0 1.0 0 tD 0 O O to O N ci N N N N N N N N N (V N N N CV O

1 1 I 1 1 (N CIO 14 pp!

1 1 I 1 I N

ý -i V'+ c-{ N -{ -I i C\j H N N c c c c c

lfl c-I 01 M C tN M N h c-{ M d CO M c i N rA (V c-i N H ri r'' r N ci c-I c-4 N

N I N 0 in Ö N lD

N 1

cI 'N N T -i cI T-4 c-4 N c-i

: 17 O O In c-4 Ný c-4 In c-4 tf 01 N M 10 C \J c4 c'4 N cq N ci N cI C"4 0 c-I v-I c-1 VA t-1 fV

(0 4 ' 4 l I 0 4 L, 1-11 N j co c-4 , O c-4 c-4 , o T -i rq N ci ýI c-I

M a) M a) In Ol d' 0 M M lD O U) 00

h 1 ý

1 ý

1 I 1 N

I u1

i O

4 1 P

h t c

c cý

N cI 0+ c c -cl ri r1

CV lD 0 t0 In M 7-1 N d' c'7 l0 r4 OD Ti ca H H rý r-I 0) r -i c-4 0 O\ ON ri rq

O I 4 m N co Ö c4 () 00

I Ln

(4 c-1 to ca c-1 cI c4 c-4 v-I

ý O

bi L Ul

0 c N c 0 N h t4 - Q) cc-I c-I -i c h 0', c -4 h co ri -i H E

N 1 I N l V V4 N 0 co 1 [ - p C E-+

(! ] c-f v-I t-1 rq c-I -(-4 En

V4 0) l0 r-4 co T-4 r4 Q) c'4 0 h m l0 l0 at c"{

1 I l I i k o 0 0 0 c (; , l C t aQ

t'i CI

0 h In M t0 0D ri O 0) In 1-0 CO CO If) h O-

C I I I I 1 I 1{ ý^" ýfl u) ý" C t l SI

In q; r ao d' h O) co M I) lo Lo qr l0 lo N t

' ý

i

N h 1

' N

i N

d [ a N

M M to N 0 CO r"4 t. O (V M in M N c-+ to to 1 8 1 1 1 1 1 1 1 1 1 1 1 1 1 1 O O O O O O O O O O O O O O O

ýo Ö Eý U Q as u w w C7 x H a o

(Y ci cf Z5 u ý w w

43

O ýi'

W

H

z U H

a

w

w

m N

z

Ln

0 F9

4

cý N to Ln M 0 O7 to N co cc N N O N C' " C% co 0 in O 0 C (n tO N L') in (; -' CV m to " " N --i Q' M

4 M Lt) CV CV C^, N N M M d' "

Erl

O\ O O 0 c- O 0-1 C- N 0 In to N N [N lO c c- c-i V' N to N l0 Ol N ci c-+ O N c- r4 s- Z

" O

" N

" Cn

" C-

" N

" M

" cý

" c-ý

" N

" O

" ra

" M

" 0

" 0;

" CV

" ýt

EA v ¢ ca 0 W w (D x H a Z z O cJ C 0 C w w

O M to lO Wp O to O O tO O tO O O O tO 0 N t-4 N N N N CV CV N N N N CV N N N 0

ý

TA N

V-4 N rl c-{ -I i N

-1 i N c T r

to O [-i a) O OA m CO N co lO si in i) N C \j ci c4 N ci (\I c-I N s-i ci ci 7-1 (\I ci c-4 c-i N

A1 to ci N 12)

4 A C- C'i T-1

" N c4 %-1 c-1

d' -

CO d' -4

o% I

0 CV

to -'A

r4 I

in -i

to 0 C'ý) -4

tV i

lO 1 ca

co c c c % c N N V c c N

I

TA vi TA c+ ca 'I cr- i

N CV tO N Co d' 0 d' (') [N O s-i In C) cý al s-I ci ca s-i c '4 ci t-4 ci C- c' ri t-I t, s-i

'

'A A 1ý cl

c-4 VA co

to c+

O c-+

d' c-i

to %-i

to c-1

M c-4 01

c-H VA

N 4 to

%-I Q- 0' C c-I

tO c4

94 4 A 4

C"i CD O OD A

A 1

r4 ri . a Q

-I ý° N m V4 C)

C\1 -+ t 0 %N '1 c- ON IH

VI C\j vi tN CO

ý ý+

E-4 C

N dl cý Ö I', [lý cý H

c-i ci si vq c-I s-4 Vý

C) to V4 N T4 r4 '00 cI W OD M 0

l0 tO 0

c-i Ö

It oll "I

IL oll elk Ln o A ''

to in - it) Co O

In CO l() to co W In ti co c-, M, ß , A

u'1 tN CN M

O t

in d' N d' tD C; N cV t-- Cß'1 N N M to to N A I A 1

9 Ln N o N

i t

d' N Ln cm ýN l0 si tD CV M to d' m ci ýt M `i

8 ö ö ö ö ö ö ö ö ö 0 ö ö ö ö ö `-,

4 cn v w w 0 X H a X z o ä c ä w w

44

z oa +

ýN

'Vol m Eä

H ý°

U) r-i O lD O N in N lO t. f) N c in N o CO O r- O llý C1 Lf 10 L! M L1 rý V' C1 N 1D

; M N -, T M M N to N N M N .4 N N m d' U

N M r-4 CO in If' M IN l0 Cl) co m lw N M CO CD N to r4 C". 11.1 111 Co N in N r: O tO M m ri M N N rý M r-I N l- 00 N N

I

4 W v w w C7 x H a z o a1 No M a s

w w O M l0 ko %D CD lfl O O C) O kD O O O l0 N rI N (N N N N N N N N N N N N N O U I I I I I 1 1 1 1 4 I I 1 I 1 1 1 -4

. --i 11 . -1 (V) co Al O II) co CO 00 M ' ý t! 1 ß 1 M r-1 r-1 N r-I N -1 N r-4 /-1 . -1 . -1 N '-I r-i r-4 N

%D co N N Cl CO qzr N M l- Il- N M d' 00 N r-I -4 N -4 -4 -1 N 1 . -4 N 1-1 -1 1-1 N

C1 1 1 I 1 I 1 1 1 I 1 1 I o" to -4 in r N l0 l0 O' N M 1-1 N /-1 . -4 N r -i u r1 . -i f -i . -i N

O eM l0 N N l0 4 in in 00 r-I N N O ; C1 N . -1 .4 N 4 i r-1 r-1 . --1 . -i . -I N

CO ' i

co M 00 cM C1 in d' in l0 CO

lV fý N to lo co -ýr O cß'1 cN S O '-t if 1' r- -4 co r4 1.4 r-i 1.4 e-4 rl r -i -4 r-1 -i i-i

I ' Ü N l -4 1 -W n lÖ

61 lo

. -. 1-1 1--1 r-1 r-I r-1 1-4 rl

r-1 u1 O M tl' d' f'M I C+') to O -t' in rl S . -1 .i ri rý r-4 ý-4 O\ . -4 . -4 . -1 CO . -I r1 r-1

V Ö N Y M ä1 lV f ( )

Ö1

M ý M N .4 N N ° %D ' -1 00 -I . . -4 r4 -4 00 01 r -I t- CO -1 -4

1 I I S I , 0 C) CA N -f in CV CD C) v Cl

00 r-I -4 0 0 ö

l0 CO 1. N r-1 10 1 N ON

1 6

1 v I-cr C kD ö ý

N M ° in tN , -4 "o M in to o'. N in u) . co Ct) 1 1 1 1 f 1 I 1 I 1 M l0 01 l0 O to 00 in ct' O tD

N N C in I 01 "Cr CO in v :r M I I

06 I I

cal 1 ;

N kO -P

M r-1 %0 M ' l*l en %D N V S N M C11 M

-ý 1 1 to

ý, b a ý O (b Q ý ý ý ý ý c C:: C) H N M Vý

Q

a m V W w cý x H a z o a of a

45

l0 f- NMN d' OO -e M 01 Ln Q' M r-1 Ol 1 00 ßl ON -q OONmm It) rV co N0

M r-i MýýM l0 MNMN d' MMMý (fý

)

).

N

ä aý

En N

MN d' N co 0'% ýP ýN O r- Om -q to NM NNQ: r-1 L. r) N OO N llý Ql NMNN l0

O 00 d' OMMMN t- . -i ON al co . -i m

p ww cý xazoäää öi

O M tO kp w O l0 QD O W O l0 O O O W O N r-1 N N N N N N N (N fV N CV N N N C H I I I I I 1' I I I I I I I I I I N N m -4 C' C; ) LO O1 Lfl r4 N N CO l0 ON M

i --i N N N . -4 (N -1 .1 N . -i N -4 . -I -1 N

l0 . -1 CV O M 00 d' co d' C) lD r-I N tf1 CO N r-I r-I N N CV r-I N -1 r-I N r-4 N . -I -1 -1 N n 1 I

L co N N CO Ö If N . -1 ri (14 r-4 r-I N r-4 . -"i 1-1 .

d' -I [N "-1 N M N . -4 I-- In Ql cN M lo '-4 rl N r-I CV ri N r-4 r-1 '-4 '-1 . -i , -i r"i -I N

Co I I 1 1 1 1 I i I 1 I CO ON 01 ri W t0 N N ß r1 r-4 r-I CV ri r-1 r-1 r-I r-I ri

M O co M co W O Ln O u M W M r-1 111 CO r-I . -i r-I . --4 .i r4 N r-4 r-I r-1 -I rl -4 1-1 N N 1 1 -

W 0 0 V ' Ln . i a h

. -1

. -1 N -4 In to N d' M N dT O et' l0 r-I O1 "i -1 r-I r-I r-i -4 C) . -I i-1 e-4 '-1 01 r-1 -1

Lo i I 1 I I 1 I I I 1 I I 1 ýO \O 0 d' d' (") CO N . -1 M p1 M dv Cs] -"I . -I . -1 '4 r-I r-4 r-I r-I r-1 r-I rl (ý ]

p Ln m M m N -i O N N M r4 CO i ßl ; . --1 i . -i N 1-4 ý--I .i 00 Co '-i r-4 0

M .a r-1 O N I I

ce u1

U) "-4 r4 r-i . --i -f

cV O N O 1- i co -4 -4 j C1 (n 01 I- Wo d1 "-4 1w

Co I

1 CO 1

01 1

C6 O 00 q6 i

W i

lf1 I

QO I

C1

O C CO -4 r- r- n `r r- 00

c { 1 F ! I { I C%. Li', m tf) a' ýD 41ö klp N

L t'1 CO Kr (i CO d' CO M UP) W U) -v M U) kD N I N cr Q' Lf1 w M r

dý N V '

r

M M N M mot' to N l0 N V U M M '-1 m M - , I I I I I I I I I I I I I I 1

0 O O O O O O b O O O O U O O O

a a oa u w w 0 x H a '4 z 0 ä ä ä

w w

46

a aN

C7 pU

H týsa

Ü

ý `a wa wwýý

H ei ý,, i7

tD Cl N N in q l- co m fei i

" z t0 in in Oý co co N tD N o- in cc r-i q t- [may " " e " " " " $ " " " " " " " " ý;

" M TA M c") c`, ' N U) N (N M tV (\J M M 4 " U)

rn c-+ r q to O\ an N -31 0 in d' to CO O M Z in to Co O V O r' O vO ý' N [r in 0 to c-I 2

o (y] G Lý d' 'c Cý1 N N N O O N q co c-I c' fu c-l TA c-I c-4 TA c-4 c-4 vi c-ý cl Z

U Ü T-4 C"li CY) &4 Ü

G ý7 U W 44 io x H a Z Z O O r CY Cl C7 U

O in tD t0 tD O t0 O O cO O to O O O to O N ca N CV N (V (V (V N N N N N N N CV 0 V-1 1 I I 1 1 1 N ! I i f I I +l ý

c-I N

1-4 N V-4 (V c-I cl t"l N c-I [-1 TA CV

tN c-I N d' co M CO N CO N c"i M d' q cal r4 N cI N c+ N r 711 c-+ ci N c-I l c-+ N

ON I I 1 I 1 1 1 i 1 I I I C) O tO .N (V [` N to 0 M fl ci N N N t-i ti r- N c-1 cI N

to 0 0 u) cI N cI t cl to ºn a- N N l0 O IH c-I c- vi N t-1 (V c-{ ci %-I c-1 ca c-4 -H ri N

CO l 0 Ö uD LI)

Q' [N

k CO U ci c-I N N cl TA c1 c-I

O ri

C q i 01 l0 ch d O d' M l0 c-I c-I to co c-1 Ol ri Cl cI t1 c %-4 VA c-i c-I c-{ c-1 1-4

-, N ý A I

Oi o 0 Ö

rý t

T-I r ca

L c-I r-4 T-4

t c'4 c-I ci cI

N %D rN W d' in CO N N "j' C') (N- ca CO c-I c-i Tq cl H H CO ca ci c-1 0) (3) ca ri

ä 1 Ö J c+l

N 1-11 -11

I I W c t 1

W W

7 Ö

C, 0 +4' IZI, M N r+ O O N Li) ci N c-i Ol ci c cl c N 1-1 rl -N co co rq c-i

o

r Lfj

in I

CV N O

Oi c- N I 1 Lr)

z a T4 c-I cl V-4 c-1 c-I H

N c-1 ri 0 ý C() l0 cI [`N H 0 ON c' to CO ON tý to 01 H

lq; t N 0 8 N CO C) to c cl ri

c1 0 l0 L') c+ to C- N" to C) u7 cc r- CC) u-) v tý O)

Cl; I I I I ! I I 1 I C, i G 1n Ca C' '. 7. 'IT iý -- Qu c- r+

in V' ) 0 d ' IN co r`') 0D i d' to to [f' M in W N J

' t '

+ f

lI N

NI Cii ý N

d co t ý lp in j

M M IN N to to cI W ri (I u: d i N N Ln 1 1 1 I 1 I 1 I I I i I I 1 c-

O O O 4 O O O O O p O O O O O O

U 4 W U W W Cý x H "4 Z O Cr ci Y u W W

47

14, ll; r w

4

0 H H

L1. O a

a4

w w a 4 w

0 to m U

z

U m a4!

U

U)

T, C- [S 0 0 15 ON CO N U1 Gl CO ci co CO CO

" 0 00 rn CO "V CO to N Co M In L co o M " " ^'3 ¶ -4 C7 CO N U') (N Cv CO (N J} N (N co S! '

ý L7

N N N (0 r -A 0 { c M d O) d CO 1N c-I pp CD N M c-( cý N to L(1 0-1 0 N O CD Z

" " " " " " " 'i

" " " " " " " '

" is G3 U r CO Ol d' CV 'r cc) O1 O CV co CO r :T

p O Ff

Q C U W [, 0 X H -1 } z 0 O" Or 0, C! U u ý4

O cM C to %D 0 tO O O (9 C CO O O O U7 C) N -j N N N N N N (V CV N N N N N N 0

1 1 I i I I 1 1 I S1 1 1 I i l I N r4 N ON M O) CO N CO N tf) L( ) l C O rý r4 N c-( N c-1 N c4 T-A T -i r-i N 0 c-I " -4 N

tO O co N CD N lfl CO lo u') N ';; V N CO ci c'I N cI N r4 N ri c-4 ri ri N c'+ c'( c1 N

1 + i 1 ON p tip C- c (f 0 (' N

N c c-{ N N T-1 N

to 01 tf) r4 to O to N V Ci ) N CO CO c'I c-I c--I c-( N c-i N rq rI Tq ri T-+ i-{ c-( c-I N

co 4 1 Ö ) 1 ci c1

1 { N to

Ö CO

c( O cI ri

0 0

C c9 cl ri cR

c 11 1, % N

CO N M pp In N d' 0 N CO N O ri d' 0 ý-! r1 ýH c-l ri cl c-4 %-I c-I ri ri c-( 1 1

i I' 1 M1 c-I{ Ili to

I I O CO

1 N

r, 1 c--

N r1

N c-I

Q cl

c' ci c-4 c-I ci r4 VA c-1 c-I

N U) ri to m M N 0 'H to N CD ci C0 c e'1 tI c d cl O) -. - c-4 t'! 01 Q, ri ci

' 4 1 1 I 1

°0 1 to

c` -+ d ° cl c T 0 0

0 c C\j © 0 h i5 N -i a rn r 4 cc-I m ON ° c N O i c c-1 zz

4 N CD' °° (N 11 °° °

(N Q

0) tp i

N ;

ýO i ; O, ID N O) H

to t0 a1 C\j

11 (X ;

d co - 0 tp o 1.0 CO cl

co O o I i d In CO

co

CO d' OD d' N N N tN M d' lD to M M to N N ý 1

CO CA In N

l l0

'4' N %D N u) tD r4 tf1 N M Ln (N (11 N d' 11

c-4 1 1 1 1 1 I 1 1 1 I I 1 1 1 1 O (A O O O O O O O O O O O O O O

r-I N c Q CQ U W W 0 x H a 2 Z 0 d U C7

48

H

in in

N

z

W

CA Q

Ln W d'

REI

W ! - 0) Ql N ! - W -i ! - ('M 00 M r1 W 00 m O : , 10 to 1+'1 Ol '" N 10 M l0 O1 M M

Q ý

M -4 M co d' N Lf) (N N M N V N N C*) d+ Uj V1

N N N N N l0 L -1 O O -I 0) --1 '. + to !Z -, v m in N N -v r-i l0 U) N in Ol 00 -4 j

O w N cl M N O r4 O Cl O M N CO O L

O rH N M Vý

ýC !ý U N t: 4 0 x H ý-1 2 O O! OI O! Cl

O M w W lo O O O l0 O l0 O O O l0 O N -1 N N N N N N N N N N N N N N O I I I I I I I! ' I I I I 1 I I I 1 I . -1

N O O Ol M 0% N N mot' N N M ýf' If1 CO m -4 r-! N -4 N V-4 N -1 r-1 N r-1 rl N

%D ON CO N co . --1 W M lp W N M qM N N "--1 . -i . -i (N ri -"1 N . --1 -1 r-I N

I) Ö O U1 th to b N lÖ r-i N N r1 r-i i N -4 -i

U-) CO -W Q'. N Ol d' N ct- 'M 0) -4 M Lo 'i r-1 0) p-4 -4 r! -4 r-q '-I pI r-I r-1 r4 '-1 r-I ? -i N

CO I I I I 1 1 I 1 I 1 t CO N f") CO l0 lC 1-1 fM CO N O

. -1 . -1 r-1 . -1 . -1 r-I i-1 r-I . -1 -1 N

m t0 N N tf) If) co O N M N O -I ,: r 01 4 CO r-1 r-1 r-1 r-i ri ri -1 ri -1 r-I 1-1 r-1 -4 r-1

[ D M N Ö J M I N L 1 --I % \ , C 1-1 rI r-4 .i r-I r-I 1-4 '-4

N d' O It) d' N N O r-i to O N r- r-4 N '--1 1-1 . -4 . -1 . -1 r-I 0) r-1 . -1 . -1 00 . -4 r1 rl I I I I I 0) I I lD

- r-I . -1 r-I r-1 r-I

S 1 ý

ý r4 Q -4 8 ° ,D ý ý (3) c C, N ý ° ý r 0

N 1 b 1 I M l 0 0 o 0 0 0

Z r-I 1-1 '"I r1 r-1

C) 1 1-4 N 0'. if) N 1 -4 -4 tN 0) I'- 1'- -4 -1 ID N G1 -i '

b Ö 1 Iýv I

b w 1 lÖ

ÖO 1 4

r- I , -1 1-4

N eN O'. i) 01 01 it) 00 ID ID ! - 00 u') LI) N M 1 t 1 ! ! 1 ! 1 ! M

l'J IL !" C" 1_ L') LI ! ̀_ l0 CO

It M N d' ID co M l0 'V' et+ ID l0 -w cM Lo N N e N

10 "0 N

L) N N Ü)

Ct) N ilo

N U) N d' t i It) M M t') 'i' M N M lt I b I b

ý

c, O 1

0 1

0 l

O O O I I

OI b

04 as v W w 0 x H a z 0 ä ä ä ä

49

6 H

H H}ý

t1ä t'1]

x

ýv N

N

L1) .i

c kc -, r co N in N tN . -e ýo CO c) O %0 "o

I

Q . -1 N %0 d' -4 "T dv r-i O l0 O 00 ,: 1, 1,0 M 6

In M l0 l0 CO ul O ý1 In l0 a` N V t1) t! 1 N U)

OD 9

co 10

(N CO

N CC)

N O

Ol OD

11) "-A

it) -4

61 N

C! ' M

m W

- M

M r

Q0 6l

('l O

r-i r- C

ri Wit' Iý O "-a . -1 M ý* N ri p c; 1ý [ ý7 rq -4 N N M N N N -I N N N N M

0

a3 U W G4 C7 x H 7 Z O Ot U c Ot

O O 'IT

lQ N

N in

N If)

N in

O moW

N IC)

O mot'

0 mot'

N to

O q'

N EI)

O V

O d'

O d.

N in Cl

1 1 1 I I I i I 1 1 I 1 i I I - N N

O mal'

dl Cl)

N N M

in C

in Co

OD N

iO M

O M

H O M

O M

N M

lf1 W

1 -4 N CO w r4 Vw d' N to 0 O Ol Ol .4 Vw M N V M Ct' M C M N M N -, v N N M

O\ 1 Ol

1 Cl

1 CO

I . -i

i V'

1 N

I '-i

I U)

1 H

1 co

1 I,

1 W

I N

1 Ol

I m a\

tV N M M - cr M N M N M (N N N W

01 N

N I

N

0l 1-4 I OD H

N M I

to M

O M I

r N

M -W I

O mot'

O M I

CO N

H C 1

In M

O M I

a1 N

d N I

M N

c. M 1

N N

N N 1

lfl N

"o M I

Vý M

in N 1 M

N

W N I

t[) N

CO N !

[- N

N V'

Ol M

tN l0 (N I

in N

tN -4

d' O

I N M

l0 YV

I cP N

01 M I

l0 M

N N I

In N

d' M I

dl N

00 N

I N N

N N

I 0 N

%D N

I :r

N

in N I

N

M M {

'-I M

N N I

O N

-. 4N N I

M N

l0 N

I M N

co M I

[C) M

Iý

[ý7 IYi

8 N

OD C'1

'p "1

Ul -4

- M

01 N

M N

-4 CV

to M N

M

dr N M

N

co N M

N

l0 N

U) N

61 H

010 H

OD N N

(N

M N N

N

O O

N (V

Ol -4

co 1

-i

N N

0 r-i

N N

O N

N M N

M 8

tr, U)

0 N N 14

CO N

CD N

-1 M

N N

N CV

d N

N -4

. -1 N

. -i N

i. p (. 4

N . -4

is -4

01 -1

'-4 M

to u1 I O N

I if) N

I co r1

I N N

I M . -1

I l- . -i

I M N

I Lf) r1

I 01 .1

I Ol -4

I M N

I W rl,

I N '-1

I N H

I CO N

LO U)

QA 4

t` .i

m r-i

N . -1

CV

-14 CV

I- . --1

Ln' "-1

l0 (V M

N

CO "-1

[N -4

%D r-4

-4

N tV

O1 N

zf' "-1

M -1

CO '-I 1

10 r-1

00 . -I 1

1ý r-1

N N

1 O N

it) . -i 1

"dý -i

l0 -1

W

t0 r-I

to

N (y

N

(C) l0

I in

r4

1 c,

O cN I

CO

-IV -1

MI O

N N i

1-4

l0 r-1 I

1-1

N . -1 1

4m

Ol . -1 1

00 14

N

1 O

U)

I to -1 1

"W

0)

1 l0 r-I

fM

('1 . -4

M

.4

. -i

t1'

N

M N

OI M

N '-4 1

01 I

Iti 0'4

°

N r4 1

00 r-1 1

f') r-1 1

CO 1

N r-1 1

Ol 1

(") . -4 I

M -1

1 1 . -4 1

4

I r-l '-I 1

. ýi

1 N . -i r-1

co in ) r-4

01 L'1 r4

r-1 ri

mot' l0 rl

CO . -1 '. 4

r-1 .4

N -4

C) . -4

ßl 0) tt) -4

O N

ß- r4 OD

ß r-1

-1 M

in N

C) C> H 4 0) 00 c -1

1 1 I 1 1 1 I 1 1 1 1 1 1 I I 1 O O O O O O O O O O O O O O Co O

a a W U W f4 0 x H rl Z O

r-1 O+

N Ck

O Ot

V4 OI

0

W rs,

50

N

H

J

x

yw x

0.4 ýv

sr N N

In <Y' ri N l0 N co C O "--4 .o O In V l0 " Ol cý? . -i 01 '-I m %D 10 O to l7 Ol l0 N tý A

v" cn lD "o CO to O , qC V' .O mot' N d' I11 L1 N U] H

Lf l0 mot' -4 IJ 00 .( N 0 In 00 . -1 -, . -1 N 00 M LI 1fl '00 1D ', ' OO L 110 rý cl Ol rý C O

. ! %

N d' co Co t- d' co In If) "-i -i 'cr ko C1 N m N - N -4 N N N N -4 N N N H N N

ý+ Lei

.ý ca U w w 0 x H a z o Of Of a s w w

O lO N N N O N O O N O N O O O N CD -! ' N to LI) Ln V' t11 d' mot' to d' in Q' V' -` It)

1 I I 1 1 I I i I 1 1 1 I I 1 -I M (N -4 d' Ln W to to tO N r1 (n 0) N V q U1 C') N 114` M C' ('') mot' ce) N M M C7 N m c) V'

N -1 O M V' m d' 'r In tO O co co j cM c. (Y) N "1' C') ý' C') d' M N (") tn (V) N M IV) IRW

01 I I I I I I I I I I I L I I I I C, O 0) . -1 -1 . -i (`1 co N O r ßl . -i -4 N N M CI -t' C' M N M N M N N C') d'

ON Co C) O N O N N "-i CO to 'V' CO C) O N -a f"1 c' e' r) -i' r) N C") N M N N r1 'd'

00 1 1 1 1 1 1 1 1 1 1 L 1 I 1 1 I Ca 00 00 t0 lD l0 O Iý r-1 r-1 00 r N to 01 N N rl M N m C') m (') N N N ('r) N N N C'')

N N to U) In Q1 l0 O O N tO N . -1 in 00 to N "-4 C"1 (14 M N C') m N N N (n N N N M

N I 1 1 1 I 1 I I I 1 1 1 1 1 1, In N N N O DO CO to In a 01 C') t0 V' N C") N M N M (N ---4 N N (') . --I N N M

t! ' lý '-1 . -1 , -1 N 61 N N fit' -V CA co N LI) M N H r"1 N M N N N H N N N -4 N N M

1 , o 1 1 Q

C) Ll 0 1

M cc in

M to t "

N lÖ

N Ö M p

8 (N r1 N r-i N N N N H N N N H N N m U

c

N d' 00 U) N C N in LI Co -4 N l0 ON N (31 N "-4 N . -1 N N N N H N N N -4 .1 N (%i U)

ÜÜÜ]]] in I I I i I I 1 I i I 1 I 1 I I 1 in . -. 1 M 'o ý 19V e 00 [h M m O N U) ('N , -. 4 kp N -4 N r-I N N H (N H H N CV . -I -1 N N

O (N i -7' C"1 N N N N CO 01 . --1 -W "p O Lf) N H N -4 N N H (N --4 -4 -4 N -1 N N r I I f I ý Ö Ö ) 1 "4'

00 c 1 N

L 1 º l n U M b' 0 0 cy

. -I N -I N N r^I N '4 '4 . -i N

tN .i N r-i 01 OA d' O O Ln N N N f') N r-1 ri -4 N 9-4 .4 . -4 "-i N "-i . -ý r-I ri r-1 N

M Ö M 1 1 I M ý1 (5 0 to N 0) CT Ch 110 C1 4 -4 rl N . -4 14 . -I

-4 -i 'i -4

c' dl in to CO d' N N -4 fit' ýO "q m . -I O r-I '-4 U) -4 CO . -i . -4 . -1 01 "--1 . - 4

N I I I I 1 I I I 1 I N N N rl '1' \O l0 N '"'1 d' O

Ö M N

1--4 tO O ch If) C) en N . -i . --4 co A N r-1 -I Ldi ri tO "4 '-1 ON 00 M r-1 -I

H 1.1 1 1 l 1 1 I I I I I 1 1 H O O O

a O c ) O O C D O O O O

J) O O O

x a U w 0 = H a ,Z o ct (DI a s

w w

51

ii

a

Co

er

c '1

" U)

M Cý

C'

N N

m

CO Cý

ko

w l0

tD

O C

a)

co co

M l-

O

t0 'V

mot'

ßl m

d'

Co m

l0

cM Vr

d'

fN l0

N

r4 N

ct'

N O

lfl

01

in

Lf W

CO

q

U)

C4 1-

, -4 N

'-! to

-' -4

tl l l' CO N

C1 NN

00 -I

Co N

Ln N

CO M

LO N

d' I

cM N

O ul

Ln N

N I

In

m N

N N

-4 In

N N

CO . -1

Ln N

u Cfl

ltd . -4

-S- N

a1 -4

Ln N

M N

m L)

co N

w u w w cý x H a Z z o ä ä ä w

(D . -4

O ct' 1

. -1 m

tp N 1

. -1 N

N In 1

N -M

N U) I

C' m

N in 1

in m1'

O d'

I dý M

N U) I

dý IC'

O I

I() M

O -'M 1

LA N

N in

1 l0 M

O dW I

N (n

N If) 1

N

O -v 1

in N

O 4

I . -i M

O VW

1 -V M

N in I

Ln t: P

O "-+

a%

O t'7

O (V

"-1 W I

O q.

M M I

O M

d' -tv I

O '

M m

M "W I

-4 ;: P

'q: N M I

M M

If' N 1

M N

N M

1 M M

-4 M 1

O m

. -I 1

00 M

V N 1

N N

O M 1

0, N

M M I

N M

'. IT

OA

CO

ON N I

N N

ON /--1

-4

p1 (y) 1

t0 M

01 N

1 to tV

C) M 1

(")

N M I

. -i m

O "W

I N m

N M I

O M

N N I

O N

N M I

0) N

ON N I

N N

I- M

M I

M

-4 N I

O N

Co N I .o N

. -1 M I

O M

M c' I

M

Co

tý t0 N I

in N

N . -1

in M

I M M

-' N

1 (4 N

M M 1

cV

O m

01 1 N

t0 m

I N M

61 N 1

a0 N

a% -4 I CI -1

00 N

I '. O N

l0 N I

in N

N m

I (5 M

al . -1 I co -4

in N

I V N

6N N

1 N N

r m 1

m M

tý

%D d' N 1

N N

lG _4 1 in I

N M 1

01 N

'--1 N I

6% r-1

CO (V 1

l0 (N

00 fV I

N

4 C')

1 d' N

N N 1

to (V

N -4 l w -

in N

ll N N

N

M N

01 N

I lD N

N . -i 1 lC -4

C') N I

. -i N

l0 N

1 to N

N tM 1

M N

'O

ul Ls7 {Y+4

cn

,4 N I

O N

'V' -4 1 A

. -. i

CO N 1

l0 N

00 -1 1 110

in N I 0-4 N

If) N I

M N

M N 1

00 ºi

in N "

+w" N

to r-f I

f r) i-1

. -1 N I

Al -4

fV N I

N

to N I

. -1 N

in r-1 I er . -i

c> N I N . -4

c' N 1

"i N

co N 1

NP N

to R

8i 'j,

uý,

l! d'

(A . -i I n V-4

N r-4

Ln N

I N N

LO "-I

"-1

O N 1

N 4

fV N 1

. -I N

[- -4 I

M . -i

M N 1

N N

N -4 1

"-1 r--4

co . -1 1

l0 r-i

O N 1

Co +-1

O N I

CO V-4

M r-1 1 N . -1

. -i I

. -4

O N

1 CO . -1

M N I

. -1 N

En

Nt'

M '. O I

'a' -1

r4 I

O . -1

. -1 CI

CO -t

N -4

Oi . -1

%D _4 I

M -4

O CI CO -4

N

OI f-4

. -1 rI

-4

O 4

I CO

in r-4

ýI'

N r-4 I

N --4

-IV

rl I

O r-4

M

OI -c

N

i[I? . -i

O (I

in -4

m

N ý

G1 1

CO . I, - -41 1

t0 r4

0) 1

t0

C%l

1-14 1-1

Vý -4 LI ''-1

M 1

l0

co r4 I N ri

N -4

I N

in

I 'IV -4

. -4 I N r4

Q1 1

CO

01 -4 -4

r-f

ý -1 1

N H

N

"4

N1

1 O

N 1

O . ý-4

I Cl

in ö

O

I O

". -1 i

O

in I

O "ý-1 I

O

ýC 1

O

V-4 .-I 1

O

M ri 1

O

'-1 'I I

o

N 1 O

CO

O

N H

O

I '-i

O

I OCC M C) W P4 C7 x H a Z Z 0 ä ä a öl

52

)

H

Z er Cä77 Ln

W

f'M Lfl mot- lO f") Ln "-I ON ! - ON to M U1 (°) to O " O 0 00 "r [- . -i lfl l- m m 410 00 r-

Q . . En M N %D N V' O C t11 to C N

4 L. ä GQ U)

co fV a0 O . --I -c7' cV C If) Ol Ln N to wo lfl -T (l -! 0i O M 0O N ß'r1 O ko N O LO N 0ý

O ýr w o0 m k. 0 m m Ul N M Ln lD Ql Ln [- N (N -1 N N N N --I N N N -4 -f N N ".

a a o

V Iii V W W U x H z o ä ä öt a om, w

O %D N N N Q N O 0 N O (N O C O N O d' N u1 to En Wit' 1n It' "V' LO V' En ! J' %r v, In O

i I 1 1 1 1 1 I 1 1 1 I 1 f 1 -ý .ý -4 (N q1' 00 00 'Cr LIl N t. 0 M N CO lfl l! 1 CV) N d' M M M d' M N M M %T N M Cr) II'

C) C) . -4 M N 1- M cl' lD En N N O E l-W M N 'i' M CV) CV) %V M N M M ' N (Y) (V) eil

O) I I I I I I I I I I I I I I I 01 m CO O En En O ('1 M N r-I 00 M C M . -I (V M m l'') M IV M N M M Cl (N N (w) ei4

co 01 N C1 VV d' 01 N N r-1 O N (N 00 N O N Cl N M M M M N M (n M N N M mot'

00 1 1 1 1 I 1 I I i I 1 1 I I 1 co t0 In d' -4 N LI) . --1 O N Q1 cN N -I N CIA M N M m M C') CV N N Cl N N M M

to CO 1' M O r-4 d' O 01 lC OD Cl O w a ko N r-4 M N M M M M r4 N N M N N M CM

N I I I I I I I I I i I I, I I I I N eN l0 N ri r' O O CO L n N M 00 (41 0O M N . -1 Cl N N M M N r-1 N N N -i N N M

M En . -1 O %Q dl a1 N N c7' lD co N N t- N N M CV N N fV N ""1 N N N -i N N M

1 t I I 1 1 1 I I t I 1 I N d' co N rl N d' \O U) CH mot' w l0 O la Ol N N r-I N N CV N r4 N N N r4 N N N

. -1 M N '0 (% lD M LA qf' '-4 M Ln N C En co N N -4 N N N N . -1 N N N 1-4 r-4 N fV

8 Ln I I i 1 1I 1 'l. '

I M 1 1 1 1 1 1 t tft C6 M O l U O m mot'

Uj rl N N N "i N r1 N CV N -I r-I N N Un

- N r-4 M N Cl ) - t N N to Q% r-1 1 1 N

c N

I . N

M .4

w r4

N N

M (V

d' 1 1 1 t i ý1 N I 1 1 I 1 1 1 I d' ý N N C N N r ß , -q -1 " -1 " . -1 r -1 '. 4 '1

N

1. O in O C1 . -I t- (V V-1 O O . -1 CO N r-I CV r-4 01 r-1 r-1 r4 r-1 a-4 N "-1 N "--I i . -i ty '-1 M 1 1 1 ! t I 1 1 I "4' - t l M

ßl ID O C' O oC) CC NJ 1-1 i lÖ N

o 0 Iq sl ri r. i . -i ri N . -t

r) U) M 0% N %0 N C) a-I N Cl . -I CO . -1 a% rl a-1 N r-I N r-1 P-4 . -1 . -i ri -1 . -I

N , ` i I I o Ö M 1 t N . i t

dý N O l '. V ' t l 10 01

r4 rl . -4 r-1 i-1 . -1 . -1

C> M a ý C-4 -W rq . -1 %D a ' 00 -1 r CV) -1 . tI P-4 H '-I o%

ý H OD

I 1 . i 1 t I I b I I I I "-1 a Cl O O O C) W-4

a vo v w w Cý x H a z o Öl ä Öt ä

w

53

0

Eý+

I

cQ ä w 0-4

z

In In

M N -4 lo V' M N CO N E) '1 ) (y M O " ?1 in r- N 00 co 00 M ýi' -I .1 . -i dl rl (71 M Q

V M I %0 l0 ýi' dl ýY' ýf` lf) C 00 (+ 1 Lf) d' CO U; U)

d' lD N ll' N [, d' Ln C) \O N -4 co C) N O O tI O O ßl O M OD L') LI C 1ý %9 O N

1 N 00 CO O r- rM Ln ul -4 mot' :r -tr O lD l0 N . -4 (N r-4 N N N N . -1 N CV N .4 N N N

O q U W rZ, 0 x H a Z O Ot 0 CX Ot

tu Lww

O LD N N N O N O O N O N O O O N O Zv N LO t! 1 to %r to C' d hfl w Ln -W mot' C If) O H 1 1 i I ! i I I I I I 1 I I I I -4 O O M --ýr OO N l0 --zr tll ct' (h r-1 M . -1 t. f) (Y)

M N d' M M cI ýi' M N Cl) M d. N cM M lqr

ü1 ON N P t, W U1 co mot' c"1 N O N O ýP N N . -4 d' cr) PD P V M N C) M w N M C') dq

ON I 1 I I I I I I I I 1 1 1 1 0 00 O co r4 LI) 'j N M O r1 'O CO M 61 -4 mot' CV M M c! ' M N M C' M N M M

00 tl- 01 N O N O . -4 N ) O U) -4 N N 00 N -1 M N M M V' PD (14 N M M N (N M M

OD I 1 I 1 I I I 1 I ! I I I I 1 OD N to dr 00 m d' O A tý CSl m O l0 -i LP) N M N N tß) M M N (V CV M N N (V) M

kD to d' M N N t'M O1 O ko 00 N 0) ß. f1 O It: N N . -1 M N (N CV) M N N N N m ý--4 N M M

r- I I I I I I I 6 1 I I I I I N '' to CV rl c" O co C d1 ý' 01 t, r 01 . -1 N . -1 CM N N M N N rl N N 1-1 N CV M

g C+) d' . -4 O (I) 01 tN [N OD M t- co %D C') CO O

8

N M N N N CS CV i N N N ; N N M U) I I I I 'o

N M CO CO 1-4 C O " lfl to r-1 LI t! ) u1 . -4 N N [ z j N N . -1 N N N N r1 N N N rl N N N W [U ý

i H

,i N N t- O N e4 to Ul O 'a' v d' O lfl l4 N r-1 N -4 N N N N -4 N lV N -4 N N N

! 1) 1 1 1 1 I I I I 1 1 1 1 ! I i ý O r-I In U) CO "D 00 : ' 'ci' O! c+1 .4 M CO to M cV . --I YV 0-4 r4 N r1 N r-1 -1 N N . -i -4 fV N

0) O d' e' N to t- M M CO N O N N V N f-4 r-4 C -4 l N NI rl j N N q .I N N

C' 1 1 I

1 I I N O1 m (I) to C') d' N N Iý .I Lt') -1 W CV 00 .4 N '-1 r-1 N r-1 N ri r4 N . -4 . -i ' -4 N . -1

l0 N [V "7' N M . -4 lD O ä O tf) -i N . -1 00 [V r'1 -4 N '"4 cl, -4 11 N . -i . -1 4 (N r-1 en 1 I I 1 1 I i I ! i 1 I 1 1 N O O N O O) o l O ý1 o l M 1 ) . -1 N r-4 r-1 N "-i . -i . -4 .i -1 . -1

ý r -I

N CA -4 01 co 'd' CO N N N CN i-i l0 . -i f IN rl . -4 co rt (A ri '1 r-1 01 r-1 l4 -1

N 1 1 d 1 1 1 1 1 1 t i i I ! N r4 110 C rn r- 'n co O 0) '

w LO CN in In r-4 N 00 .4 l0 -4.4 l0 CO I`N 0) '-4 OD

1-4 1 1 1 1 I I I I I ! 1 1 1 H O O O O O O O O O O O O O O O O

ofý . -C N M 'AT

04 CQ U W Cs4 0 `4 H f-7 : - Z O a a a of W G4

54

er

hT ý N

H r+l

f -T l9 tß O ýv Ln (n uD tD N 00 CO ;7 O t. f) " r) t'') N N O O a Ol N Ln r- d' t11 d' L! 1 N Cý

L n M ý10 t- ON l0 O d' L'1 l0 qr [- LC W l0 o Ui C! 1

c') r1 C`1 Ol CO O Lil Lf1 ') ct' "ý ri . -i M M M M O 61 O r, t, CJ r- r- C' N qtl M p ýy

-4 ;) O Olt 110 to W l0 V' 4

N M r: 01 '! N N N . -1 M ý-4 N N N N N N N "

a oa _ N M d+

u W 44 c7 x H a 0 8

a Of Of 2

O lD N N N O N CD O N O N O O O N ,!: r cV u1 I U') .:: r U') -1' d' Ln eT to zr d' ;r t! 1 O

I 1 I I 1 1 I 1 I 1 i I I 1 d 1 -4 C 1 M o

M N , in ('m -rr (1) cV M M N: r M M

O N M co N IIl '. D O N r' N C1 O N tn M

M N cV (Y) "r ('p) C M N M M M M M M dý I I I 1 1 I I 1 I 1 1 1 I I I I a

cy% T% -4 C', 'r C, M Ln M m' C) LO co m CO C> N N i. M er M N 1 ß m'1 M N N M "zr

co CD -1 m . -i M M d" (N to 01 Cl N CO N ON N N i. M 'N C+1 -ýZp m N M N M N N M M I I I I I I I I I I I I I I co

CO Nl- CO IN %D (V C) N O M CO . -1 mot' tN ID C14 f'7 N M M ct. M N M N M N N M

t0 O'. 1- l0 tf1 "-1 O, '-1 01 fV I- O M kc . -i U1 (N i CO N M M M fM . -t M N !M N N M M 1 1 1 1 1 1 1 1 I I I I 1 1 1 1 rý

N LI) U) cyý cý1 . -i O ' I' O CO 0) ul CO O r') C1 N ' cr i M N m M M cM . -1 N N N N N N 1'x 1

.V tom- M N O a% M 41 N N wr N ON N 00 ý--1 N H M N M N M N H N N N H N N M

I I I I I I I I I I I I I I I l0 N O O'. to N Imo- r 1I) LI1 M cr N O t! 1 co

N I"1 r4 CV N N N '--t (V N N ri N N N

-4 to CO N S In CQ rr 4

N N

N N N ' N N ri r-{ N N

' I I I I I I I I I I I I 1 1 I u 1 U) O Lft N %. O N d' O u1 N r-t O O V' N N C") tV -1 N rl N N CV N -4 N N N -4 . -1 N N

M d` lD [I) -4 M d1 <Y -q O ON ON M "p -4 N -4 . -i fV r-1 N N -4 N -4 N --4 '"'4 r4 -4 N N

st' NI 1M

C" 1')

d I'

aI 01 1 1

l0 CV I 1O

CO i 1QO

l0 1 1N

M tfý Cb *. -r

. -1 -1 CV rI . -4 . --t i-1 CN "--i r-1 r4 r-t -4 . -I _i r-1

M ý co U-) +-1 N I- tF ri N %o 1'- A N ý4 . -ý "4 N 01 ri rý r1 . -i . -1 . -i , -I

C") I I I 1 1 1

1 1 1 I 1 to

1 M

I I I 4') , -1 -4 N r'1 -4 N . -i -i -4 -4 .. 4

"^. O C O lzr O C' ;, 3 u) c' CN M M i. 4 rt N r-4 -4 -4 Ql r-i . -1

N 1 m

1 Ql

1 00 [

i O C") N lD lÖ

ý' M Oll CO N N

4 14 H -4 r-4 1-4 -4

r- C%4 tI M N ri 00 40 ri l0 M H %D "-1 : It . -I -1 00 0) N -4 CO

-4 1 O

1 O

1 O 0

1 1 O

1 O

I O

I O

I O

I O

I O

I O

I O

I O

I co

I O

H

a a H N M c!!

+Qi AIi Cý W W C9 `"1: H F-7 . -`ý'. '1Zi CJ Oi OI a a w (z+

55

N Ln

11

Eri

.i N

u

O M Ql (N 00 O 'd' 0) Ul 00 T N O) 01 (C) ßl Vý ß1 in M N- O1 N '1 O N X11 to 10 ý--1 N A

Ln N l0 ko 01 in O in [- Cý Ln l0 iJi CA

O 0 C' 00 to M to --1 co cN r- 00 CO N O --" d1 O CC) 00 O ý-I 01 Ol N l0 p1 M M M %. p (y

-4 Ln ON 00 r- In Nä in 4

N -l 4

w o M 00 (N -4 N r-4 N N N N 1-4 N N N -1 r-4 N N

a a -4 N M "

4 W U W fL4 0 x H Z O OI a a a 1 WG

O l0 N N N Cl N O O N O N O C O N O d' N in in U) d' in Zir 'M U) ý: v in w d' in O 4 I I I I I I I I I I I I 1 I I

fM . -4 M N N- U) N- to to 00 (''1 r-1 00 '-i t11 to M N 'd' (C) M d' CC N M M d' N M () cr

N O N I t d' to Ln -'1' IN N O I- O mot' cP C') N d' (Y) ýi' M (C) N CM M ItT N M m d'

O1 I I I I I I I I I I I I I I 1 01 O O O (C) cM 'V' "cI' N N O to d' O0 N . --1 m .T M I' () "14' M N M M M N N M

0) Ol 0) 0) N N M M r-1 _ 01 U) C') [N ri O N .4 M N d' m w M N M N M N N m ds

co I I I I I I I I I I I I I 1 co co t- I DO N -I Ol (3) ob N r-1 l0 O N- N M N M M M . -1 N (N M N CV M CM

N- co tD in r- r-1 l0 O co co IN '-f O in C) lfl N -4 Cl N cM f') M M . -1 N N M N N N Cl

N I I I I I I I I 1 I 1 1 1 I I 1 N- in r- m N co 01 . -1 01 N to in cn C7 ,

M N- M N r1 M N M N M N 1-1 N N N r-1 N N M

. f' %D N r-1 N CO O O tO in C` CO 00 CV tO N N 1-4 m N M N M N . (V N N -4 N N M

to 1 I I I I I I I I I I I I I M O 41 C P% . to to to U) M (C) LO N 6) to M N C+) r-I N (V N N -4 N N N r-4 r-1 N N

8 N V) a% OD co in v Ul Q' N N mot' to OD d' CO f! 1 N . -1 N . -I N CV N N r-4 N N fV '-4 rl N N

in 1 1 1 t 1 1 t I I 1 I 1 1 I I ºn U) 01 to dO CM C") C1 d' M O O '-I L11 N N r 4 ;4 N '-1 N N . -1 N N N N '-1 '-i N N

co f'') in ill N N CO Cl N 01 a% O 'a' to r-4 f') -4 N -1 N N . -i C, 4 -4 -/ N -4 r-1 N N

'd' 1 1 1 l I I I 1 I I I 1 1 ý -4 c , N 1 N C)

N -/ 4

N "-1 r-1 -4 -4 -4 1-1 . -1 -4 '-4 -4

U') . --4 (n N t-- Cl N tO lfl r-I N tD 0') --i _4 N ri . -4 -4 . --! N -I . -i . -i . -t . -i r-4 . -t -4

t''% I 1 1 1 1 I 1 I I 1 l I i I I <'1 ý1 Qý O C N O1 C% OD mot' 1! 1 C ) O -1 . ' d' In .-i -4 . --1 ri V-4 . -i -4 rß-1 . -4 -4 . -1 r-1 . -i

Cl O 00 f"1 to co M N' N O M st' . -i r- 4 ri . -I rl co "-t N- -4 r-1 .1 p . -1 -1 r-1

N 1 1 I 1 1 I I 1 I I I 1 1 N p 10 r cl to rl- t° -11 - Co ao O -1 -4 r-4 -4 r-4 ". 4 -I

r-4 in cv) W C) C14 9-4 DA . -1 to ON "-I in .-4 U1 r-I . -1 01 t- 01 a% 1-4 t I I 1 1 1 1 1 I 1 1 I 1 1 1 1 '--1 O O O O O O O O O O O O O O O O

a a N

RC 011 U W W C7 W H O Z O a a a a Q

CL

o 9 Io 11

N U

CJl

lr)

O . -i o CO N to O) CO w al co l0 ri m 'V' 61 Ql N in N O -4 O O N Ln l0 W N '--1 00

, av cri %D .o ao U) C) -cw

4 in

4 r 4 in cri r: U;

rt M O U) N CO ßl N M U) ý' . -4 N q M . -4 Cpl 1. CO 10 1 c', ' r1 (V r- c'7 C% N mot` D

ri M (N 00 lp ,v N C l0 O N Ln lU Cl N Ol N -i N '-1 N N N N 1-1 CV N N rl r-1 N N

a

V FC W U W w C7 x H 1.7 X 2 O 31 a Of OI w w

O 1ýD N N N O CV O O N O N O O C) N O V' N U) in 1n d' U) mot' V to mot' U) -' "Y' V' U) O r-4 1 1 1 1 1 1 J I 1 1 I I 1 1 1 I I -4 N . -I r-I d' % V r M M N ' V Pl O N ý-i M Ui

M N d' M d' Cr 1 :! ' M N M M -N N (r) fM 'T

. -1 O O M M M N N lo M O O', l0 O (V cr M N qr m 'd' (vi d' M N M M M N M M d'

O' I I I I I I I I I I I I 1 1 1 I On 61 Ol co C) CD N Ol -4 d' O ß1 CO N -4 M tM 'i' M M fM N C') N M N N M

00 00 N Cl 01 . -1 OO O M M 00 kO C'') N O O CV r-1 M N M M M M N N CV M N N f'1 qM

aO 1 1 I 1 1 1 1 I 1 I 1 1 I 1 1 00 N to In to O -W 01 N l0 N v . -t U) O l N N M N M M Ci N tV N (V M fV N N M

kO F- -W `C d' a-, M 00 --1 U) l0 M O d' 00 'p N .1 c"1 N m (14 r') N N N N M N (V N M

r- I I I I I I I I I I I I I I I I t- Ul lD N N co co 61 (v1 LC) . -i al M lfl d' N . --1 M N M N CV N r-i N N M -4 N N C')

C' to . -l . -1 O LN N l0 co N d' O co N In M N r-1 M N M N N CV '-i N N M r-1 N ('4 M

tD I I '

I 1 I I I I I N

I- 1 I I 1 I to K,

8 N N

H r--4.

CO N

m rl

N (V

U) N

N N N r-i

1 r N

M N

N iV

N . -1

O N

M N

O m Wi

ý

8 U] . -4 fn N co l0 C' -4 d' '. O O N ' t0 Gl N Ol E

En CV

I 4 N

fi . -1 1 N

I N M

N l

N i

r-q 1 1

N I

N I

N I rl

I r-1 1 N 1

N 1 En

O d' O ý M r - M ' 1.4 . ßl . -1 (V ) c N t'ý 1 r- to N (V . -I CV N . -I N 1"1 -4 N (V ri 4 N fV U]

ß 1 N CO N . r -4 . -1 N -1 . N 1-1 N r-I C14 r 1 -4

N N

l 1 M N o 6 1 1 I i 14 t m 0 , al -4 t p N

. -1 . -) N ri . -1 N r4 N -4 '-1 r-1 r-i -i r-1 I N

l1 C) C~ ('t CO O? (N Q --i V) r'- i vi N r-+ . -1 cV rt ri . -. .4 N "-i "-d . -t ri . -1 . -4 r-I (N

cri I i I I I 1 0) J I I ci Ln -4 (n ß O "' l0 r

Ül N

4 . -4 _1 r-1 . -4 -1 . -4 . -1 . -4 rl r. 4 r-4

d' CO d- un N M Ln M O i dT %D . -4 41 r4 Cl r4 rl co r1 M r4 . -4 4 f-1 r-$ -4 r-4

N I 1 ý

1 1 1 '

1 I ý1

1 Iý

1 N

1 M 1 1 1 1 4 1 N

. -1 l Iý N d l0 O 01 O C -1

O V' . -I lß) d' . -4 N tV O ri 'O . -1 '. D . -I -4 U) .' l0 -4 r-1 co i

1 I 1 1 I 1 I 1 i I j j . -i O O O o O O O O o O o o O o O O

a pppa

U W w U' '. I+' H ºa Z O .4 a

N

a M a

ep a ýw w

än

W Q

x ui

Lr

H 4"1

. Nr In ýy+ M C l4 M CV <3' O I'D M N ON al in lD M N O r-1 C- M M N -1 N d' lo

cM lO l0 CO in Ol P C' U-) 'T lý ýI' to tf1 N t/ý M

-4 l7 l0 6l QC N Ol tN tO to c' N r-1 tO M -l Ol ýI' N CD O fN M r-1 ri Ctl it) CO : 7` d` in r-1

.{ N M tll '7' "A ý2' C- O N l0 lD O N O r N . -t N 1--4 N N N N -4 N (N N 1-4 N N P1 ý,

a a

cv U W w cý x H a Z O ö' ö' ä cz w

C) tp (N N N C) N O O (V O N O O CD N O '1' N Ln LA to mot' lf) mot' C, Lfl mot' lt1 "M V C' I) C)

-1 1 1 I 1 I 1 I I I 1 1 1 1 1 1 1 ,,.. t co M 7' tO N (V ý4 N l0 ri mot' LA M .1 M M dý M d' M N M M eM N M f' ) Vý

O CO tN N M LO O l0 O -4 in O M N M M 1! ' N Ct) (1) '

O1 I I 1 1 I 1 1 I I I I 1 I I 1 o' M M P M (3) r O's IV N M

c M N N N Cl) N N M

CO N , 4' O% 00 N ýD Ol M CO CO l0 N CO O e--I N '--4 M N M M M N N N N M N N M "V

Co I I I I I I I I I I I I I 1 CO l- M LA M O (V N to r 411 -i to c 7N N M CV M M M N CV N (") N N N M

l0 t0 N d' N 01 . -1 co . 44 to tD 'ýv O in CO CO N . -'1 M N M N M N N N N lM N N N M

N I I I I I I I I I I I I I 1 I N t(1 Ui O N 0 CO "0 l, O M -4 al V' to to N r-4 M N N N CV N N N M r-4 N N M

": r 'r M "-i Co l- If) tO a1 N LI) O CO P U'1 'IT CV -I N N N N N N ' N N M --i N N P

tp I { 1 I l I I I I I I I 1 I I I M C) S C) 111 to . -1 ý N O M CO N r-i M . -4 N '-1 N 4 N N N N 't N N N F-i N N M

8 8 N N to Co tl' to O 'ý7' N Ol N N to o N O 8 N I N N N N N +q N N ri N N M

LI) I I I I I I I I I I I i I 1 I ºn r-O . -1 r-1 tý1 to r-1 (7 N M t! ) QO I eý v r to CO O N N F-4 N tV N r N rq CV N rl -i N N

O O N in O N to N d' N O ' 1ý G1 to N .4 N r+ (N N -4 cV -t . -+ N N ri . -i . -4 N

mot' 1 1 I 1 1 1 1 1 I I I 1 I 1 I 'd' Co CO O rl N l0 O o M In n m i-I N "-1 . -t N "-1 CV -1 -4 -1 N . -4 . -1 ri N

N O M N 01 N 0 N Lfl CC) 0) N ci %D N r1 Co N "-1 r4 "-1 . -1 N -1 N

M i I I I 1 I C 1 I I I I 1 I it) N '-t ' Co 01 O 'd' l0 . -' M in Ol

ef' ýO O cal N t, M u7 t11 O N r co I- I N . -1 '-4

N 1 1 1 I 1 1 1 1 1 I I 1 N { N N to M 01 O tO Ui W N M 0 ý &. 4 r-i r-1 r-1 ý4 ri r4 -i ri , -i '-i

CSI i M -I (n N "-i M ' 1 t) rr q CO C) . 4 CO . -1 . -1 r- 4 O1 O's r-4 r-A I I I I t 1 1 I I I L 1 1 1 I I r-4 C) o C) O C) O O O . O O 0 O O O

a ao N M c

Q W U W W C7 x H ºa . ý z O Ot OI d OI

w

SR

ÖM C

Öu

W LL Z

W ýT JN

W- I-- U.

in co Z

W U. a

CCD J Q. w

ä F- = 7- Q

t- 0 '. O rn \0 M '. D O M '. p co v d - in

I

C] O\ O Oý Lf 00 19 u - M Vl M pý p

u? N N M v M Itl M M Nl M M N'1 M C cr;

V M M - N O\ M CO - - u'1 Z M N ü1 O tt) N - N O t` N co L", U'1 N CO Z

W O N c* N to 0 M - rn - O - O O O M

Of 0

U < m cý w lý Cý _ - -ý Z 0 Oý O? O" d a L i. LL-

O M %D ýo %0 0 'D 0 0 1-0 0 '. O 0 0 0 N - N (N N N N N N (N N N N N N N

O 1 1 1 I 1 1 I 1 I 1 I 1 I 1 O - I N M - ul 0p O\ I` O t`ý - CO 00 C

- - N N N - N - - (N - N - - - N

'. O N 0 d' M CO '. D O\ 0 t- 1- N N - N N N N - - N - - N l I I f 1

u - a' - N N I ' 40 - ON '. O %o kD - N - N N - - - - - - - N

O co N %0 - -r r- to Co u\ 0 N - N - N - - - - - co d

- 0 -

I C'

1 N

1 -

1 u'1

I O

'. O -

I N'.

I '. O

1 1 ul ul 1 co

M CO %0 0 -1- m u1 N U-) M ýp eT ýT co N

1 I

I I 1 I I 1 1 M 1 1 1 r- C14 N d' O' Kl 10 4 - "t N ul - M M 1-

w '. p Nl co N n Pr1 0 M N N N '. O w O %O - I I I I d I I I I I 1 to O

N %0 O ' N - ý

Z Z w w

(1) 0 v ßt1 O r'I = O O O O M u~i cc) 111 1

p% N 1

M 1 1

ON 1

O 1

1- 1

O 1

ON I

p1\(17\ 1

M 1 tIl

N M 0 -

d CO %0 1 - 1 co 1 - 1 '. O O\ CO

Oý 1 CO

co I CO 1 V

- co r- co co 1

0 0 N - N - ýO N uý oo tý N tý ýp r- co M 1 Lrl 1 I 1 I I N I I I 1 1 t I r1 ýo a'. 10 c'. Ln 10 v N 10 10 'o In '. o w

00 %* v ko Ui to Ln N If1 a0 I M f I I %r Lr% 1 N N'1 to l l M M

Ict M iN of %D N N M N Na- M N qt M r: t N - a a a a o a o o o o o o o ö l a - )

0 U < CO 0 W 1- U' M - -J Z 0 O" O p ä

U- LL.

59

t-i

Ca

W

aa gö uw x .ý H Lv H

Lnn z

, wa cý. w ma

ä

1

iEl

U) N

0 6

" C" _' , .. :f LD :; -T -H c- Cý S cn 5 H cß'1 m r N 0 CD UI N N 01 °O , zT ;' O 0 `ý- c. P O Lo U; 01 N M

4 4 M ul M c7 01 M

4 01 c+1 M " 1!

?_ c1 N 4J :J 0 07 It ('`) r H N T M

Cl M

Cl) N Q ri Cl . -i r-1 M W

x O O Q Q CO U W Z O

0' 0 O

0 v

C-. {?

O cri W 0 0) O 1 O O 0) O u O O O 00 N r-i N N N N N N N N N N N N N N 0 I

t- I i N 1 I

Gl 1 1 I I I I 1 I O

.! N N 61 ID O N 0) N 0p (p CV c rl ri N N N ri N r-1 H N H H H H H N

to N H N co M co t! 1 O Lp

1 I I H N r lý

.4 1 (D CO I [ý I-- . -1 m

L H H 1

a r! r-I ý-1 N ri N H N r-1 N H r-1 H H

0 H-1 H

N

0) H 110

O1, w H

N Ö 11) 1-0 i . CO l r-H H ri O H H r-i r-i '-I r-I r-i r-H r. {

CO UI lS U) CO Ln H L(1 LI) M N if) Q) M rl

ON ti

i

10

r! I r-7 1 i

V ri O

r-j1

d' M

Hr

'

'i N

ri 4

Hj i Ö N

H r-1 r-1 H r-i r-q r-i r-i d r-1 r-i r-i

dý H

t r-1

WA tY, N lQ M in cn Ul m M N fM H M M Ln O . -i . -i r-i q . -I rl r-I M r-t H H H H H r-i O UU) r4 Ln

I N N N CO

N H N d N 4 L N

E-4 0 d' 0 CI) N , -4 r-I H M

Ln I 01

N I M

I ON

I -1 H

d d N rI H O -

`ý d H d H . -1

' r i CO H C%j '-1 H H

co CO O r-'

O 0 M ON C- M

O H %D C\I

H 1

0 !

N a) to 011 I I o d

CO OD 0

ri

CV) l0 U1 Oi

A t0 L-I LA lÖ CA

co 5 I LO N 00 I `

M t

LO Ln C j4 l0 d ' l0 lL lC (N 1 d'

d' CO Cr 1 d' l0 i

Cý1 d' N

tL

ui .4 F

M n L N L I to

cn 01 !5 M M U1 N (n H Ill d' M N V' u'1 cl'

d d d d d d d d d d d d d d d 0 ý pp Ö Q

6 CA v w t=+ 0 x H 4 Z z 0 5 cl Q

60

N Ul N

0 Q-

E-4 Ha

is.

.¢

w w

CD Or)

z

a

, -1 lJ) H : ]D C. (r% ßl 0 (\J v' (V N L l0 r-'1

Vl N N 'V' d' d' M LU cM Cr) cr') M 'd' N M M LU C")

Z CD r-i

M N

in Cý

N l0

N if )

0 LU

[`N l0

Ol '

M (10

N IQ"

ON N

M IQ"

0 Q

I N

N l-

O O 7

C CL O N O ra N (N ri

r-I tý N

1-1 H H

ri r-1

N ri r-1

N

r-{

M H st'..

x x

< m w w 0 x H a Z o ä O c 0 U

w

o r, W to LD o CO o 0 ýD O Uo O O O to N r1 N N N N N N N N N N N N N N

lI N C

rý N

O1, 4

O, J) d

Ol d

in CU 01 M 0

r-{ N N N r-1 N r1 r-1 N H N H r-I r-i N

<'') O H m CO tN O d' co N N N N N H H N H H N

CU LU H

H r--!

I N

I I al ', r

1 N

I tý

ICZV r-{

O1 H CO

1 CU

I M

I lý

OD r r-{

O' N H ri N r-I r-1 r-1 r-I H N

co E- r-i co 0 N r-1 H N H

ý H

Co r H

0 . "U

1 Q1

+ lp

1 N

CD 1 Ol

'C H

1 A 9

Ln r-i IL

N r-1

C 1 r-1

U) 0'

On

r-I r-1 . -1 ri r-I r-i H r-H H H

M

C' CO

1 LU

1 CD

II LU CO

1

1U J

O to

I 1-11 ' LU

H LO

I L. n I

N N [N M .4 H N ri Ul H H

4' H Cp

H H r-I H H r-H r-i r-i r-1 r-1

w aQ

t0 t7j . -a

r-I cU

lp ri

I LU

N e-1 1 r-1

fß'1 r -i 1N

d rl

cU

U) rl

M H

1 N

ON

tb

d r-i

1 M

M r-1

1 N

M r-I

N

Q

rl

(31

M H

i N

'Cr H

I M

LU . --I

14

ß', O m 10

r-i H ri rl H H . -i H H H r-I H

47 E1 E+ En

0 r-1

IV H+

M H

0 rl

oll

r-H r1 0

N i-1

r" 1

H

N H

O H

r-I H d

H

1

N H

-11 e-i

r-{ H d

H

r-{ r-I

0

40

N r-1 r-{

M r-I

rl H

r-I

r-i

Op I

OD

N ý

O'

9

0)

CIO

I'D

In

0

Ol

01

0O

N H r-i CD

H H

Q)

C

M I ýD

to I ON

LU 1

LO co I I

in I

(U OD N I

ul ul cN I

00 I

LO M

N N [i A Ln N Ö ko I N N l M 1 CA

v tU tp N C') in N C r-4 Ul to N lV It lfl N

`ý d d d d d d d d d d d d d d d d ý'

ü

w ¢ w v w w t7 x H a z O

ä ä 9 o

Q

61

)

°ý

a4 H

z H E+

0 a

H E-4 H

4 r4 Ö W W a ýi w a z 4 N ik

cý M

z

ö ý3 W 4

" '1-

N C!

CO r--1

a' Qa

00 N

in N

M 0

ý1' V

Ol N

ty' C; r

-1 to

Cr) rS

In M

in CO

' Cr)

Cr) (l

(V co

" Ci

U) (*1 N M d 4

Cr) LI) Cr) Cr) M M ' fJ M N d in

Z N H

N O

CV M

co M

-' N

N M

O Cr)

CO C)

00 to

VT 0'\

N (-

Ol in

d to

r- in

r-1 N

61 0 Z

O H

N 4 H

O H

O

. -4

M H

N 1-I

O ri

N ri H

H H

H r-1

CO rl H

Cr) s -I

Cr) s-i ..

O O E-4 "ý q U W LL4 t9 x H a X Z O o ^ý a

Ö Q

W w,

CV r-I to CV

W N

W N

O N

to N.

O CV

O N

,o N

o N

to CV

o CV

O N

O N

WD N

N ri

N

r-i (4 N

d

N

d

CV r-1 -4 N H

I

H

J

H 0,, H

N

CV LA, r-i

ßl H

A

H

M

N

o

W 0) ON r

m C\j r-4

H -q

- C\j C ON

U) r-I

H H

r-1 N

1

Q) r-I

1

co H

CO H

1 . -H CV

1 tý r-I

s ri

co 1

1 H

t co r-I

d' r-1

H

9 H

00 H

I

H N

O j

N N r-I

0 Iý

to M N to r-I

Cr) H

O N

co d' r-1

O rI ON

I I in

I N H CO

I In

I fV

I 10

I In

N H

I (V

L0 H

N '-i all H H r-i -4 H "! r-i H r-1 14

M 4) in t- t1' r-1 in to in l0 co

N [!

A A -, lSi ýTl O C7' r-1 " .4 U1 lD

H rl H H H r-4 r-I r-i r-4 r-I H r-I

w

O tD H co I

to r-I

1 '

N H

1

CV r-H I

I! ) i-+

L r) H

N H

ON I

OD (')

Cr) H

M ri

I

0 H

1

m r-I I d'

in r-I

I

W

O ID O C1ä

r-1 N U ) r-I

H H

H r-I H H

H (V ri

tV ri

dl N H

H d' H

U

W vI H

0 r-I r-H eN-1

r-4 -1-1 n

C\j H r-i

r-I CO ri r -1

M H

Cr) -t

Cl)

(ý1 Ql d d

Ol rl

rl d

lý ýJ O N A LA

r-1 ri r'1 r-1 r-. 4 H H r-1

d' co u) 1 QO f kD ON O

I d' 1-4 10 N r-i o 0) 5 H r-I

ce) r-

0 LI)

0 r-4 0\ U') l0

M C) S 00 to Lo

CX) N

0 -i I

CD I M

Q1 O l to

N 4 c'1 co

r) 1 I 1 N In tL

, " , " 1 N

H Cr)

d N

d lO

d CV

d N

d tO

d N

d Cr)

d H

d d Lo

d d"

d M

d in

d t-

d M

d H

4 ao v w s14 cý x H a z O a ö Q w w

62

)

0 In W J CIO

0

äl J

aD O a J

W

W

2 N H_

ti a I'D

Q

0 w

w J

w U-

0

w J

I

M , Kr

N

Z

. 67 a-

OD rn M ýo N in Co co ao 0 mt ct v 0 O «' D 0 - r 0 - - M 'V' - Ltl O, N 0 M Co ( L1

to M N M d ýt t lfý M M M N d M CV 't c! )

Co - 0 Un e M 0 %1o n ON r - 'o Co co Z O N r CO CO O\ t- 0 N1 00 - OD G\ M 1- M Z

W O O p - N O r N M N w

O 0 - N M cY I-

U Q Co U w LL- c S - -t z 0 O" d O- Ch. 0 Q Q W W

O M CO CO O CO 0 O %0 O CO 0 O N - N N N N N N (N N N N N N

O I 1 1 I 1 0 1 1 I 1 1 1 1 1 O 1 0 - r - Kl 01 O N d CC) u1 O CO 0 to 0' (N N -

- - N - N N - - N - N - - N

%0 N 00 O' ßr1 O' m ýf' - - N '- - (N - - - N

m O I N 0 o't I mot' N m M CO I (j% - - - - O

- N - - (N - - - N

0 CO to O \o M r r r Co rn W d O' 1 1 1 00 I 1 1 1 CO 1 N 1 1 I Co

[O ul N - CO - %O r Co

M 00 M M r r cf - Ul n CO - r

r 1 aD 1 1 1 1 1 1 I 1 I I 1 CO Ip 1 r N r N N '0 try M 0 d 't M O - - 10

W ýp - - V N M N d CO CO W r l N M - ýO O Ü

- I i O - N M M

z z ~

(n O d' d' - N O N N CO

M Ö Ö - r 1 - t11 (7% Co O t41 > - N -

a 0

in N-

r r N - Co 0%

O Co %0

O N (IN le a0 1 I e ' 1 0 0 d' 1- 0 0 % -

$

O r CO CO - ýo r tn 00 rn "U \

'c7 O' to CO P N 010 CO ct CO

-

I M

In a' w'l 00 d UIN Co r v r rn e N N M N ý n M O i M 1 1 I N d t r gzr C NÖ t l \p Co M

M - r N - %D N M N e ul N (N CO r N

O O O O (DI 01 O O O O O O o

CY- Cif

Q Q m U W u- CD S - --t Z C, 4 C),

M) O

U- L¢

63

0 0 w J

Q F

Q Tg

I. L

0 LL

J

w U-

O Z

w -j Q

C' 1ý Oý N Oý N ýU ýt O Co h"1 tt e M N D - O CO CO - CO tO - t r-% N N M t- Ln - 0

U) M N N LC1 M Lf, z; M �l Nl M M M t! 1 u;

ttl 1- 00 'ci' N e U l0 M N Co ul to %0 r- Z ttl rn N N O Iý ul 00 ýO ltd O - tO U CO tO Z

4 w O 1- it1 - N N M - 1ý M - O - N N LL!

O O Q

¢ in U w Li. - CD S - --i Z O d d O" ä

Li 1L-

O v', w O tO O O tO O i0 O O tD N - N (N N N N N N N N N N N N

O I 1 1 1 1 1 I 1 1 I 1 I I I O 1 O - 1- (V t N 't ON U1 O\ tO - O-, O 10 O' N M -

- - N N N - N - - N - N - - N

M - M lY u1 CO p' to Q\ N N N N N - - - - -- N

O% 1 J ö 1 1 1 CO 1 O I O\ - N 1 - N tO N - CO

(V (V (V (V - - - - - (V

UN - C 0 Co N M rn N 0 CO 1 - 1 1 1 1 N 1 1 Ul 1 M 1 tý I CO

v - O N a) N O - N oo - uý - tO - rn

Oý tD 1- 0' 1,0 - tý e -e N ul tO co N M O i 1 1 1 1 I 1 1 I 1 I 1 N

- - cc ý- ul - N ul O tO M M - d' u1 tO

w N I" M Lr \O U'% N O M ul W er - 0% - - - - - - - - Q O %0 U

1 -

1 CD

1 tD

I -

1 N

1 M

1 ýt

1 M

1 OD

1 ýt

N -

1 O

1 !T

1 N

ýt -

1 to

O U

to - - - - -' - - - - - - V)

W W

() O O _ N 1'1 N M _ M N If) r _ t 1

O' N 1

vý 1

O 1

- 1

- 1

O 1

tO 1

N 1

O 1

a0 1

tý 01 1N 1 ifl

M O O - - ý _ 1 _ 1 1

et CO ýD 1 cV

I N N

I a' Co (0 O CO ýO

ýp t CO

1 O cO

t- - t Co '- r- u1 r- ~ M 1 ul t 1 I 1 I 1 t C N- um 1 t

i j tr ;

ýO U Ln N- to ýn r1 v ýp CO to

rn le to Wn (o v N - N W I d I d N I 'p M Kl ul I N CD

M U% nt r- e

M

' M N (4 M d tO 10 - to um CV N d- W N

- a a a a 0 o o a lo o o a 0 o a o -

ö ö < CD O W U. - (D M - -1 Z O C), ä

d o Q

LL.

u_

64

H H

Gaa a °9

99ON

z rj w N u) aÜ

tn

w l: i

ý4a äwS U)

O "

(V O ül <i' d' 110 co 0) N lD CV CV Ul lD d Ol Li 5! )

" rn

" N

" m

" d

" U)

" M tll

" M

" M

" M

" m

" 4 M

" s-n

" cn

" ' : r)

z N m in 0

H N

O O1

O r-l

M l0

o d

U Ol

H [

O N

d 0)

H 0 to

rn lI)

(o Gil Z

O O CV N N r01 N N O O Ol H CV (V (z7

o .ý co r-) W 0 H a z O ä 0 0 0 0

0 N rl tO N to CV O N to N o N to N o N ko N o N O N o N 10 N O rl

1 Lý

1 N

1 4 N0

4 1 Ol

O N

I r-1

I CIO

1 tN

! O1

1 (71,

1 N

O r-I H H (14 N r-l N r-I N 1-1 CV H r-I r-I (NJ

r) 0 r) cfl 1ý

in O N m IO to ON W

r-4 H rI

1 N

I H

I N

co +"{ N

W r-4

A

r{

1

H

1

r-4

1 ý

I

rt

1

ri

CO rI N rn

to H

H CV

N r-I

H N

lS r-1

H N

N H

00 H

t- r-H

V ri

N r-I

0 N

OD [1' 4 0

rf 1

O 1

IL 1

Ol 1

l0 1

0) S rl

I rI

1 S

In ri

1 tD

I 'O H

1 l0

1 ß\

OD H N H H rI rl rl . -i rl ri r1 º-1

t'n r-1

0) H

to H

CO H

CO H

ID rl

d' H

I[) H

N H

ul r-1

CO f-i r-

CV oll

CIO A I U')

r, I t1r)

rl O

L In rl 4 LO

rI tÖ r-4 r-1 r-i 1 H r -l r4 r -f r-1 r-I rH

lS r-I

N , -4

Lf) , -1

d' r-1

ºn , -1

M H ON

d' H

fV H

N r-i

0 s-I

C') r-I

[N , -I

t. I . -,

14 x

a

7 r-l In ri 1 -I H ri rl N A

.4 t. 0 0

t ] H . -1 1-4 - r . -i r-1 r-I H H r1

En Ei

0 d' 0 (V N N r-1 N O O H r-4

N 'i

cy) q

v)

ck 'A A d ý d d ý ý 0) (3; N

O r1i " Ln r-1 r-I H r-I r-i H r-1 r4 r-l

11 OD ý N

c0 L

a% I

A

of

rn I

0'

CIO

W

ui 0 co

co t°

ON I

O i

O

l t ý ' C O C rn co

m p Ln r.

1 ýi t0 co

I L: ) -l

tO to LO r

N In

.4 .1

01 d! '

A

d'

(A

l0

if M

to 4

M N

N Ln

[i' A

d'

CA Lf) W In

N Cc) ! to

ý' CM

d [fl

d C-

d N

d N

d d-

d N

d cßl

d r-l

d Ul

d -'T

d C \f

d N

d [j'

d CY

d f')

d `ý

u < cn u W w o x H a x z o El cf a ü w w

65

ö Fl

9 a

ä

N ID

ri Hý

c0

N U

N

4a äý

O O O

w co

ý M

0 N

cN C %-1 M

0 N

d N

cA (M

C. (1i

O (1

O Cl

rn V

Ir 0 co CO

"I CO

Cl N (*l Ci ýJ M tI} M C*)* M: M M M M d' l1

Z N C

d pp

r- in

m CO

m G"

O N

r4 (1

N co

in M

0 to

r4 (V

to to

CO N

CV d'

cM CV

N r4 : 7-

cý s7 lD V'

cI C; c-I

(V ri

N [-I

N c-{

0 c-I

CO cr c-i

ca c-I

rH rl

01 c"4 c-i

tV c1

M r-i Z

o < as O w w CO x H .a z 0 C C\l

CY CM7 CY Q,

)

O CV

M Vq

in N

in N

tD N

O CV

to CV

O N

O N

in (V

O CV

to N

o N

O N

O N

tO N

o I I r+ 1 rn

I co

I ýo

I a, 1m r1 1

co I

rn CA o

T-i T-1 N N N ci CV c-1 c1 ri (V T-i c-I c-4 N

C\j N

N

.N (

OI l

(IN to c{

0 N

CO c-i

1

N

Lý c ci

tý jpp I

-

[-

i

j

N Qý

C\j c c-

O N

N c4

rn c-

t- ri

O (V

'. 0 ri

d' r4

'O c-I

to [-I

N ri

d' c-I

0 N

c-I Oi c-I

lÖ T-4

ýpI c-I

tÖ T-4

Qi c-I

ul t'i ri

ui r4

ui c-{

tp cl

il

cI cc-i c'D -I

I 0' c-I

co

T-4 TI N-i 0 TI tý 04 cc) ILO--l

I ci

Iq 1-4

A

14 4 4

IL N

c-1 IH c-I ri c-4 vi T -f % c-ý c-+

W Q

t-i

ý

w ri

N -

[1 c4

d' 'r,

in c-i

M ý

O

I

M

1 N

M

I O

Cl

c' 1

LO

[+7 Ö

l0 H

N ý0 d [ Uj r-1

A A A CV 01 CV TI (V t-I

! N c-i

4

r-1 c-I cH c-I r4 cl r-I cl r4 c-I `Ei L I

in E in

d' c'l

O (V ri

N N c'I

ri c-l CO

c"1 ri

c-i ri

ri Q

c-i c' 1

N T-i

M c4

01 kO ck Ol tý '

O O '

O CO '

4 4 O O

r4 %_4

CO U"

c-+ -

co 0 ?4

i 0% N

co [N

kO

ti

O, I

0' N o

01 I cl k 10

N t N a) O t OD t C D c

Cl '. O '1'

0 r4

I ko

I C-

I , I W

I ýG ' or"

I NN 0 in I I 00

I M Ol in in 9 to N lfl CO l0

N In

( M OD CA

N 4

to

t1

d'

CA

tI

4 ( N

l0

Ln

W

to

in Cl'

EN in N

cl

d lV

d N

d N

d Cl

d dr

d (V

d Cl

d V-4

d d'

d d'

d tn

d en

d -0

d in

d N

d ''

a 4 m U w w o x H a z o r o a ä d u w w

66

)

0 W J

F-

o_

J

3-

J

W Z wI

I

N

CD

U-

CL 10

Q

tL

W J

Lß. 1 U-

O Z Q

W J Q

00 Ln

Z

W

c2 Z

O

cn N

J U

J

U)

0 trý CD Co ýKI- Ln c, r-- n -e 0 N t 0 N Q Z; a) O N O C O M M N -. u N N M C M ül M M M M - M M M ý- cn

%O N N O CO ' N - N - OD 00 ,O r- o M Z d N Oý 00 O - 00 M 00 N ßf1 N U1 - ýD Q ¢ w O %0 M O N N - - OD - - N Oý - N M w

(Y ý

p O Q [a O w u- -1 57 Z

Om K Q lº LL.

p M \0 %0 ýo O %0 O O %D O '. O O O '0 N - N N N N (N N N N N N N N N

O 1 1 1 1 1 1 1 1 I l I I I O I O

- '. 0 - N - M ON M CO ýO c o 00 a0 - Iý 00 N M - - - N N N - N - - - - N - - N

p N N LC) O % N N 'N N N - N - N

Oý Uý O I 1 1 Co I r 1 1- N I 1 N CO C% - - O aO - - - - v - - rn - - -

N - N N - - - N

ON r- 0 r- o 'o %D %Z ao nr p N

C0 Oý 1 1 1 1 1 1 M 1 1 I 1 ýO 1 CD - 00 U1 cc) '. O N ul - t[l Ul 1ý M - - Oý

M r- r- N Ul Ul% OD

r CC) Z 1 le U - N ý K'% WN - Ln - M - N 'ct -t IN

n N ýt d M O M ýt M %0 w W O - N 1 7 1 1 1 1 I N 1 I O 1 ºý1 ýO O

M N N O\ - N M - N - U to - - - - - - - - - - U)

Z Z ui LU N

M O - N - - - - N - N M U)

- - ' - N p1 p O O O O m O N

O

t- a10 Ö+

r-, m' Ö ä0 12, 1 1

,0 N- CO C; Iý CZ %z Co CO ND I- le 1 r n 1 l N 011 l 1 1 N 1 M

0 % n

1 0 . e u -% 1 - t0

%0 M CO n Kr ýo N ýo 1, N to M tf1 C Iý N

In N N M M Ul '. O u1 d

sT - %D M N l! 1 N e t N d' LID d' M 'e %0 ý O i O O O ý O O

M - O O O O O O O

tY ý O O

ä ä ä Q ¢ m U w 1i- t9 S - -i Z O O

U LL .

67

d1 kD W J an d F-

O

I-- Q J ZD CL. O G.

J

W

W

N

I -- º. r

Co

LA. C i- %0 I-

C) U- W

aJ

U.

O Z

a W J

O M Oct n

z

v N 1

L

C

O DG CD

W CB Q

'5 cM l0 %D d' -t' l0 F -i' al to CO /O O u1 00 r-I . -I '1 O 01 M I- co co co r-1 O o r

t: m -I c. r M U) C'') N c N ý1' M C) cý ) ýf

N U7

O Ln W t- 0) 01 -I co N O N tD l0 N N M F- l0 LO %D -4 N l4 ;r a l0 Ln l0 -i d' '-1 M

LJ (n r- -W -4 10 ß Q1 01 Ol Ol co O m 1-4 rI 1 -4 r -1 -1 r-4 1-1 -1 ir

cr_ a O

Q PQ U W W U' w H F4 x z O d ä ä a

Q

Na: LL- (sr

O c'r1 %D l0 t0 O l0 O O O l0 O O O l0 O N 1-1 N N N N N N N N N N N N N N O I- 1 1 1 1 1 1 1 1 I 1 1 1 I I I -4

N N Cr) r 1 In N U) N U) O l0 r-1 l 0 U) 00 M . -1 N N N H N -4 . -1 N r4 N 4 . -t 1-4 N

l0 -I N O d' l0 k0 d' ON U) O U) N N r-I N N N r-1 N r-q -1 N r-1 r-1 N

I ; i

I CIN N N . ß

-I . r -1

U) O r-1 co N In r-4 d' M ýp m 01 d' M Ln r-i r- rl N . -I N r-1 N ri r-I r-1 _I . --I . -1 "- I . -i N

co 1 1 1 I 1 1 1 1 I 1 1 1 1 a0 Rd' ON l0 . -1 O M N U) N M N ItT O . -4 -4 N CV 1-4 . -1 -4 r-4 N

M CO to O -' C1 N -4 -' N lC N "4 Cr) al "- (3N "-4 . -I N . -1 . -1 . -i . -1 . -1 . -1 . -I .ý ^ý . -1 r-1 d O M ' [, - N N ' N C ) L)

( 41 O (N U)

1

l0 c'') 00 N l0 ' O N '-f . qr O . -i N LAJ r- CO ri -4 -4 . --I r-1 .4 -i -4 r-i Ol ri r-1. cl: to 1 1 h' I I I I I I I '. o O U) N S f'M O . -4 O M 01 lf1' U . -1 . -I . -1 0-1 r--1 r-1 (ý]

N ýi

8 LAJ N C) ß r-q O N N

a ý U2

Ln I [N - i1

I .. 1 I

. 1 I

i- I

-i 1

O1 1

01 1 1

--1 -4 1

co 01 -4 1

-4 1 L1

rj F ] 00 O 'n o O 00 ( m (a O lý p1 H 1)

CNI ý . -I f-ý l0 r-I 01 " i ON al t, N CO N p' N %D CO 1 tD ( N

N O

O LO Ln . -4 [- -4 co l0 lfl kO t-- lD In lfl in l0 ß1 C") 1 I I I I I I I I t i I I t r` cY' m l0 O N d' ß. l1 Ln Ul -, Ln LC) N

r-I

cM d' CO U) dl ko M V ' ýf' w t ) M R: r M .0 l0 N 1

MI ,

f N N 'P N I )

N M ko d' 1- in H N M M M ri M N M d' , - 5 . I I I I I I I I 1 I I 1 1 1 O O O O O O O O O o O O O O O

o 1 Q co 0 W P4 C7 x H a ,Z o ä ä ä ä ILL- w

68

)

uý ýo w a as H

Z1 C>

a

OCr ! Sa

tý.

vi H

t4

0 cti w a ýi

w a z a w a V.;

0 Ln u

fn j C%l

S (3) CO m In a) l0 tD N cc) lD co C\ N

M N W

to M r1 M 4

V N En M N M N d' N M M U)

C rn U) a) Gl rr P7 H lý ýIl N ch V"

d' O 0 l0 u (V (V a) O a' M

N U)

m M

a J)

¢ 6

L, 0'. N c'

H H H

d H

N r-I

N H

O H

cD O H

O r-4

H H

CO 0) H r-1

M H

u .y 0Y1 U W G. O X H ý-7 Z Z O , --{ C7 O

m O

Irv CJ

0 O N

I

ch H

I

c N

a) (N

I 1

w fV 4

O CV

I

0 N

O Cv

I

O CV

1

a) N

I 1

O N

1 0

O cV

O N t

O CV

I 0

O N

1

0o (V

1 CY)

0 H H N

r{ N r-{ N

' 4 N N

CO r-i N

CO r-i

U) H

G H

l H H

l H

co H N

ON to H N N H

0 (V

1

M N '

M (V

N I s i

lD

;; r H

'A

CO i

9 V) 0 ;

Q

LO H

4 N H1

'

(V

C\j 1-4 H

N 1 0

H 1

(-. j H

CV r1 H r-{ r-I l

H H C} H

r-1 H (V

Co

U) H 4 H

O cV d

CD H Ö

H N d

H

r-+ N

Ol

ul r-I I Q'

c i

H 4 H A

co

N1

Cl)

I N

cf)

1 N

Lo 1

In

0 (1

Ol CO r-i

, H

l H N r-1 1-4

r H H r-H r-I r-I r-1 H

lý cM H

1 (71 CO H !

LO r-i

01 r-1

i

to r-i

CI '

CO -1

I 121)

M r-1

1 N

r-I H d

M H

1 N

n1 lfl H

1 H H d' 00 t-1

1 S N H

C H r-1

CO H r-I r-1 r-I r-i r-1

1., H

r1 r1 r1 H

to H CD

W

I N

fn

I CV

r-

I to

M r-1

ßf1

I1 A +-4

1, a

H 8 d r-i

-H

M

I H

0

I (A

0

I 0'

M

1 N

w 4

.4

a: OCý

0 H

c7ý H r

Ln

Q

0 H H

0 cr r+

A N r-+

H N H

d O H

all CC) °`

O r-i

k O H

°° co

N r-+ r-i

cn H

H Lo C

d' tý lD ,I I N

CO

[

H

H 0

ý

r`

co

r -I

N co g OD

CO I

N I

OA

Op

H

Oý d'

M { `n -

o', in

O

Cl')

co

Ln vi `n , ,

-j Ln

j r

ýr`I M

N N %" 1 dý d M d' a) 1 cV

'ý H

d M

d [1

d M

d In

d to

d CV

d M

d M

d M

d q

d N

d M

d N

d 1'

d cn

d

ö U

w Q ca U W w O x H a z 0 Erl ä c ;5

ö u

w

69

V

H

n

gu Cöz

40 xg1 H1n

a 8

NQ 16

to ýQ

" l0 V` O H N co rl M M 0 M :. l) l0 CV H l to 0, - 1 r-i to N Co M rN rN lD co (' IT H

" " " " " " " " " " " " " " " " " " U) M H C d V N l1) ) N M N d N m M L J]

N cn

Co Q

N Ol

a) cý

a' lD

a' 01

lD .i

N rl

O V

O N

O Co

a' G1

d' N

a' u')

O N

'. o O' Z

C1 N C H M N M C; " O" O" O" Co

" 00

" " "

L. .1 r1 . -I r1 ri r-I H

Ö

Q m cý w w o x ý-+ a z o a c ä ö 0 0 U

w w

o M tD lD ,0 O 1.0 O O tU o u9 0 o O l0 N H N N N N (V N tV N N N N N N CV

0 H

1 N

1 N

1 rý

1 M

! Ql

1 Lý 1 t11

1 H

1 1 1 1 1 O H r-1 fV N N H (V H rl -I N r-I A -, N

01 j

In 'I r -i

N N

j

a' N N

pp A

Cl) N

Co r1

H H H

to

C\l

Ol H

rl co

CV N

d' H(

rl

O r-1 CV

CV

CC) r-i

rl

r-i N

01 H

N r-4

tý r-i

N N

N

IC) r-i d' H

ri Ln rl

.; T H

co rl 1 w ri

CM r1

m rI

IN r1

r-1

r-i (V 0

rl

00

r\ °ý m H

r!

ul r-i

4 Cn r-1 c H

0% r-i r!

(1) H-i N

H H d

d' r-1

M N r -I 4

H H d

N r-i H

q) rý

'ö N

w (Z o r-1

rl

OD N

lD H

CA H

M r

O

l0 i r

N

r-i

d' H

('1

lfl -4

J O rd

° Co rý

N H

a' ccl)

N r-i

O

_q

(') r-i

-4 CD

0 H

CO

d r-1

C\l -i

r-r-4 1

I

CC)

H H

w O

d' lD

CV H

H

ý1 OD

' O

H r-i

I 00

ýI N

'. D I- rn

uj

CC) '0

O

Ok

O

ap

M ul G' L"

O ri (31,

t

tj

Gý 10

c

C ri

Ol

r-

t0 Lrý

'D

uý trl

Ln u1 co

[1

C- M

N M N I

co

[N

tl 4

N

'L I

U) N

Ul

VI'

'i' A "

kD

U) M d

M N l0

U1 1 N

'ý eI M l0 C') u1 N

d r-1

d M

d N

d c+)

d 4' N cri r-1 U) N

H d ö d d d d d d d d d

ö v

w ¢ ao V w w C7 x H a z o o

ä ä cr p

o w

70

p L

O en °4 co

cv

p wz

En L;

tO zö

ww4ý 14 1.11 W

" rn LI)

rn CD

() ri

U) O

M 0

' 01

H ßl

0 Ul

'0 LO

u) t0

r- CO

w co

C) co

N -+

' M

io O Cam.

V1 M , Q cj [i N ti) M N M N Cf N M M Ll (I)

Z M C1

ul N

H M

Ul Q1

d LU

CO In

d O

to H

l0 M

l0 O r{

0 CN

d O

ri O

0p d

H a) ý:

ýu .1l Iý ý7 1-1

0

ri N r-i

M H

M H

O H

0) O H

rt H

H rl

OD O (N H

CV H

i, ]

O o

4 x' U W C 0 x H a £ Z O 0 Ö 0 0 0 ü

0 N

lÖ

M H N

U) N

'. p NJ

W N

CA

0 N

Gi

LO N

0 N

0 N

'. D N

al

0 N

'. o N

1

0 N n

0 N

'

O N

' N A g

ri "--I N N N l

H N rl H l

H CD H

1- 1 N

I r!

U ) r-I

CO r-1

C N

OI ýn H

N C, J I

G` r-q 1

N -C\j d O

M CM I

(V i M -I

CO I '0

N r4 I O

N

1-4 I CA

d' 14 I th t -

N N 4 01%

ri H r-1 N CO , -i N

H N

r r H H r-1 r-{

r-4 - N

d' r-i

(A H O N

(A N H

ý m r-i

CO H H N

3 +n H

4 , U) H

Un r-H

' 0) ri

NH cN , -I

I H H 0 N

) r-i r-i r-1 H N H r H _q `4 I I H a H

C-1

00 H

Ell

ul r- i

14

N r-i

11 Ir

I

01 r-1

ri

r-I H

M H

C\ 'i

tD H

r-4 H

r-

Co r-!

v-1 r-I H r-i r-i H r-1 H H

OW r-i I

co

lD I

fý1 I

d' I

U) CO I

N H

s 0)

H

OH N

Cl I of

O,

d' I

M

tf) 1

M ID

a0

rH-i Hr-I H ri H r - ý r-4 H ý-l X71

Un o0 r

d' 0

r-4 N

I 1

`ý I

N O

1 [0 rn H

`J d

O

I 0,

N ýI

N N

N

d

W Ul

Ln rý ý1 r ri H e e1

d' ýO

N

I r-I

00

0

1 01

N

1 . -i

t I N

co l N

I op 1 ý1

W I v-I

I 8

ON 1

d' H r-I r-{

M in Q

r- ul

cG tÖ

113 H ßi

:0 ül 1G :f

u I o0 S

.) v M N

N ýp

m

N ýj

N 1

IL d'

14 co

ý1

N l0

Vj

M 11* N CO

t1l,

N 4 N

rl

d C7

d U)

d c7

d c/1

d CN

d r-1

d to

d N

d M

d U)

d N

d Cl

d H

d V

d Cl

d 4

Q Q M Q ca V w w o : H a £ Z o

är a u

w w

71

°ý W J

m

a- J

W

S N I. -

m

Lº. a 10

m

0 LL w J

w w 0 Z Q w J

I

un 00 M

N

z

0

(9

tO r- mot CO Lfl M j- M ýp Q tt\ M Oý ýO D t! 1 - N 00 0' CO tN M n N N ! Il M - Qý p

V) M N M N Lc M N M N d N Kl M ýt U)

O' O' O tO M O N N aD - tO - - O' tO N Z Qý tr OD N Oý M N 10 N Q 1- tO U) - p1 M

ui to to - O O M M O N O O r rn N w _ _

cl: O O Q CO U w L- C9 = - w Z O Od

0- O"

0-

Li.

O M tp 1.0 to O '. O O O to O 1.0 O O O '. o N - N N (V N N N N (N N (N N N (N N O 1 1 I I 1 1 1 I 1 I I 1 1 1 t I O - 10 -e O O Q. to co v co CO 0 v h rn N - - - N N N - N - - - - N - - - N

M O' O' ýP N h O' M ýp - N - - N O Ö N ( N l', r- - M %0 - ýO - O L ' -

pý N - - N - - - - - '-

- %0 N ul N ul %0 N 00

O i to V l ' Co rn O u - O ' e n tO - tN - N - N - - - - - - -

N 0ý e u1 '. D O' M Wl N to to tý 1 O M t 8 M N M O i 1 tý

N l If' I

W %0 N N 10 N - M (ýJ M w O O h - 1 1 1 ON O N - O O

O - to

- N

U

Z Z w w V) ý1 O M O M - U)

in I

%0 M 0% 1 N1 1 0 I O% _ _ I I- I 1 1 wl

Co m O> to N Q\

N N O 0 - d 1 to 1

Co I

to 1 -

CYN ýD

CO j - CO N - N

%0 :: h h _ N co jl) %0 'ö 1 0

1 %0 `t

N 1,0 ýo CO "'

LCl lfý cr\ ýj

Kl , T' c% Ul - V_% '0 CO `d

M Co u1 0' «t u1 t[1 M ct r- M N N M 7 ` M e h N 1 1 N

t t M ýO N

- (N u'1 M M %0 N t) M N1 %0 - M (-4 ul -

- 01 CI: ) b 0 c 0 0 0 0 0 1 1 - 0 0 c O Q

Q Co U w 1i CD S - O" O" C), U-

72

Z O

H

J

L". º tfý

U vL

Oz

ZW 03 W U' W"

J.

Vf ;N

W V) LL Q

J

110 WzU J WS äW

N 2- - M v -- co ON N a J D

.D N N M Q\ M (T N (T - cj d o

N r, 1n

C/') M - t; ý; ' v M Hfl Nl N C M C "ý1 M rr1 X11 _

r J - 0 0 \D C o\ v ': O, ! r, 'O G' r^ iI

(N ( \! :n tC N 11 O C`\ v CO CO M in ` R_

O o

- N M C. ) Q CD 0 W U- U) 2 - - 7_ 0 Od Cr Cr Cr U

U-

0 t '. O %a 'O 0 1ý0 0 0 '0 0 'O 0 O 0 %. 0 N - (N N N N N N N N (N N N N N N O -

I 10

I N

I I I d'

I C'

I t11 C7'

I .1

1 -

I 00

i -

1 tN

I 10

t G,

i trl

G - - - N N N - N - - N - (N - - - N

CO '7 0 iý 0 '. O (N N N 1 ON ir\ - N O N N - M , 1 L'I CO

1 G^ N - t M 'p

J, . ) N N - - - - - -- N

C\ - t+l 00 U1 co v 1ý O

CO 1 1 - I N I r-

N 1 1 N 1 1 I d 1 1

11 CO M N I- O\ - - ý1 - 10 d 10 - M '. O of

0 '. D CO %0 0 _ u1 u"' t N rN h N O 1 1 1 I

" t ýt 1 ! M 1 1 1 u1 i h co ýt %D U \ Q) - O d - d N 10

_ I- Kl ü1 ct 1N tn V) N M - O '! ' Ln w

I + I U O '. O N t l ul N N 0 0 t l d- Ü

U) - - - - - - - - - - - - "- V)

z ui Z 1- Ul - N O N N

LU v)

n o

. CO - I

- 1

- I N

- I

- c0 - f 0

- O\ CC) 1 1 c) 0

0' N o 011

N n

M O - O NT N I I I CO ' (IN

I CO t "

2 O\

10 (-4 ý 0 3

CO c0 '. O I

'ID $ ü1 p'

1,

= c ti, i C-3 C ý.

N CO d d ýO

t d

M ifl

M C N to pn N

d t- M M ul N M N d d N M - d N

- O a a 0, a a a a 6 () 1 6 - 0 0 0 0 0 o 0 ¢ m U W LL O S - z O - G Ör Or U LL- I

73

zi ! 0 H H

M

H O

2 w

ý7 Ü

U) w

Au 4

U)

" n

U)

rý N

M

O OO

ci

(Y) C-%

m

CO N

Sr

Ul N

LI)

N r1

M

N CO

U1

N (9

M

O) CO

N

Co Ol

M

Ln -1

Co

ON N

V'

V4 O

Co

k-O M

M

O N

Co

CO G`,

ý(' :!!

Z ., Zr iO

" rn

0 0)

" N

Co (N

" ,n

Ol Sr

" +

CN Sr

"

O l0

" N

Sr Co

" M

H M

" O

Co Sr

" OD

Ul Co

" O

(, j CO

" O

C) N

" O

v-l O

" (

M u)

" Co

0 01

"

V' N

" N

yý W

N C:

w Q w O w w O X H a Z 2 O 0

N a

M a

Sr 0 V2

w

o

0 N

c-I

Co ¶-3 N ci

l0 N A N

to N

I .ý N

l0 N

1 ui (V

O N

I ON c-1

W N

Ln CV

O N

I co r-i

O N

Lin r

lO N d N

O N

I oo c-3

Co N

r (V

O N

'o c1

O N

"c1 c-I

O N

C) c-i

Co N

N

Oo

ON w vi

ri 10

N "

0 N

01 c-ý

Sr CVt

M N

d' N

c) N

t- r-i I

t'1

S' r4 A

c"4

0) t-I

ýpI ci

N ci

lÖ ri

O (V

011 ri

ul %-I 4

e'1

cn cj 4

V-4

ý

N CV

cA N 17,

co Co c-l

1 t-i (E: l

c-4 N

q N

(ý cý

c-4

N N

d N

N V-4

ý s-I

N CV

d (V

N

4 c-4 r

I- c-1

LA l

Lf1 V1

14 ci

co cl

1 T-4 c

M c-1

N c-,

N c-4

to %-i

0 N

0, c-3 00

C- rl I (IN %-i 1 T-4 Sr V-4 CO

cý c-ý

V4

CO c

c-" vi 0

H M

c-I V-4 %-I

cl r4 r-I

W a

Uýp Vý

.l

En

d V4

CTl

`-i tl V1

Lo c"4 4

IH

M ri N c-I

T-4 c-I d

Iri

[ý

to t'-

%; r cl (N

ri

V4

c-+

[N rA

1 d' VI

M t'3

%-I

N %-A I

r, VI

0 cl Q1

C)

N %-I I

cý C+

O t'1

Ol

N N-i

I r c-+

O c'4

Q)

M TA

I c-i V-4

O C"I

d1

O V4

I C))

O 1-4 Ci cl i

CO

N

d' r-4 c(l) V

N ri ý

LI) %-q

I A

c'i

N ri

8

W

cOQU CI

ý

E

`r t- to

r r4 ON

co ci of

Vol ö ýq

0ö 00 r- 00

N GO to I

[- I

l0 ýn

01 co

C" U) to cl c-i lS CO rn N C^

U) N N

f

N fý1 d' J 1 Co l0

N 4 Co

N Sr M 4 4 N

N

d C'1

d NN

d Co

d fn

d Co

d N

d N

d N

d t+)

d Sr

d VI

d c+1

d N

d C+1

d Co

d

I 6

w .ý ao v w w ED M H a X Z O ä ä d Q

w

74

z C I-. N

co N a igz

äü HN

öý. w

" ch O N cc 0 N c1 d' c-' l0 cc 'D " O M ýD " R v Cl O u7 Gý CU M r- N cc cc

" r+

" d

" co

" " " C! )

" (n

" c-4

" p'

" :'

" -t'

" N

" fl

" M

" N

" (Y)

" N

" Q' CV en M ý' :)

Cn Lf . -I N 0 cc cO N rn o al rn to 0 0 rn N t W N c-1 C) : f) to Ln 0) Cpl 0 in 1.0 cj 4

" " " '

" " " " " " " " " --

s " " -,

" i Q, iN Q

n ci c-1

C C1

N c-i

N c-4

C co Q-. 0 c iý

00 CO c C'{

M c-1

L cam...

y .r

C3 `ý Q ca u W W C) .' H ,. a Z Z 0 (7 O I Cl o

0 m co cO to O cO 0 0 c9 O 0 0 0 0 0 CV r4 N N N CV N N N N N N N N N N

O 1 ý

1 I 1

N i I [ý

I A 1 1 I 4 AA I I 1 0 -1 V-4 [

a-4 (-, j c N N

OD cI N c-I

L c-4

00 TH

lo cI

1 CV

t0 c-4

0ý c-R

M C'!

7

to M. O e''n c d' [N 0 d' to N ti N N N c-i IH cI N c-1 c-4 N

0 N ( N CV

E- I

N "I

t M

V) i Ol

M 4 -1 ' in

-I N pp 1

c- fV ci -{ -A

c 0 -I -I

c 1 N c c c c c c

IT T-4 r- rq TA NP Ln N m vi N c-q N N c-I r1 c-1 I - l CM M 1 ý d N C,

d ý N -I N

I 1

c T-1 IH O T-4

c N c- 4 c-I

c c-i

t c-i s-I

c 0 cI

ß -1 c c-1 r cH c"1 cI c-i rq r-4 c-I ri c-I

N C. rl' 4 9 d' lc\lli d N N ä (N cI c--i v c-l ci c-ý c-{ c-I vi %-I cI r

e-4 tD C) tO in c- c-I M M to W

rq c c W. co (Y) ON ON 0

E/! Cý ti v c c c-i c c cý t U)

N 0 O CY) H N

c-I i c

1 1-+

I co!

Cl) ON 1 A ý CO

ö

L cI

rn cI c c-I

0 % C r c-I cI

N r' C)o OD

O mal' l0 to

v CD t. 0 1 O' d'

tD

in

"' C"' N O' 0' 14)

in In '-0

'- in

"' L') .C cc

s

V' c, to cc co c .4

N el dt CD 11)

.4

i N I

N '

to

N

M

N

d' N M

l0 N

Lf1 4 N

N M t- In

N cc en q tp c-I M M M d' c-i N N d' C11

r' d d d d d d d d d d d d d d d

U Q CO U W Ci, C7 x H ,ý z O Ör

a s

U

w w

75

N E W J

J M CL 0 Q.

J

W Z W CD

N

h-

U- d

ýo

m

0 LL- U; J

W

OZ

W J

v rn c

Z

W

0

N ¢ U)

U

J

U O N

"o a' V 1- ',; t ON O I5 M - 01 CO O', t"', N N O :t 0' M co qt co ; - O u1 M /0 t0 X11 0 M M O

V) M - M M d N U M N M N -Y N M M "t to

Z CC) O M 0\ U\ N N 0 T '. O CO 0 N N %0 ; - Z < In to '. O q; T M U'% %D 0 I- v NT I- 0 I, - cc) < W

rn ýO M O M N - O 00 CO 06 - M

cr- O F-

- N Kl -' < co U W U C) S - -i Z O d O" d O"

Q LL

O M %D ko 0 O 'o O O "D O 'O O O O 0 N - N (N N N N N N N N (N N N N N

O 1 1 1 I 1 I I I 1 1 I I 1 I 1 I 0 - N - N O M O\ M 'o U I- 1- N to co M -

- - N N N - N - - - - N - - - N

%0 -- ()N N CD CN or %0 to - (14 N - N - N - - - N N

0 , 1 p M Iý I 01% to -

O Iý 1ý - M ul ttl ý -

- N - N - N - - - - N

d O' '. O 0 0 00 N M % ý 1 I -O I N r I I N U O a0

t 1 a 0 u - r l - N

r- d' 00 N M - N M %O c- C' Iý N 00 1 I 1 Lr1 I 1 1 I 1 I O - 1 I fý

- '. p M '. O - - -

O - N -

- - M 1-

N Ln d ct O %0 LU W O %0 1 s 1 1 1 1 1 - ON O - I 1 I N I %D 0 U O - d M N - V) OD C' - Ul U V) - - - - - - - - U')

Z z W

N1 O M - O N -

0 1 1 1 N $ CO ON I 0 t-- co I I

cn tr%

co G% cI O 0

- co - co co Oý tý Oý - I O Iý

O -

I O $

l N N CO Iý %0 CO

0'

ý %o Llý ko %ID %. 0 r % 0 M I rn I C ý 1 1 1 u; u'. N I rr)

v ao Ln IT 'n L(''. u-1 ýo

CD N1 to I; r qT 0 WI N M M %O N N "Y I I lq: r I W I (N

W I

N N WI M Ul 'O - ýT M M -7 N M N Q M

O O O Cl) O O O O Ö O 1 O O

fr 0

in U W LL U' S - Z O O' O' O' O' ä W W

76

0 H H

a 0 a

M tn

aD +

0 a N z "r N

Ui) sý

4

Cl) u) Orn Qý Gý rn G) ýn c+ "+ M rn ' cq co " N 'c i C) M [- lD M N O N N c'1 l0 Gl t0 "

" Cl) M W rN cc) Lt) O '. O in l0 d' Co u) in u) co " V) c-ý U)

7 (N O

LO C%

U) O

M tD

01 S

c in

0 [N

in N

d 0

0 C-)

CD (p

u') In

0) N

C" Ql

r- CN

0"1 N z

- 0 t, O) 0) c-I c4 U) O OD N ON M O) 00 O CD N r-' N N M N N (N cI N r-i N c-I ri N N

x x U O O . 0) U W v+ C% H r] 'F. Z 0 O' G C7 0

Ü

0 l0 N N N O N O O N O N O O O (V O T N U) Lo u) ;: Y LI) d' d' Ln C in d' d` d' u) 0

c+i tN 1

O 1

CO 1 A) i

rD 1

O 1

[ý oll Qk 1

ci 1

ci CA 4 M (V d' d' d' C') d r*) M c') N M M M M d'

O c'4 fM 01 N N in N ON l0 00 co O O N M M N d' M d M Kt M N M N 0) M M M

CY\ c) o I 0) I

in .+ ci a ý

cat 1ö oli co r+

.ý °i

N N iP M [}' M V M N M N M N N M

co c-I M M d' cn d' N N M N M N N

O M

O V

OD 'A N

!N M

N

M cä d

O (N

ODI M

00 N

4

N CO N

ui N

M, M

W, N

A

N CIO N CIO M

CO

Ln N

pý c+

l0 cI

v M

O IV

N (N

N M

CN N

M N

[- N

d' N

N M

U) N

V N

N N

N M

4 tI

A t- L

AI N 1-1, to A CIO

M N N

N c'4 r'1 N m fV M N N N rV N N N N M

94

'. D

M N

1 r4 N

to c-I

! U) T-4

N M

I O M

l0 N 1

r') N

to M

1 tV (n

[t' N

1 N N

r M

rV

d' N

1 N (V

0 N

W c-I

d' N

N

N N

1 O N

N CV

N

N N

I O N

c-I N

1 O N

M N

1 c-+ rV

co (Y)

I Ql N

kD

Ö Ö

z N of

0 N

c+

cI C) (V1

N

rV N

N

ca f0

N

ca CV

V-1

in N

C\l

c-+ N

cß-4

N c

t1

c4

(N N

N

O' TA

NI

r+l N

I N

Oi r4

0) t-1

Iri

O N

co

0) (V

N 111

1'

N

T-4 1 ate-

M 1-4

C\l 1

\1 to Iý

Cj (V N

at r-a 1

r- %-4

r- C\l

4 rV CI

0) 'H

i cI

N r4 4 c-C

Ln ci

M s-1

0) c--

r ci*

N ca

rn c-1

M c--

to c-I

I- V4

vli 5-s

to

ri

ci

1 cc-

N

N

t" c+

- cý

t- rv

"o c+

M N

V7 i c

M CV lC r

14' rý

ul ' c

N .+

, +

Cl

t{ C1 c-I

ý' r1

01 H

G ci

cý c-1

ci c-+ c-i c-i

c-I c-4

M

c1 'I' c-+

N

%-i c-+

O r4

O c

M l4

00 r4

N t-I Ol

O c-1

O ri

M 0

N c-i

O c-I

N Vi

O ri

T-i %"-

M

N Qk

o a Ul) ol r! a0 Oli Ell 1-11 C)

C) cy"

N

Tq CD

a CD

a ;

o C)

a cV -I

o at

a '-O

o r-

G ü)

O N- ýI

a i

0 '. O

o 0

0 co

a CO

a 00

a 0 u w a w U w w 0 x H z o ä C\j a ä U w

77

o ca a+ c'+

ýd z ZGcý w Cý

M

VH! ZN

H d' Hw ý'O

4W f. ri

CO 10 in 0 N M OD 0 Q' 0 0 M C\j Oh CO in " Oý M L') O 0 N N 1O O O 1 O N 10 d' Ll " "

ul L4 0 1O cl; 0. l to O

c-1 10 to 10 4

CD Lo Ln un CJ " Cl)

Z al CO ": 4" N O O) d' W d 0 lz: i' N ci N

N Z

z O N

d' %-i

Ol N

N N

N N

d N

ct N

c N

10 c-i

N C\j

O N

N N r4

O N

0 N

10 N

r Z

O Ö r-4 C44

Ü co ü W W C7 x H a Z O d G Ü w W

O l0 N

N in

N to

N in

O o

N in

O V'

O N in

0 d

N in

O V

O di

O d'

N to 0

VI N

cn

N

(1i

A

d' O 'C'

A

d'

A

0 u, d'

4

0

9

N

9

(+) O M

L5 (M)

O (. )

A

m

4

Rl

A

q

T,

r c-I N Ql d' C' [' M to l0 OA 00 01 N O (V O N d' M q' M % M N 0 N O N cM (y)

0 I N

1

N C

I

Cl)

,

d'

.

Cl)

.

d'

1

n1

. N

dI

rM a)

(

N 1

c ar)

1 N

. N

1

O

1 M

N 4

M O

%I 0 r4 C\J d' r4 co 'I N N-4 cO L" N

COO) N N M M

CID N N

[N Cl)

1-1, Cl)

r! M

O

Cl) tN Cr1

CID CV

C) N

j

CV N c1 c)

4

C\j kÖ N

Ö

M 111 Cn

co

iO ýp 10 O 1O O\ l0 cN N CO to O M Ui Q' U1 N c-I M r1 M CV M N N N N C') N Ctr l N CO

C- } 4 'A

N [t cä

! tD

8 1 4 dl 14

t M

N r- N cI M N C7 N M N N N N N N N N M

0 N

to c4

C\j M

in N

v4 c')

tO N

O ('')

in N

0' c'4 (V

cn cm

l0 N

O N

N N

tll N

r M

W rý N

in 1

cI

Ö C)

N N

co 1 N

-4 N

Ln, N

N (V

C1 v4

N N

ci C11

M N c4

r1 N N

CIO N

4Z W

O

0 N

d' ci

0) N

ca N N (4

ce) N

d' CV

c-I N

0 0 c'4 N "O' N

N N

[N H

O N

n') N

tN N

O

H M

In I N c-1

4

N N c1

4 N

c! 1 H

IL

ejl N

A

Cl)

, qr H 4

c+

ci N

cý

N 4

N to t--

N 8

(V

N

OA v4

v4

U1 cý

ci

to cý

ri N

cý

11-4

l-, cý

s-I

l- cý c-I

r4 4

cý

c--ß

to rý

N

cý

N

pp1 cý

d'

M cI

vi c-I

C-) N

c}' ci

Ol co ci

C' cI

rY c4

-! c+

to ^{

ID c-i

M T-1

M c-1

C. r-i

- c-{

N V-1

M 1

CI

8

c-I

"1

N

N

cl It Ca

NI ri

d1 11-11 cl

Ö

c'

Ü1

ci Ill ri

14 ti

cý c-I

N

c-I c4 C4 c-I

0

0 O'

0 C\-4 4 Ln

c-i w OD

C) H Q1 cl

cl r' ý c'r, -l

in c-i

(n cl N

cl L" r4

all 11, Irl

L A I

j ýN

C\ll T--l

010 OK Ö VA r4

Ö N

d co

0,

ý

ö co

Cl)

0

ö cl

d in

0 00

d 1o

ö cl

0 cc -I

d iN

ö co

ö 0)

d cl

d ON

d

R

uý I P'. -

w u W w 0 x H Z O ä cm ä y Q

4,

78

z 0 H E1

I ä °+4

cv

ýwa

xä n' HgM

W t7 ýä `ý w

to e C) ul c% N (N G' C fh N lD N 00 M 10 " lL M r-ý N ý"ý c") N cP Q`; O L7 M In ü) lD L)

" oZ

Ln cI r N- Co un O ci

to d' -D «"ý to d' to to rn U)

c r, N-

t0 C5

M O

L- N

" l0

r CO

v. tD

tS c-I

e N

N Op

Co N

w N

N C)

C% Ol

N 00 2

Oý i

a1 N if ) U) l! 1 ca ýA N c1 N lD O Cý Ui c-i _f N N [V CV CV N N N N r4 N N N , t,

O O Ü

Q 03 U W W C7 x H ,ý 2 O O

0 0 d

U 4 ý

O lD N N N O N O O N O N O O O N O N in LO If) d' to d Ln d d' LI O a

c ci M O f1 +

C' O ci

i

Ol lý N

tÖ u! M N ý' d' d' M ýi' M N m M M N M cal

M N V' Cl d

M e M N M

OCM co N M CAM de'

1 a) N

I U) (+)

1 O M

1 N

I M

I N

I () M

I 01 N

1 l0 M

1 Lf) (N

I 01 N

I (V,

I c-i

C

I1 Ol vi d' pp cI c-I O C'1 N U) If) d' Co M O N c ý' M cý1 M CP ti cm m N ei cm N M d'

(N Co M

Ö

(I) M 31 M

(X) M

ßl

N c-, A N

J

N N N

c-li M

1

N N N

ci

en

L

M

Co

U) ap N O' cM O 5 0) O CO '. O O O t0 O u'1 CM c-I M N rn m (I) N N N N rl N N M M M

[ý Oý O1 lÖ U1 t' C1 1 4 J

cý N

N c-) M N N N rI) N N N (N c-+ N N M

N N

tD vi

M M

LI N

M N

CO N

N cn

U) N

W cI

U) N

d' N

l0 N

ýp c1

M N

CO N

O M

w

Ö

N t11 c')

Ö

f") N lÖ N

li N

t- (N

N

N H

( A

N

r lý N N

tý t-f

c14 N

lb (V

tD N

10 w

O x

o ý pý , -i c-)

pý N

i N N

ýp N

tp N

c-) N

l0 N N

N N

N N

l0 c-I

O N

Lt) N

u) N z

H U1

tn [ý cý

tý N c)

Vi N

t M N

Q1 c)

0% cý

d e

Ö N

(D1 N

1 CE) r

LA c-) + N

N N

to W H U1

l0 ci

c1

ý1 i

cý

tp (N

M N

N i

Ln cl

O N

[ý c-)

N N

N

pp c-)

%-4

CD v4

Ln %-f

tý! N-1

1 N rI

Q) X-4

1

c-)

OM i

[ vi

N r) 4

ri

cam-) N

rH

ca 4

rl

N

Ol cI

N

c-)

(I r

r+ r+

N N

-e r{

%t`. cl

O N

cn c+

r" c-ý

, c-a

I. ̂ c-;

Lr: ; -t

cc- V-+

CJ %

ýn r`

in c'

Ln c-)

1

M c1

Ö Ö N V' OÖ c4

cI

N

r'{

O If) %c-4

In, c-)

Ö

c-1 c' c-i

c

c-i t0 c')

CV c"1

M

O c1 vi

cn N- O vl

vi ll 01%

d' -r-1

V' IM 0)

O O U) r+ N

41 %-4

011 tý v4

U1 Iri

Ql W M

K-4 t [ý Oý 01 t`

N

0) 1

0

Co 1

O

tf c) 1

O

W I

O

O ci

I O

ct' c-) I

O

vi I

O

CO 1

a

N I

O

N c, 1

C)

(N r,

1 O

l0

O I

00

O 03 I O

M

I O I

O c-+

4 ¢ ra v w w x H a z o ä a ö a Q >ý w

79

}

EE+

7- 0 H H

0 Cý +

4

O w

a

w "n w

N co

tl z

d' in

a

0

LýJ

Q') O Cl CO N N M V' IN in C )'N c-I M O [N CO " ` ýý N N Oý r, O m W N c-t in M O + Q . . " " " . " " " " " " " " " " []

to t"7 N N C- Sr O 10 10 In d" co Sr tri to (7) N a

z c-I ci

C m

CO in

[i' m

Co in

Sr CO

co in

N c-t

Co o

l0 N

a) Co

c-i co

CO in

L to

N t--

c") O Z

ß1 d' CO N

ca CV

Co N

LO N

in N

cI N

l0 %-I

N N

N N

N N

l0 c{

Cl t-I

I) N

l0 ("4

E-4 Q ¢ m v w w cý x H a z o O O c ä v

a w

0 0 Sr

c0 C\j

N in

N U"1

C\j in

O Sr '

N to

O O d'

N in

O Sr

N in

O d'

Sr O Sr

N U") 0

cA N

(n ci 10 I M

CO 4 A c,

LN N

lD u7 .

M N Sr Co C}' Co "'i Co N Co Co V N c')1 Co d-

N cli C\j C) N 1-4 M am N Sr CO

Sr (') V' M N Co C" ) Czt Co ' d

Oý 1 Co

I

Sr (A Co 1ý Co CA Sr

1 Co

1 N

I

Co 1

Co Co

I

N

i

N

Io

Co

1

Sr

N Ol 01 CV U) Sr N O Sr 0 a Lt') N co m Ol lV c-4 Co Co M M Sr Co N Co N M N N M M

O In Ql N (311

N 'A N c)

r! W Co

N 'A IL C\13 tD

N (0Y N Co Co M N N N N C') N N Co M

Sr N 19

w M

CO N

c Co

(N Co

t0 Co

IN (N

0 N

LI- (N

N (V

c- M

0 N

to N

cI Co

t. 1 Cl)

Co N

rý c-I

N

CO LA N

N

N

I

M 'A Co

to N

Oý N

1 cb N

00 N

(11 cI

'A N

Öý

(V cä Co

W ö

N N

I

tV

Wp

to r1

N Co A

N

Sr N

T4 N

tN N C'

N

Qý N

co N

N M1

(0 N

Sr N N

N

00 c-i

to c-'

b' N

M N

to N N

N

[N N N

N

Co c-I

t- 'r1

N lV Ö

N

Cp N I

N

O Mt

lQ N

Co Ö

. bi m Z 6 01

\-4 Sr H

CO (V

O N

c') N

N C\j

U) N

c-4 N

Ul T-4

N N

N N

N N

W ci

CO ci

to N

in N

Ü

, Ö I ý LI &'I t! j

tD 1-4 c'4

10 N

0ý c-I

Ö

N ýi N (V

Ca 'r ca N N

ýp ci

LA ci

l ri cm

N

N 0

Sr in

T-1

N

1 T4

ßt1

1 N

N

! T-i

a

1 V-1

Sr N

N

Vq I

c-i

N T-1 I

c

M T-4

I tr

Ol ri

ýp1 cr

O CM

1

N j

c-I

Sr `i

ci

Lb V -f i

c+

en N

N

c-I N

tl T-4 -1

rn O Ti

ci Ni c4

ýn ri

cv N

ý' c'

C T-i ca

N \-+

N c- +

z r

(' 7-4

Y t-3

O CV

y ra

M O

cI co ca

14 c4

N

cI

Ö

CV O cl

N

c; Ok

4

c1

( l0 r4

N

cI tI 1-4 I

N

cº rI

N

c-I

M

N CO

CO

ON

CO

N c-S 14

V-4

O 'ri

col

c-S 4

0O

O) c+

T-4

C) i

0

Co M cI

u) ci

c-i i

CO

O Cý

%-I ;

0

N ý .4

N ý

pk N

T-4 N 1

o

N t

o

fn T -4 N 1

0

[ 1

0 V4 c

8 in 1

0

CO 1

o

in

O1

N c-I 1

o -4 cc I

0 CO

0I

01 8

M f

0

V) c4

1 0

CO i

0 -4

ö g Uý Q as v w w C7 x H a z o

ä ä

Öý ä o

W 4

80

tJ -V O VýI' N cß') : -+ 01 J' N u() (IN ßl c+) " CQ 'D O O O Cl M 0

. fl M M O ri (Y) C Gl "

" " " " . " " " " " " " " " " " D " U, M t- t- S

4 O tD V ' tf 4 N d U, t7 C) "

U; ci y

LU LO CO co N lo v, . ýý a ßl lD O Ui (V Cl tUi Ol CO N co lfl co 111 F..

co N C1 O O N 14 i a-ý U1 c-: l7 c4 c-{ c-+ N N N N N CV r1 N N N c-i N N N

er) o 4 C1 U w iL, t7 X H +-a Z Z 0

a Ör a

a U

w w

)

)

N N

w

0 H H

ao

Ö v'

sa z

in a 2ö

aw 4

O W N N N O N O 'O N O N O O O C\j 0 41 N u1 U! ) tf) d' I d' d In dl to d' d' '' U) O

Ö O CO

I M t W [-

M N d' M M c*1 d' c'") N M M M (V M c*1 C

O1, 01 rn N tf lO N I'D Cl) N to In (V Q0 O N 'r d' M M U C U`1 N M C*) C) N M re)

0' 1 pp N

1 ci d'

1N ef)

NI M

1N 4' cI{

r''1

4 N

I c. M

1 O rM

1 d' cn

I M N

1 O rn

1 uý r

1 n m

N 0 ci c-1 d a-4 O U O ON C') N O') 'T t0 N lq M M M d' (In N U (V c+) N C\j M M

CO N ll,

C rt CID

N [N (n

N Ö Ö N

N N

N T-1 M N (V U M N N N M N N M M

V to N l0 N c. 4 '. 0 [5 O t" CO ON 01 l0 c-4 N N ri m N N cy) M N N CV CV (N N cn V)

N N to1 d'

4 O1 O1 Ln! ollI

1ý I r--

' 1 1 1 1 N

N 0 M N N M M C\1 r-' N N N ý N

m N

N d' (q en f'7 Qs Ol d' 0) ¶' l0 Ln N ;; N 01 CO N ri M N N (v N N ¶4 N N N T-1 (V C\j N

W

ID ' Ö

N c1 t"4

O U

O N N

a) N

to N

(1) N c-4

N

C\j N N c4

N

(V r-, N

l1 N

ILD

O

0l c+

lV ci

01 N

01 ¶

0 CV

N CV

d' N

cV (4

l0 ri

-4 N

U) N c4 C\j

to r4

c. 1 N

0 (V

q CJ

ö

CJ1

In N

V4 T4 "

Lot N

[N t. 4

OÖ ¶4 N

Ö

N Q1 c,

Lf1 c-4

Ö

N

N

N N ['H

4

¶'l Os c-4

Ul N

01 ¶4

N In

10

t0 ¶4 N

c-1

O H

to N N

N

l0 c. 4 4

at

N ¶4

,A ¶

U') N N

N

C') ¶

c-,

CO ¶4 I

'CA

¶7' c'4 N

c'{

O') c-4

r-, ¶

c-+ N

O N

t0 t-4 A

c-1

M c4 N

TA

co ¶4

[l, cH

d' N N

N

00 c'4 4

c-i

(11; N C') 'z? (\: C' L) 'c kO (,. N Vl N Cr)

U ¶4

'

V4

Gl 1

CO

N

1 N

¶ 1

VA

¶

1 c'\

1

r4

¶4 1

-

c-{

c-I

^{ 1

'

c4 (

cl

cr !

¶

¶

I %

" c-i 1

ý

N 1

ci

¶4

I C''1

N O') NI

l5 c¶-4 NI

O

Ia)

O

CIO

ca IN

co 1CO

c-4 1O

O')

CIO

¶4 1N

cl

U1 ')

01 I N

ý

aIl

%-4 1

¶4

H

1 m ! N

%-4 0 t in

i O

'- c4

1 O

N 1

O

N 1

O

¶

1 O

N 1

O

m 1

O

NN 1

O

T-i c. 4

I O

14. ¶4

I O

ko I

O CO

1 O

O r{

I O

t0 ¶4

1 O

l0 1

O r1l

aO ä

w Q as u w w 0) x H a Z z o GF Cy (y c( Ü

w

81

00

4

Hw NHA

L Al

" to C)

co N

r- ll

C1 cH

LC) N

d' M

CG QD

It '

r1 %

N (V

0) co

to rH

O : f)

cc) M

w LID

to L, "

a in M N co C' to O N if) tC d' CO Ln 10 10 C)

Z 0) IN

tD N

cH cc)

CO CC

c-1 N

N d

C' d'

lC CD

M r-I

N cH

qll 0)

M O

d' O

c-H "

7-1 O

CC) O 2

" " " " " " " " " " " " " " " " Q' %D O N to Lo N N LC in 4 H ýp O If ((I

6

cH TA M N N N N N ci N N N cH N N N

E-4 T-4 C\j cn U Q U W G Cý x H i4 Z 0 C% C7 Cr

CI U

O to N N N O N O O N O N O O O N O q N In in ul -' in d' ,: i' u') d' Ln ". N d V in O

co (f) N d' d' Q' M

w M M M

M M o M

o am C\j c '1 N d' d' d' ß'r1 [N M CV ()) c) N M

cN m v t A o 1 cl, 01

N N

N

t}' [ M

N

Iq

4

q; r M t N

lÖ M

O co

N

M 00 (V

1 4 Cr)

4

M

a

c}'

N N d' m V M d' M N M

N M N

N M C-)

co rt Ck

A tÖ

N 0 'A (V

N r-I

O N N N 11

CO N C') M r7 M d' M (V Cl) N M N CV CV) M

to Q1 (ý lV to N 0) O cH ý to 0) ; 1' to v-4 in N H 0 M M 0 M M N M N N N N M Cr)

N rn cl N

I

C"1

o

M ui c4

0I N

OI c'H

all N

1I N

11 N

ci N

4

N

a

c*) 1- 3 W

H

V1

lý

t'+1 N

O) N

N c'H

to c' ,

(4)

t0 N N

N

C' N 4

fV

O, N A

N

w M

W N

r- N 4

N

CO ý

It cH

CO CV

u1 N

g N N

N

Ul N

cli N

O N

1 ý

(. 0 N

ci N

0) N

N

O M

lÖ N

t0 Ef

C1 -$

to ci

O m

H N

M N

tD N

N N

C') N

3. O cH

";: f' cV

N N

O N

N rH

O (V

If1 N

U) N

N

ci P-4 N

ýR C\1 cm N

N N

ýv. N N H

ý

H

ý

(ýV

N tH

4 eH

'4' rq A cH

N N

4 N

[N cH

tÄ cH

O N

r-I c-H

(V N

d r4

O N

[, cH

01 r'

If1 TA

M cH

1-1, cH

" N

C1 c'H

Cl cH

ap cH

l0 H

A cH

to V-1

A rH

to rq

4 rq

N N

ý rH

O cV

IL cH

(4) cH

N sH

C4: N

-I t-H

to !H

Q1 c+

t0 c-{

C' -

C) cH

00 'H

N T-

N rH

fV tH

M c-H

N r{

to ri

M

T4 cý H

J.

N

N

c-H 'A ca

U1 r4

N

r4

N

rH 01 Cý

cH c-{

a

c-H ci cH

r+

r-I cH

ca

to

m

N O sH

O eH

O N

rl' ca cH d,

N c+

CA

d' r'

1 T"H 't-4

1 00

c'H vi

1 O%

co n

to `

un C, t`

O e-1+ OS

O

! C)

c"1

I

O

U) N

ý N

a ON

a s'-H

a W

a GO

a r'-H

a N

a co

a 3. O

a T-4

a T-1

a to

a co

a l-

a 'H

a '

a ri 4 4 to V W W C9 x H ri Z O Cy O (Y aU

Ü

w

aý

In H

hH

A

f1 CD N L') Co to o1 m N N N N L7 lD cn c-1 " N c{ r- C' C' C% Ol C) 6' c-+ m - d p1 cal 0 "

" tD M tJ N ß'. in O ýS t0 in CC in tf) tO 01 " N in

Z r4 G1

M ý7

0 ['N N

N tO

c-4 N a) M

cl if) N

ire ýp

d' N

0 tO

C' O

cl O

c1 (`'1

N N Z

I Ch Ln 0 M

N N

N N

1 N

l0 N

c'ý N

to c'

M N

vi N

vi N

CO cl

O CV

Hfl

o Ö

O ¢ U =ý w c7 x E-+ a L Z 0 vt ä (M c Cl

a

o O d'

O N

N in

N cn

N in

O e

N to

O O 2 N in

O -r

N in

O *

O e

0 'cr

N in o

N 1

N 1

t7 {

Ol t

l`ý 101 t

lý I

to I

O I

r-- 1

N 1

ßl 1

O N l

tt) M N d' M M Ci' fn CV M M m M M M d'

c-ý c N Co to fit' tJ C [` ýG c Co Cý cV d' (n r) N d'

c+ C')

tC

V' '(r) L'" C") N

N

il)

! 'n N

M Ö

M

t)

N

CO

M

Q1

Cl'. m

[t' o 1111

N d' C7 cJ' M d' t) N M (+l Cl) N N cM d'

N N d(D ' Ln ri c114 .4 (M ('I) ti -e M

N M

N M N

Co M m olý

l4 N

Q'. ti

Cý r7

cl M

mI M

14 (r)

CO M

j

(Ni

ý ,

C'h tý N

Ö

M to N N

cý (n

t11 M

Co

u) N

ý) c1

N Cl

o C

N Cl

O tM

IN m

co N

O N

ü1 N

to N

o , N

d' N

un N

O M

[r Cl

r 4

N d Cl

C-' N

ý'

M

J

N

A

Cl N Ci ra

N N

uj N

to N

d N

C

N 00 N

c m

Cl N

N Cl to N

m Cl)

Co N

N Cl

in N

W cl

tO N

«e N in

N O N

Cl N

tN N

p Cl

(rj

3

(V t c-ý

d

fl) (N fV

a) N

to

N CDI N

N

N W cl

14 CV

N

N

N

N CCI) di

c+ N

Ul N

tD N W

C1 VI

to

O cm

N rl

ºt1 cl

O Cl 1

Cý N

ca N

1 di

t- N

1 N N

trl N

N

r- N I

-, CM

N N

cl c-a

ifl vi

rl

Cl (V

f. "

N

c-i N 1

c ci

vi N

1 N IM

P t-l

Vi

O cm i

Co -I

d' N I

N N

ýn N

N N

N

[-t U)

1r4

tý-ý

cc-i

rýº

N

N

N-4

Uli

N t-' di

N ýý

N

%. n-l

di

ýý

di N

c

c

1101 di

Lný

T-4

t- c1 r4

TA N

1

vý N ße-1

a

r; %-

N S.

(') N

C' VA

l! ) v+

O'' X_<

C' n

' c-

r' c-.

" c'

t-^, -'

C') v-i

N N^

^1 ci

[N t--)

it'. c-

<'ý 1 ! ! 1 I i I 1 1 1 I M cý cý

N N c-i c-I

to c'I a-i

ti d

Q di

to ri

C c-

O %-i

ci ci in c-1

(4 TH

Cn Q N

'rVi 4 ri c tai Oý Cl

r %--t (A

ý 0 d 'C I

N 1 to

1 G'.

1 N

1 O

1 C'

1 N

1 tý

1 O'.

1 [

i N

{ N 1 [ý

7 CA

1 1 ß

1 O Oý

N

t- OD

C1

m

01

t-

01

OD

O Ö

(D 0

OD 01

0 ä

w

01 cc) 01

cc) 01

c% Ö

Co 0

M.

4 t10 t! W G, c7 x H a Z H O Ö'

a Cl O U

yy

83

x 0 H H

R

5 N

o1

HW s1

w a 4 w

4

a N

Z

N U N

qV)

U a

8 U)

to t- c4 u) N M a lo In N N N lD v'1 tO "0 " N c-{ dl if) M ' M O N lD N t- to `Z3' Co "

" V)

In M %D t- Co uo O ci

O tiro U1 d GJ d ul in Cp ti

2 Ol r:

CI 0

N mot'

'j' 0

%0 to

(Ni in

O O

%. D (''l

Ol O-

M v'

M in

0 t0

N d'

N CU

O CU

M Z

a Ol

j c? N to d' in O '. D 4 4

N N cl th '. O cý N N N N N N C\j VI N N N c-I V-1 N N Z

O o

w < rý o w w C) x H a Z z O (Y ä c! ä Q

w

O O d'

l0 N

N Hfl

N u-)

N ul

0 d'

N U)

O V'

O d'

N If)

O cl'

N to

0 cl'

O d'

O d'

N tr) O

Irl ri Qý ni

N üý ci

ül c, 4

C') N d' M cj' M d' M N M M M N M M 92

U) N d' d' C') 'd' m N c'

UM Co Q M -e

OD CV

0% [T d'

to M

O d'

A

M c d'

c M

tf, N

O M

0'. N M

to N

Ci N

(N M

O

c'

O

N ý1

in N

ýp ri

O

1 N c'1

d' C'1

1 o M

QA M

M

N M

i M

O d'

I M

O) N

I N

d' N

N

ON N

I N

C) N

1 N

ý C')

IN U)

d' N

N

CO N

N

¶ c+1

1o M

0) M

1w M

00

d' N

N c-I

to c'

0% tV

Ln cM

Q) CV

lD M

tp N

N N

l0 N

l0 N

c-I M

N CV

l[) N

OA N

in M

r- 11 IL 1-1 a0

N 14

Ö 4 t1

Ö A t,

N t- N c-I M N U) N M N N N N N N N N ('))

w

w

N N

O (N

Ln c"1

¶4

N M Ö

M

LA N N

N

O cn g

N

N N i

N

ri M

X11 N

I'r) (V

cä fV

C) c1

ll cI

M Cu

c+ N

d' N N

N

tN N

M N

01 c-1

CZi vi

N N

O N

l0 N

N

VI M

tý N

to w

Ö O bi

Z 1-4 VI

in

() ci N

t-I

cM vi

O) N N N

c-+ N ý

c-1

l0 N

N

d' N

1M N

e N tr

I

c-,

0 N

1 O c-I

O C1 N1 V

O N 01

1

ci

ci N 1O

N

N N Ol

1 VI

N c-' 111

I1

ci

0 c-4 [Iý c-

M CV N1 N

lp cV

I M N

ýn

2

"' Ei H U)

to VI

14 Im

cy c4

14 -I

to N N

t1 cm N

cm N cm

ýp ti

vi

N

ý

"'

tI

CO

cI

C) N

vi

W cn ý

V' ci c1

tD c-ý

c-I

ý"i N

c-C

(V N

19

c- ¶4

0 ¶4

N N

V' vl

N c-+

O-, ci

C' ¶4

th -

v-, : -4

tD ¶

'. 0 M N -e ci

r c4

r %-i

M

¶4 0 N

N

t', Ifs r-' s'-

o

r4 l' r4

O vi

4

c-1

"

%-4 li

IM

! tp Ti

N

c-I

M

N C1

c

0%

C&

c

v4

V-4 j

c-I H (3)

IN C)

O)

010

¶ r+ (N

9

H

Irl

Ö

%-4

c1

4 Ti j

`l Nd [S

a rd

CO

d d Ti N

o tO

a aD

a S

a a ý

ý t0

a Q)

ý 0)

ý r4

a Co

a ý

w 4 co u w w C7 x H a Z 0 ör ör 0 ä

0 w

84

CD

z 0 E-+

4 0'

F

m

4Q

u

" N v

L') 1-7

d' n

n 00

UO r,

c-1 )

N l0

t" %

O N

O O

a rl

0% CO

t7 0 c-+

C to

UD Cý "

" LIl M ko to CO CC Ul r! ' cn d' N V' I I) t1) CJ "

;z N 0

C0 n

N 1p

0 cl

O LI)

M to

M ! fl

M N

'r-4 U)

(N n

t0 a'

V N

a' M

M M

co N

n Ln Z

Cl N n u'l G" C) ri n O c-c p` n 0 M n G7 .Z c'c T- N N (N N N N c-c N N N c-i N N N Z

O Q

s G+ U W G., c7 x H a z O ä ä C!

CI , zp

0 c4

0 c' I

O

l0 N

1 O

N u) A

N to 1

q

N U) 4

O a f

to

N to

LA

O ,*

14

O cl'

CIO

N in

c-1q

0 d'

1- I

N in 1

0)

O V t

co

O V' i

O

O ;r

u)1

N cn O

M N d' M C C) -z; ' M N M M M N M M "; r

ON ON N n M c! ' cl' 0 S O O CO E-- C) cl' M N ri d' M d' N) ct' M N M M M N N M d'

a i N

f p ppI

M

1

M

i PD I

cl

I N

I

N cl% N

1

N

10

c*1 Ull)

I

N

I N

1

M

1

v

ý 0

n N

N a-i

N C")

(N M

a' M

N M

ON M

CO N

d' N

0) N

co N

tf1 r')

'q N

n N

N ('C1

O Cº

co It N M CIO N .:

f, n C) -A C') LIn M

! ý N N

I co N n N

A cM NI CV

1ý N oý

1 N M

o0

E-

cf) N

'A CV

lD cI

ea

M M

14 M

n N

t1 N

d' (C) Ö

M

0 M

CIO N

d' M

CO N

l0 N

L) N

N tV

01 N

L17 Nt

V' (V

lfl fV

to N

CV (C) Ö1

CV

rl CV

O N

In N 4

N

CO N

l-I N

to CM N

r')

r,

4.0 N (4

1 V-1 N

'd' c-+

'A T-4

O M 0

CV

d' N

14 N

0) (V

t) (V

n N

i) CV

n N

Cl N

V' N N

(N

Q\ v-I

W 10

M N

I'll CV

d' N N

N

CO N

to N

O' cI W

c-I

M N

ci N

kp CV ý,

N

Cy M

1 N

lD 94

j 4

0 i

H U)

Ln

0 "

I CO c'c

(N cI

N N

I to CV

O CV

I CO V4

'' CV

I t-ý C4

d' N

I N N

N N

1 01 t-i

r+ N

1 0ý ci

n c-I

I cf) ci

O N

i 0) cý

ri N

I O N

d' N

1 I

(N

n t-c '

cI

O fV

ml v-I

CV) N

1 '4 N

r- N

(v to

[H /)

N cI

ti c

c a-I Ö

t-c

d' (V

ci N

n ci

to r4

O N

QO c2

C \l \l O N

a) c

t+ d'

40I c -

ýÖ c-c

d'

t-i

0ciQ

cc

c0)ý '

Cn t

Rl

'4

U) rc

ci ri

cODi -

tp c

O N

I

ý-1

(C) N

Oi c1

c'" C) N t- 0'. (-', c:: r" L) h CO

T-4 '{

'.

'4 Ti 1) ci 1 r1

r4 'A

c"c r4

M

c-i c+ n

n c4

N ci

(") c-+

l0 c-+ Ch

N c4

O ci

M 'H

LO c-+

(N c-I N H

t! 1 c-I

N c-+

N CN ti

U) O o

.4 1 Ö ck

.4 . 31,

Ö p Ok

N

CO 8

u1 0

c

8 0

0

01

01 8

u1

01

O

01 CO 01 01 01

OD

01

0)

01

01

01

ci

01

() 0

. W [i, 0 w H .7 Z 0 O Oc C7 C7 U W

CQý+

85

APPENDIX7

DESCRIPTIVESTATISTICS:

PRIMARYSOURCETRAITSBYSOCIALCLASS ------- - ---- ----- ----- -----

86

N

[ýs, ]3

H

Cl)

NU

U)

U)

U

C4

i4

0 w

U Q W U W f. C7 x H I 2 O

C1 O G G 4 t:.

" rl t- ýn '. O 0 u ý` n co q CO Cý d' cý V) 0 tD cV N N O N ci N M

0 N N M 4 4

M In N M M 17 M f7 14) N 4.

c+ä N

U 0 CV CO (V ' fV 0) '. G (V CO lD l. 7 N 0) lD N

Z 2 cM tD N -4 CO Lfl (fl Q\ N N CO N th m CO CG v Q " " " " . " " " " " " " " " " "

zC3i -i Sý M O O\ M ¶ N 0p c' N N N N N ? ri N ci N r1 N cý N N N c-i c-+

" N 0% N N N N Co 0 w Lfl [i '. 0 d' N a\ LO A to 01 d` N N r Zo ul d' N O CO N W O' ri V) N r: M M d' M V' N M M M Cn N 14 N '4'

N A

u l5') U) 14) 0 t0 OJ d' 15) O N N Ol N T-4 CAD In

2 2 N V-4 CV t- O CO O" O N CO O LD tN In t5 0 ,

0

c t0 C Y) N r O M 00 O N M co c-1 c-a to VA N N cý N N ri N N N VA N

" N p In Q N H N d' co to a% cq W to 1. ) c-4 A N O W o'. W N ci [N If) rn c. - 01% 0 cal (A In n " " " " " " " " " " . " " o " " " " N V! N N M M 'd' M ifl N m m M Cn M N

4

N 01 M

O p un cn

r, M N

r+ 0 m

N c)

O 0

00 O

lD )

Orn N to N CO .1 N; r ' .i

ap

O l0 M

N

Cl N

c

c c M

(3

N c1 4

CO

N

"

pp T-1

l0

N

V

V z N t N c1 c-I N ct c-+ 5-I s- N ca

(5) N N c*1 d' to 1, O 0 0 N to N r+1 c-4 CO A co 01 O co M In N ON (n N N O O U) N ko In N ca ý4' ch U) M Lo N N c4 M Ct M M M CJ'

c5) V4 N

d eN M d' 0% O co O O s-I N cn r N 'N rn tO N M l0 t- O o' c-' kD M rt M rn CN d' O O Z s + ' - 4 i " " " " " " " i c N '4

Ö to r cr 9 N N ý

T-1 N ý-i T14 c-i N -1 T c--i r4 11-4 -4 c-i

" tý dý N s-i c-I d' in 01 N CO CO C") t0 00 0 A 0 c+ co tD r+ 0) 1. D i0 m o N O M O U) In M N M d' N M u1 N C4

4 M M C7 M U)

CQ

Cp ri cý g l0 c-

h c-1

N rn

l! T-I

) Op ý co z z O r O r in aA O c\t 4 4 14 r: t4 c! C4 C 1; 1ý C4 4 r: ; C4 14 V4 T-1 V4 T4 ri r4 T-4 T-4 Tq c-4 N N

" cv G^. Q) C' co ci 0'. C -, V cý cv t- r A r, rn N a c,: C t-0 Lrl T c-+ ca V n ü

c "+ N CA M 4 C4 In 11 M ßf1 14 M <h cal C") C') M M

4

4 in

A N 0

2 c-i to a) M co to in ;: r CV ' d In t. 0 V' 0'. U) " " . " " " " " " " " " " " "

T-4 N V' G1 CV fV fM M N U) ri 4

rz s-1 e4 r4 t4 c'l N vi s-I ri r4 ri s-i c"1

9 t oo v w w 0 x H a X Z O .a en N: r a cl

87

CV)

E-+

U,

U

HUý

4I U,

H

U) W

w

A

a

VA

En

° Q U) U w ct. 0 x H a z o ä Ncl ý

w

" o

ci to

t! 1 0 m

O c-i

(n tO

V' in

S7' Co

0 W

"d' M

(V l0

00 CO

M U7

kc) t0

cA M

c+ h

to r

M V) N N M t1' ;r n1 L? CV CV 14 (V 4

N M N c

fl

Z z C) M (V

M 01

h 0

O h

r' Hfl

<J h

UO Op

01 h

In CO d' co

LO co

CO to

(0 O,

lO "

O ýLD N 01 %-q M c c op (n c+ CV N o) ('3 V

U') to c' C. N M Ol U) O O O co M h to cl,

n A s-4 Qi to V" fV O O\ to 10 d' t0 N 1.0 OD M N (V " " " " " " " " " " " " " " " a " t-{ N M ý-i co co d' M 4' N N M N CV N m V N

A

z z

ý

Q h

O s-'

ltj d'

tp

(13 dl

N 0

h M

Co tl1

Co t'4

0 lO

cv c1

01 c

O c-;

l0 M

O "

W u')

C;

tV d'

Oi

M M

O c'4

co C')

rM c-I

N Oi

h

CO N

co

Cep M

p c-I

M

in C1

A c{ O

f- CO

h m

0 CO

0 cC

N OD

O\ lD

c"S h

N to

to M

Ci in

c0 h

r4 co

OD Ol

c0 M

(1 C3'

o 1.0 U) M c-I M M d' N t r; N M N

4 N CV M d'

M N U

Z Z N N

O

'd' c4

h

tV Cep

M "

tV t0

Ok

'0 CV

11 vi

M (0

N

c4 c,

c v

%D v4

O N

M" W

ca t! 1

"

c-4 u'1

"

M Ol

"

d' 0

co "

M N

" Mcc) w

c' 0

"

ri

CO (D

"

-. 1

ý. "

A t0 C

QO %D

(V U)

N un

to 01

0 co

h 00

N 0 h 00

Om 1.0 H h

h . " " " " o . " " " " " " " . " a r1 (! ý M c-4 M m d' N (V N M N ý3 lV M C) Ci N

cý U If

z z i) : ý.,

(7\

u, A c+

t4 a) ýo tý

h to s3' c

CO . fl of

to

%41 3 c+

oN Q o cri c+

N

co c-4

0 cv . -4

ttO N

d OO

N

" OT-4 ý' q

"

'. 0 to

" aD

O

" aC)

C) inn "

cc-I

cn c+ "

r-I

" q

CO p')

In 0

l0 c-{

s-I h

Q') N

It) O

(V -Ci

CV CD

l0 O

CM O op

'T' ý om' V O

N

cri N

" U) co

" 4" V'

" 4" d'

" " M

" lO

" N CV

. " " ci'

" M" M"

" d'

"

eI

Z v

Z h C0

" O

th CV

" oD

d' t-I N

cý

O (0

c

Ln h

ý, ý c+

co Ol

N cý

W h

M cý

O h

N c+

- h

z h

to d'

r 4

c+

-P 01

" v4

rn co

"

ra

h M

" a%

d' N

" °c)

o N

"

. -I

o to

" v4

" U)

W CO

CO to

w ü1

h kp

CO CN

N d'

c'1 ß')

CX) -1

E- 1.0

N (0

co GO

l'- lD

Ci N

N M

CV r:

o tO

C\ U) M rl M M c3' N U') M N M N 4

M M tM ý3' t1'

4 0 V4 0 ý. {

ýR

OD 1n 11

'T

h to 0ý

ýO

M s-I

M

'IT T4

C4 r4

(''1 1-1

to h

, r, c-4 c-i

0) O c-I

O

M r-1

T-4 pN

cN co

m

r4 ci

rn M C-i

Ü 4 [A U W W t9 x H ºa Z O V-1 C\1 (4) O v v

88

D w

N

cr,

U

8

U, wl tlý I

wl H

[s. ä

to

U,

ra

ö w

C H < C) U w cL4 o x H :. z O n N ý N C

" d C In N IC) 0 In Ol "ZZI Cý Q c-i In In O M lJ (N Q' c-i in fN In N to In

N Cl) c' cM CO N a) 4 C d' 4

to 4

N 3' In lz7 N W

II N N d if) tC ý4' l0 ID N N co N N -;: r c-+ T tO N V' O" % N N U1 In cl ul TA l0 co N N ri l0 Gl 44 N M V' N O M Lo Ul r1 tf) CJ N %- N T-1 N N N N r4 N N N r1 N N N

" d' O N f'') ON NT LS r1 l0 l0 Ol N '1' co N (y) A N M r1 OD O C4 CD OD co N N c'1 0 N d' to

T-1 In V' M 10 I) CJ In CJ c'ý1 d' to ý4' N C7' If) U1 N

A N

Q to N lD fV CV O N In 41 c-4 to m to m ,o Ol

z z CO to C4 O n M O O O M N c+ to cm co 'i'

t N '. O ON In 4 4

N O N N tD O O N c'I N c-I (V lV N N r1 N N N c-I N N m

" 0 r as. co (V In CA CO In dl cx In ci 0) V' O1 ý-. A 01 N to tN O , O M ON (V lf) In l0 cm ci co 0 " " " " " " " " " " " " " " " " "

In 4' M '. O '. O 0; Ln O d' d' S d' to to N N

M

O t-i M O t! l l- OD 01 S C') In d' N-4 t- c cy) c-4

z z l0 ()N to CO l0 r+ '1' N N M N l0

W c4 c4 S C x) lfl C" N C4' lq O N In l0 Cl fV CSI N r4 (V ri N N N N c-' (V N (V c-i N N

" O M ON N CD 0 d' O) to co d' N O1 0) M ON "" A cP Q1 in c7 V' N O) LS V 0 c\t IC) lD iD vi N

" " " " " " " " " " " " " " " " " [-I dl t! 1 N 1p In Ol In O d' d' tp U1 N l ul tD 00 N

T-4 U 11

0 ON d' CD U) CC) to C4 ap d' IN N CO N O d' z z ps. O c'1 CO co 0 O N '. D Cs. rM M c'"') l0 N

" " " " N Ln v N ri lp 0) cM CO y N T4 N vi (V N N N c-, N N N %-I rl N CV

" to ßý 0 d' N N [M N d' Ol to M r1 lD t0 Q

S d' m co Os. d' 0 vP m 00 N LO to N tN O

In L4 M tD N CO tD U) l0 'i' N L 10 t0 Ct

PO h

N U1 L- lD M O In h Ln r W (V z z d' 01

A N d o kO rn r r d M c O ýO

.r " . " " " . " " . " . " " " " " c+ u) O O W un tD to d' d' N M tN m d. 'D N c4 m N N N N N c- N N (\I c-4 ci N N

" t- In In CV cc) It) N ', L '. co c'i CC' N 0 C.. ) CD 0 O c"i L- v'{ IC) co 0 c-i co N ('f: M' N co [7

rn vý ýO M N Oý d O ýn to ýt N uý u1 to N ý

co 14, %-1 0 (7) co LO ý Z O r4 d ci N Ur In N M

CO L' Cl 0

1'1 LD C1 0) 11) N l0 ýD V N d lt) Oý d' CO tV cI (V c-I (V N CV N c-I N N N c-i s^a N N

a o 4 m U w w O x H O H ö c

r

89

co w

U) U)

U

8 U)

U) EJ zU Wa

Ii

114

U) w

d

w

C4 0 H Q `= Ct W r, 4 (D X H ,ý 2 O v1 N M G O' C C

" c Cý Gil N In CS' tO N N a) t0 Co r- fý CO co %- Cl N N r-1 0 ON Co O r- GW

Oý (1ý N 11 M M 1Y' N to (fl ('7 M N " N c'f) N 4

N

u LU N N c-i (fl 0 CO Co Co V' m Ln Lr) d

Z Z In LU 1. D M O Ql N U) C T Co LI N l. n c-I G4' is

Co LUU LU 0 O (0 N Co N c-- s-I Co 0) N V) O . z. c-I t-I cý c ci s-I c - ci

" N Q' 0 crU 01 c'{ 0 Ci' to N N (V tJ ; i' 0 G' O Co O c-4 CO N cM c-I N ra co c-4 11 c+ T-4 W

M U) r1 s- I 4

.4 J' rl t! 1 M M M N 'C' M M r) 4

N

a n N ON -N Co In 00 'd' IN 0 0 M -t %D -V (-, i C\j

Z 'IT L d cý %p Ol CO M ci 1. CO Lr) ('4 M M tf) v Q: " " " " " " " " " " " " " " " "

r 7 Ql l0 d' N N c1 N O" O' t1 O c-i c; c-i N N ia , c-i c4 s-I N-i c-I c'I t-I c-A c--i c-i

" 1 TA VA M N Ol %-4 l0 N 01 a) O CO O N c-i ^ Q Q\ O N c-ý N LU M N co Q ( c-i Ul 0 a' Q " " " " " " " " " " "- " " " " " "

N En N (V M d' d' c'M to M M M M ICi Cl) M M C

(V d'

U u Lo Ln Ln un LO c-+ cn O ci+ op 0 M c' to co co

Z Z 0 al m CM L(1 LO N CO N N U) M CM to N Ol

Ol tp Ln N fV N M pp co 4 O O O c"4 N v-I c+ C+ c-I c-I c+ c'i* V-i

" Vi Co T-i 01 t-i U) z1' -q 1 N N Co Co Co V CU A (y) c+ co Ln w N CO Lr) 5 un VI 0 d, w I'D O

d' Vi M N M V' U) M to M M M M 4

Cfl M M to N

U 11 N M ON (31, VI O N (fl s-i N co N M N ON C-

: 7. CV l0 '' c-I CO ci c-4 N U) (V O 00 (\i tD lD co v Q " " " " " " " " " " " " " " " "

( ý ] Ol N N N N N M ßl M O Q' O VI N VI CL 7y

c-I l-1 VI t-1 VI C-I C'I VI c-i VI

"' cH l0 c-I t- ri LL0 0 5 ti' '. D c-4 N T-4 N u'1 A N Oo vi VI r+ CO 'd` 0 '' N T-1 N O tD tD c`1

N U) M 4 1f) tM In CI' M M M d' M M M U) cI

il l0 m M to lO V M qi' M co l0 V' L(1 lD t! 1

Z z CO ' i' u1 d' N .I OD d' q N to m t0 0) (V co v Q " " " " " " " " " " " " " " " "

( = ] Ol Ol ýL 7ti

c'i VI c'ý ci c-I 11 %-I VIA

" O) LU C- r+ 0 Cl Co N LU lV N ', ' 'v M (IL) n Cý 0 O G' ci q, (, -) C') N N p' '4' rU 0 cc) CO C.

I! ) C/A M N ch U) Ln t LO M M N en d` en M M G"

ýr Q Q

CV V-1 OD ' C; ) O C\j co CC) 0 .; Z d' d' O' d' O) O) 1. O CO 1. o r i d N O M Q1 -c: r

O N to cn cI N '}' rn N t7' O m 0) ci cV r+ t cý ci c c'ý VI r-I vi ci ri c'i

LoY.

C9 x H Ü ýt ft1 U W W ýi z o d (7 C7 cr

G4

90

CD W a H

M U)

z H

i,. rYý Li W

f-i

S

c:.

cn

C c:.

U

O C7 C 0

C r1 C, In M Jl 0) V L') -, z' Ln r, l0 c"i ON Ili CC

O t1) <`'1 N G` M C'? (V (Il 1'1 CV M CV '' N N M ., W

ii

<' [ý

M (V

" C)

0 M

" to

N 0

" ýt

N N

" U

G (V

" v"1

Cl) if)

" (V

M 00

" M

N lf)

" Cl

M M

N"

N tf)

" Ol

N c-1

" ca

CO l0

" Cl

C') lD

1 Cý

CO N

" c)

CO to

e N

[N Q)

C)

^ "

C: i C- V'

1l a) C)

G) LN

(o l0

c-a CJ

0*1 CO

0 C'

co tf)

; i' VI

ko h

a) G+

m t))

q VI

to (N

c' G

O 1] Cl) VI (Y' rl G' N tr) N C M N C CV (4 M Lr) N

F., co M

CO Lý

to M

'i Q)

CV CO

N N

I-i' co

CO tD

' O

VI In

I rl

C" tV

M M

O lp

Cl) CO

G) CN

v !ý " " " " " " " " " " " " " * " " j l+ Co '. 0 q: ci b rn rn O O co co

. S.. . -i VI VI c+ c4 c-+

" O M CO cl 0) i- a) c-+ ý' M O M 0 rn rn G' ^ v-I

Ca a

tD "

Gý "

C) "

co "

N "

ON "

Ln "

T-4 "

N "

In "

Cl "

M "

ý" " "

a.

Q] U) CO IV, M CO a N to r1 N M N ý' N M M `. M

(V O 1

C"1 iD to C+ O N CO N [- t- (71 l0 c-{ Ol d' O'. CO TA [N Lý O ri` 0 al 01 (V N CO r'1 0) VI N

O

. ý'ý.

N N lf) v1

Cv VI

e' N lI

co CI

tN CO 0 ri

0 v-I

ON co co N C'{

c-+ C"I

" d' i'D d' -, v c-+ to "' % O O M cn d' m ^ C. co co c-# N Ul N UO tf) lf) vi C') i'D N d' LC) co

N V) CO 4 4

4 mot' CV) 10 r N ý' (4 c '' M M M ci' (V

U tp 0 w N C) N N 0 L- O V vi cm w l0

7.. Z N c-1 CC) N U) 0 11) to tV N U') tI d' ON N co ý. / '. " " " a " " a `" " " " a " " " "

u. b O L) CV) D' N G co Gý N O Cl O '. OD CV V-!

^ U

" (3

"

0 Cl

"

U) l0

"

N U)

"

r{ C)

"

M M

a

0 N

"

G" Qý

"

c-1 11)

a,

N pp

"

c-I M

"

V M

"

Lt') O

"

O1 qv

"

v-1 U)

"

M C'1

"

Ul CC)

" Cil cn cI VN CO tt) M 11) r'1 N d' M tl) M M M ul

Z z c+ (V

O co

in CO

d' N

co CO co

Co co

In N

In CO

rn ()

r+ Cl

Lo d'

to M

O 0)

CO l0

co r+

co 0; ICA r{ c'4 ci C4 -+ c

C(' ! 01 ri

61 Ql co c1 v

('1 tl'

a

V)

C

M

t

VI

c 4

. ý, (Z) ?

M

t

L)

'i?

M

cX

N

Ch

M

C

M r1

c\i

M

10

ty

r; Q

t7

't `Z

X

VI N

CO

in M

CO

m [-

LO

N lQ

c", r+

CO to

C\J r1

O LO

ci

N O

to

M ON

Ol

IT CD

CO

CV O

C\J

lW CO

O

p r' i

m

[- v4'

M

(') Ul

CD

O lp

(' I

O M

T-4

O w U w w c! x H 1 z o ä cl ä c`'1

91

N GD

w a r7

Cl +

4 f=. a 0

0 E+ E u W L U w H . s. ý. 0 r4

cl i'V 0

M

cY t

o'

" V Lý ý" c-4 M ON ui to CN N M N M a92 co Cl ü ti C (N O CD Lr) 0 d' u) M to C Cs, M to u

! CR U; 'C' M N l0 U) L" Cý if) CJ' LI) (N C- 4

d' Ul a)

W N

Z N N O 61 q (r)

v) tO N 0) Ul 0 N d' M C')

l0 to lU O O

Cr) Cr) CO C 4

tN c-i

0 VA

O (-1

:i N

4 N

ti N

ýU N

CJ) r4

11 ci

C N

Cr) N

G) c-4

to r4 N

N N

G CV

" N v) 00 M T-4 Cr) r+ ýý ßt1 G) N V-1 N Ol C) a U N C) CD N to a: ) 0 LU In M En Ln r '' c,

O 11)

i to M to lid 0.1 In Cl Ln vý V V N "4' to to a)

N

v Q W :4

ON "

r

Clý CO

" Cr) .0

Ch tai

" co CV

O Cr)

" ci' N

q If)

" tO N

m Oý

" CO (V

LU "

If) N

N C)

" r c

O') 'v-'

" co c-4

to c-I

" T -i N

c-, c-I

4 c-{ N

N co

" r1 CV

N tO

"

v-I q

to 0

O" N

C) ý

" Cr) N

0 ý

" U') N

" i) r+ N Q O

O to

() to

CO O

(T) Cn

M O

pp r+

N tO

N to

M t0

q tO

M ºn

M ß'i

T{ " ül ("1 tD N CO I) O V Ll lf) 11' S 4

to l() C;

M (V

2 F

W

c-, Cl

l0 %-i

O c+

C c-+

d' ci

v-i M

Ln O

uo N

Ln tD

tO N

q

'i' N

co M

iý N

tO tý

U) c-4

O N

S vi

CO Q

N N

r t-

O CV

to ul

G) cI

O tD

CO "-I

d' N

O N

O Ol

N

r

(N CV

CV V

Lr) M

O r

M O

Cr) 'Z4"

N c-4

tO co

CO v,

0 O

t0 O

c'i N to

to tD

vI c+

(0 Ci 0

O " to M tO co O t0 "D Ul t0 to N L tO tD Ol U 0

(V Tý

u 11

Z v

k

to

Cr) O

q v-1

CO C)

Ln s-1

co to

c-+ CO

N LX)

L N

LO V'

r'! N

U-) N 4 N

V) co

N N

Co N

N 'H

r-1 N

N T--'

0 Co

c' N

N Ln

0 N

LI) r-

q cl

N t-

rn c-I

0 to

0 N

111 m

v' CV

c'i

M N

" c-+ N c-(

N tU

r O

to C

c-s N

O LI

r r

d' M

vý M

0 N

r M

O ';: M

C) O

ý' ''

Ln to

O CV

" If)

to cn N N C1) u7 O t. ) 4

tf1 to CO In W tO O

11 Z Ln O kD I-

tO c-1

ui ev

c-I M

O N

U) N

M ev d' N

r d'

In CV

r en N N

CO eN q -4

co to N c-1

d' 0 U) (\I

0 u1 ci N

q co q c"l

'' o C) 'r-4

r co O N

(; ) co tl) CV

C) 0 M N

(v rn cn Cr' C) C) ,:, Cý C' C) C) a7 C'! " N (' ý N U; Ul Lrl ("1 V' C) N l; J C\

"" " tD M N q C) N 1 t0 LI) U7 S U) lD r a) M En V'

It

z z is

,£

tG

19

C

Cn r-i

l0

N M

c-'

tO N

d' 4

CV

4Y

c' N

M

O M

CC)

C" V-4

U)

tO ca

Lf)

tO N

d'

ri CV

t0

t- ci

; ý'

q T-1

Hfl

O CV

;r

to N

Ql

N N

Ei Q CD u w 9z4 O x ri v ýN ov w

92

ca Co

H

U

a

8 F4

DL4

ID

U)

H cý

Q z

w

U C U W [l. O x H .: Z O c+

O N C7

rt C

ý* C7

" r tL O c C1 00 U1 C (V co c1 co u N C) r-- co to r+ F. M N c O N N O Q) ( :.

Ol t1: N N .4 C' G 5V) L-1 5V) (*) M [h 1; [V) (V) N '

i.: :I t0 Qý tD ON V-1 'J Q` tV V' Cr) CD 00 c ý* O N

k O ýn v O 0) C'`) r-I c co co N Z c" c-I " C+ c-+ c-4 Irl s c -I T-l N -i --s c-s

" %-4 (ý cc c-+ c-4 co co (0 m [N M in C) l0 0 OD U1 O N ci CO V' O M M N Ol t-{ O C) O N

N (f) N N M 'Cl' L" M In M M M N d' c*) Mi M C

A It

N h -Ii' N to m 1- lfl 0 0 Q M N M ý} c'I 2 Z f7 m CO a% e Ol CO C\i Ol N -tci lU [2' -t C) 00

f O ýD m O N cI c-+ c+ ap ra c-I N al r+ c-+ r) It-4 c'4 %-4 T-4 c'-I c'i N-4 N-4 V-4 c-+ V-i C'1

Q Q to 0 N

N C'

1-4 en

O N ct' co

M N N

O cn

O Cn

M '

ýr O

CD Go _ CJ U CO en to

N V4

" U)

" M

. CV

" M

. d'

. V'

. M

" U)

. M

" M

. M

" cn

" G2'

" en

" en

" M

" C

co N U 11

N `., S M m O c"I N tf 0 N co N CM N 7 z cc co N Co C N M CO M Ln N tf) N I N c4 v . cC

GJ "

Ol "

l0 "

t}' "

O "

N "

N "

N "

O "

Cp .

c-I "

c-4 "

c't "

C) "

cI "

N "

e'") .i Y4 cI [-+ -q c-I c-1 c-I 0 r-+ c-A c-I

" in C) '. O N 1S In cI O In If) In O O ci N co A N O 0) C d' I'D CO Gl N '. D N N In t0 q. C'.

N " 14

" M

" N

" M

" V'

" LI)

" C`1

" LA

" 1.4

" M

" M

" eel

" d'

" M

" M

" M

" ý'i'

U p en N u'1 rc O C <'') O O c-+ 0 d c-+ d' Qc co z 2 N cr 0 N m c-4 ' rn N N m 0 tf) r) 0)

u C Ln 0 r ß 1 tz 0 O 0) c c--I c - c-I O -i c c-I c4

) w Irl Ln ý 00 M in co r- r- ý

O Ch -i

CY) (+l ;: r r-4 f qT (N n A t- l O l0 i cu D r CO t ) T-" to

" C!

" M

" N

" en

" Ci

s to

" e1

" U)

" " M

" C4

" M

" V' M"

" M

" CM

" to

N

u In a) O en c+ t. ' w co c-+ m .' 1n oý 4 cý 'o

z z-. 'i' O c"1 r+ 0 tO cý N S C O N N u1 co In

C cc to c-t N N M N N M t-I O QJ 4

N (N Z r" r4 c-+ c+ c-I c-+ t-+ r c-4 rq V-+ l-+ r-+

" G

N N

rJ C

"n V:,

r-- (",

c CV N G i

L- 11)

Azz. M en c cn

tl tv

. ', G, fN i

N C! M N en N In M 1ý C en M M 4

M M 14 Q' 01

11 N Il) to M co

-+ m Ol

V O

'r V 00 N OD O co N

Z Z U) l0 N LA c co N Ol O) M s 01 Q\ v Q

(yam]

w

" Q s-I

" N

" If) cI

"

cý T-4

" N TI

" N 0

" V c4

" %-4 c-I

" N

" m t-I

" 0 c'I

" 0N

" CO

" T-4 c-C

" N t"t

+ N c-5

cý x H Q m U w w a Z z o c-i C m or

93

m co

1-ý

F

F4

to W

Gc4 [+.

P4

ci

U)

, Oil

rn

Q: 0 w

` ca L w v a 0 cr

" CO C! -, N Ll Cl M N Cý CD in c-n N i- c q c'ý U N C) c-I to to Cr) (+) N tD Co

M U) C`) N ýi M C N to M N C`1 (N c N C') (q-) J) n

11 N ifl Cl Lý O ýO ý; ' C O N l C^, [N C'

Z Z C- Co (0 In V' c-I lf) O\ C -e Cr) u) CN N t- C 2

GR L1 m cl M CV C Cl [ Oý N z ci vý c-; c-+ c-i

" vi to CO If) N M U-) O to N Q1 In vu 0 d' c q if) C cc QD d Ol ('n O U') N C0 W l-0 O M 03

CI N

. L7

" M

" ci

" M

. M

. 'V

. N

. If1

" m

. N

" Cr)

. N

. Q'

. CV Cr)

. C)

. t'

C' A

U') N Cr) in N cI v-I M 0 l0 V' N N Ol CO Oý ice. Z N CO CO to to V' Ln CO CO e Cr) Co c-I l0 Co O v W

Ol l0 M O m N v-1 m " "

O\ "

0 " )

CC) "

CO "

O "

Z v-+ cý c+ rt c+ ý

" c O r tn O N 5 de H u In Io c-, 0 M .. Q ýO rn O vi (D rn co Cr) N mot' N OD co c-I e CI) Irf ý' Cl) c+l c-1 V' c V' N L'1 C'') N c'') (N CN N fM m d' N ` (V

ü u m Ln c+ N O CO to cm C- O rn G% Cj O O Gý Co N t-; Co U) in lf) Ol m 0 cý Ln w vi

Q W

" Ci

" N

" C'

" c-I

" d'

" N

" N

" 01

" ap

" 01

" O

" " 00

" GO

" v-I

" ('`)

c-I TI VA t`{ vi Ca C'4

" 5 C Cr) Co U') N (V N Cl t! ) L) 0) c-I lJ 0 CO q N q O-, N N --I Cj) l0 lO O) �- LN O m N O%

(V ýw

" U)

" M

" rd

" M

" c'

" to

" C')

" In

" M

" N

" M

" M

" V'

" M

" M

" M

" 'd'

U a ý, O M rn O ýr c-+ Cr) U-1 N Ch s -I Co O v Z Z l0 O) (V C' cl' lO CO M d' Co CO N O tf1 Ol N

(: 7 rn N uý r+ v" N rý O Co 0 0 O O, Co c-I N , e, cI cý cý c-I cý c-I IV-1 cý ci tý

" l0 N (y) N N O CO Co (V Ol O O N d' OJ

I

4 C0 cn N m O M rn C- O\ N c-+ c u') 'n N ýY

M ýi Q I() M to M N CN M V M M c7 º1 N

p O cý. O O% N e Ch w c-I CO CD Ol tý N

,Z Z Co N In [N Co to N IS in Zr C) CO M in c-I m

iä 0: cO L1) vi Cr) N M O CD c-I O O Ol Q) N N c-4 c'I vi c"I 5-4 c-1 c-; c-i

a C- 4 G M C^ M C. N C^ 1ý Cl N N (: ) N C-'

N [n Co c- ýN d' ''M In M (N cß'1 N d' ý'ý) M M '' C'

Co 9 ý rn ;Z Z N N In O CO c-I tO r N Co f'ý) N N

`ý t C) m In c-I cv m d' r+ C) c- O c-+ Gi " N c-4 -c- c-i c-+ c-I c-I t"'ý v v-I v- c 'I

aO C U Z M ü G C1 ü 7 W ,a Z 0 vi

a ci m d'

Q a cf

94

0 o, w

(h cn

U

H

t'?

H

EL)

iL

1

H

rý a

0)

X:

a U

4 WC w w ý7 x !i a Z 6 Q c' N U) 1: 111 c a a a LEI

" in it) CV c-i it) N N C' N 7-1 M 0 c-+ vi Q CJ 1"N NN (-I c-. U7 N Ln C ;J M 'f` c-{ M [J 0 a' C: J

u) ýt t*; 9 CO N C' Ol t r) Q' u) cf CL) [t ul V U)

11 in in M Irv L) N Ol in N Cpl a) O in Ol CV 'Zr

ý. fn M (V 6' in a' r in N l7 Lll Q1 C? % U) Gl

O cý ap Oý rI l7 V N w O M N '1 " l0 z-' . ý- N H N H CV CV N N ci N N N H N N N

" M Q co O s-+ C' 0 in M N O 0) m N N "ý A in N LU (Ti (^ U) N' N C (n C C') c-4 ß. I) it) N (Iý trl M W Co CO to Q1 Lý d d N Q in in CJ

A 1 1

q O c-l d !M iý Q c" CM Ol Cl to -t' N c-I z Z. co 0 U. ) tD c-+ H M 0 (. o N in 't "J V-4 N O

z (yi Ol c4 S H tD N' M c-I [

d 4 4 N O N U)

TA c-{ CV N N N N CV %-4 CV C\I N c-H N N CV

" Ul [N c-4 LI) N M C' kD U) t- N N .0 ci ko k-p A N s-1 Ol L7 ('') d' M H O N l0 N N Ln d' OC

v) to M tO r U) LO O in un in d' c; ) mot' Ln in C0

(V U II

O\ ci CV N' Cf) N O W C% M M O b V- 0 co r-+ O "r 0 tD In O M rn d' Ln tO N' co CC) M

Iý C1 4 (ý N '. D d' u'I U CD r+ c-{ N N O\ C') '. G

. 'ý.. ' H H N CV cV N N N c-I N CV N c'i r4 N N

" in (ý r- U1 co ttl Ql M to N C- N in to m c-a ^ A C- cI Ný Ol Cl Q' 01 OD Ol H M c1 d O\ M 0 IN " " " " " " " " " " " " " " " " " M U) l0 M lD t- Ol in 0 l0 kp in Cp LP) in lD Q'.

'i c O tl

ci C') O N N H CO H L7 to 0 U) " H CV ý. (Ti to d CV lD i ce M C) CV lD N l0 0 Cl C'`i N

v ., z " " " " " " " " " " " " " " " "

W U) to 0 N N N in H lD C') c-I c-{ a) O d' in e H s- r N N N N N c--I N N CV c1 CV N N

,

to cm (IN 0., C') Lr) -x LD m k-0 r-' t- Ln LO ^ Q M N a, N c') N H M '1' 0 N

in d' l0

0 9

U) in M N lz C-, if) O cN Lf) in Ln 0) u1 tD t0 0 N

A 11

cy1 OJ N Q w c'-4 CI N 0 w t- l0 ci c-i in in I- ('') co CJ c') N O in M O1 Ol (7'. N c-4 dl a)

O ký Ol lD O M to Ui i N "C ) V' 4 O N O d' v' ý., c-I c-I M N N N CV N tl CV N CV H N N C''

" in N S CV N' V-4 N: ' U) '" U) C"" 0 0 M c-' U) (V cl CO [- M M cV Q C (- in c-I M N N

V) ýLD M LO U) 01 l0 N in to ý7' CD in CD to CO 4

II C7 IN CO u) O co S N' lD c-I H H O EN

Z (7\ C1 tS to cc) Oti U) t0 LI) Cp N CO c-i Cv v> " " " " " " " " " " " " " " "

O'. in O N d' u') C) M to to c-i s-I tN Q Ln Cn H H M N N N (V (V c-i N (V CV 5-I (V N N

LL' o

4 CO U W C 0 X H 4 X z O ý (N a

N' 4

95

APPENDIX8

DESCRIPTIVESTATISTICS: ----------- ----------

PRIMAR YSOURCETRAITSBYAGE ------- ------ ------ -- ---

96

ä Ri a

T-4 I rn a

N p4

w

w

m a w w 4

W

a o Q Co U w cL. tý x H a z 0 ä ä ä Ü

orn c-' u, Ln tD kD O M pp co ON cr to to ' 0 A CO M (n co 0 CO M co :r co O w Cl) V

0% aA N N cal A 4

N N N M M N d' N c) N C'

W 10 U' lD u 4 N N c-r Oý c ci Q d' t0 cal t0 Oý Oý

' rý h

Z z l0 (V OD (NJ 0) O (IN t'4 l0 0) tD d M ' v Q

Gcý O In d' W Ql fml c-I N Co c-I N 1-4 a, cat Cri (V Z r! c-i c1 r-I ýi cI c-1 cI ci t-I c -i

" in cn ý CU c- Co tV in ß1 N 0 '

a O % 4 t-

N Q -ý A 'D O U. ) Zo h cV co tN co m 0 % V -

r U) N N M M M M 1 N V tS N M cal N

uni mot' O m co d' N N r4 q M ;; r ON 19 Z Z co 0O ON o0 0 U) co to m c ICT

. cý ö I m CD I a, 1; 0; 4 C4 C4 tz ri

rl rq T4 ri T-1 TA Tq IH r4 r4 r4

" d' tl) cI N 0 h h 00 l0 In c 0 N qi' cc N 0 q a, 'ri 0 c-l N N N 01 to M ON to "T h N

. d' it))

" Cl)

" N

" CZ

" " " " N

" In

" CV

" M

" CM

" C")

" (4

" N

" N)

" N ' V

N Ln 'r-4 w LA N N M Q CO ON C' h N ON N to

Z Z Oº ON W) w V (0 M N V. t! 1 a t0 ul 01 r-

0; C4 a h am (4 4 h rl cl t -, c - 4 c -4 l-4 e-1 rl

" a) in d' '. o 0 d' o1 c0 N

rn h M N +

0 h

r (0

rn 0'

g ý" A in ON O N o0 cI u1 I M 0% c

IT Omi "

U) " N "

cf "

d' " d'

" ý! ' M I! 1 cV M (v) C') d' N M N V;

M H tn O cr+ c-+

(7% m

ei c-+

V N

OD in

N to

ý vi

d' c0

to to

h d'

l0 t0

O'. rn

U) Z

" " h

" " " . N c-1

. " co

. s

" N cI" N" h" i" c c-"

" d'

%-I c-I ri e-1 c-I t-I Ti c-I t-4

0 O A CO 0 00 0 "o cl h m a, co M

IV ON "

Cl) "

N N" " M

" d' d' " M to N co M M co co M N

yMN

, 4 cm W CM N N

O (0 10 a, c0 ' °M `4

N m M co c

z z O Q t-I h d' O\ M c4 cV M h t-I c-I N 00 c4

l d'

I r4 ci cl cl c. [-4 c4 c-1 ca c-I c T

" h cv) h l0 M in ri A ýA U) C O N ON N 10 in f cm N %-4 t o N

to V

0 ER N

" e1

" fh rn d' (0 in N M e4 M d' 1"1 co M

U) N e'I

i9 p QN Q c N M %D N CV

( pý 9

cm Cm Ln t- 0 '

(; % N CD T. O p" in m , 0 V e M N 1n 10 vJ Ln

t4 J ý O ; e-4 N eý1 W

4 4 4 C1 O O

T-4 c. + It cI e-+ ei 1-+ r4 l T-1 O O

h M

tO d'

N M

O Q\

(ý t0

N c

N N

CO N

c4 0

t0 in

h N

M M A to O o

Cý Vl N N

m C4 4 A

W; N co co M co A N1 M d'

N ri In 0% ¢ý ` CA 0 0 0\

cn r4 ýp ýp co i c"I z N ` a% M Ln lý v c

< co U w W 0 = H 64 z Z O V-4 C-4 c d !i 94

Or cy

97

U)

N O`.

Aýy

G`1 is7

<1 0

w c 4 r-4

N

w

ao ö

G.

I ö Ü < m U W P. 0 x H a z 0 ä ä c (

e-1 to to LSl r-I lD 0 to N. N 0 N -4 ýp 0 N N it c) cn N M m . -a to 0 to O co co

N to M N M M e' N V; N N C') N to N M N C'

W t 0 ul II < '. O M ri l0 (0 co lo O O CC) O r-i [r) M m N WT

Z Z .ý N o, CO H (N 0 w N to 0 C) d -I r-H G) "ý " " f " " " 6

O '. p M U) ri M (Z N

CO O: N 11 h Ol M . -i H H 1-4 r-H H H H r-1 H-i rl

" Q oR rn O C' N N 00 co Q Cl) O O ICP A . O H d' co C% v co W Un 0 to 'IV CO En (n N CY M c') N U1 N N M (V C' N N M

4

U) to W ýD Oý C7 J

co r-4 cn N. C\j N 00 t- 0 ' LI) Z %T N N m N Q N N co O) H v 7 r

.. (c] O U) 1+1 Q) H M N N N Q H N N N M H r-i rl r- i H r-i r-1 H r-1 H r -i

" -. N tD U) n'1 0 0) l0 N. cc) co N LN Q N Q) 9 O Ln O O W O) to N. ul U) rn N M U)

0) Uý N ri ý7' M d' M U) N N c') N d' N M M v

W al -4 Qip 0) O O O r-1 O G) (n cV O u1 U) , -4 N N '. p

d Z H H V' N W 00 W ) N H r-H O N i-i r-I r-H ... (ýi O Iz PA1 0) (N ri ri C: O ri M CO N V'

,2 r-i r-I r-i r-4 ri H r-4 r1 r-4 H-i H

Lo ul O d a0 d O O a0 CT u, d' rI - t. 0 M N OD A 1 9 O N (n "o "1 14) ýo N dý N M 410 w p U) (") r-f V" M c3 N Lo N N M N V4' N M M .4

'a) Wd QLne

N. 0 H co H M d C'*) Ol N in LD 00 W Ch '7' Z U v N N N N cj' co N H N ül Ol N l0 b'

z M M " " " " " " " O N 4 Qý N 9 N O O N °0 H

rA H H H '-1 H -{ -t H

0 t- C- tD co tn OD 0 0) U) m 0) H N Lr) A H a 'i' 00 d' co O d in to l 41 r-4 M

U) V) c+1 ri M M Vi N Ul M N M N C N M M d Ea %1: 1 14

Q LO II N v ., 'cr Ln %D rn d' H ON 0 in '-I co '. O M N ; Z. 01 d' M Lr) in r-. .

l) l0 O 00 U. ) tn N M 01 lp Z v O N d' 0) d' N ri H m 0) O N "

c0 "

00 " O "

e, . -I r- A .ý H H . -i # H H

" pý 0 N W 0 N ON N N ON t- In N r-1 ! f1 C"1 N N O LI) M CO 01 d' l0 U) l0 H H tD C+) d'

U) M H{ 4 to U') N N M N M N M c 0 v

N N C) C-1 M y S O) N ( to d' 0 U1 l0 [N Ol tl' tN H N

z 4 z " " " " " " " " W O t c4 0) ut H H O rn rn H N al? O .4 ri 1-4 H H r-i 1-1

H

0) t! ) m l0 00 () C') 0) to Op U1 co Q1 N U) 0) A N 0) Qý Ili M v S CO N CO 0) N ON N lD G)

E Vj M I M (`l d' m to N N M N d' N M M M H

W

N N p r-1 Z N Ln to O N: CA co IN CY) In U) M M N

" " " " " " " " " " " " " " " "

O N M O H O O co O rn M O) co C ßr1 rl H r4 4 ri H H H H

a 1-4 Q ra v W w 0 x H a z a c c f Q i4

98

ON

H

4 Z O N

a

U, zA Waý

H! i W t7 4

W

01

w, al'. 4

Ö G4

a p F < CA V W C 0 x H .7 Z Z. O

O O Lti

iy% N. Q " O N N tom) N Cr)

Q ýn O tý rn c*ý <" 7

N " Ln rt tD tO N ;7 O ý' V ýr W to ýn C Co cn +N W to

U) C'ý lD h

31 tD ca N N N C r-i Co ' N 4 CO tD 1. ,

M

O O N Co N N r-i UV H 0 OD '0 O CA e uo [. 7 .ý N ao N O N cn .n tO cn c . -i t c

(N H N rA N N N N ri N N N r-1 N N CV

de Ol Cr) N lD CO --1 r-I N r-1 U) O ri N Cr) CO " Co Cr) O C' V' N u1 lD c+1 d' N O CO 00 N Ol

C In " c' M N to lD Ul Ol d ýS to ci CO Cr) c' tf) N

Ww c*I vi

in tl

ß ý -i N l0 c' In cm N 0' 4 N r

Cl N

Cl O

t- d' 0 in Co N O t0 rn r-1 in 01 ch O c ) t

ri CV N C r-1 N N to U) e-4 V in in Ol L N N rI N _A N N N CV rl (V CV tV r-i e N (N

en to d' lo Cr) LO . -) Ol N p1 l0 rn Ln Cr) to O

.. . O r d' to O Co d' N r-t to t- Cr) cn tQ cD b rn

a (rj c1 N (Ö N d' Ö

cl in Lri d' N d' In d

c0

U1 N adu N 8 r-1 c1' N O' l1) c% to N to t0 %O d'

z Z alb ý 0% m M CO (3 Cr) Q ri -. O N O to N O'

O V' to 0) Cl l0 M to in N Cl l0 Ol in N (V r-1 N ri N N (%j N rl N N N e-1 r-4 N N

C"') N CO (0 O Cl Cl w 0) CO c+1 lS r4 N a% u1 " 01 N CO lD d' CO N cl' (l1 Cl d' lO N 0 lC to

" d' Cr) tp l0 o d' O d' d' t0 V' N d' tD u'1 Cl) ie H

<eN N e-i in O'b CO Co

+ 'd' O

' IN 01 '-i Co in d' to cn1

Z Z C- in «e N N c 1 r4 u 1 r-1 N U1 r-1 Cr) N N V) "

r-f ".

CO " "

00 to ü1 Va' In tf1 M N in U) (3) M 0p N N r1 N H N N N N ri N N N . -1 ri N N

Ln d' r-1 N W N Cr) O\ 0% -1 Cr) l0 1 U) '4' l0

.. " rn ri H Cl . -) H cn ýo 0 0 N to O' w N N 14t

; q

" " "

eý "

l0 "

ýD "

CO "

it) " O " d' " d "

ýD "

V "

N "

d' s

t! 1 "

ýn " N

Vý t nN W

V

4NN tD d' C) 1 tý to N

cN in t! ) to Co r Oý C1 Co v cu d' [D QO N d' fi lA U) e-1 e-!

i l0 Ol N 'cl

FF.. N ri N ri N N N N ri N N N r-1 r-1 N N

(A Co ic2 Aj O

N N ' 'mo ' . -. " c% Co N N in l0 Cr) N O C) M V d

-e S q"

d" ' "

N "

1. O "

If) "

CO N" "

O "

V' "

d' "

in "

V' N" "

d' "

ul "

u7 (" tIý Nr ta t

t0

11 0G QN M ßN ci q Vn CO Gn ý, C N ri e Co C% CO

z N l0 C Oý rn V O Co Co 'a' vi Co N (?

r-4 CO C') rn N cM ýO C; ri tn r! cý O O N 1-4 (4 (V (V CV N rl fV N (V rl r-4 N c(1

M r-4 In in w to u) U1 t0 r-i 0 n "

N tff U) C N (1 l0 Cr) d' Co O p

o% Kr q"

in cn tO t0 N ut O d' u1 t0 V' N ' In ýn N r4 y H

ON 9-1 4 Ln t- to (V CO l0 O' 0 d' en d' M r4 N O

Z 0 tD en u1 00 fit' CO N lO N r4 O-, Cr ( v r4 M u1

N , -ý N N (V N N r-4 N N N

r"4

9 ¢ on u w w cD x H a Z Z O CM ä Ict

99

H

ON cýa o

ww

w a

O Ei q U W [-, C7 X H F7 ý. Z O ( H C 0 G

" ('4 GO W M O LA CO O\ CV O M fV N. O Q N rl to e-i [i N in 0) Q' d 0) O C` M O) Q':

v) M N M 'J' d' M LA fV N M N cT N M (V ý* +H n L U tD

Lo iD C C' M VT M M . -1 to 0 V' 0 O'. Cl) fV U') CO VI CO "1' V' N M M ý3' C' l0 to O)

o o c O rn (0 rn OD m co O H ý r-4 ýj

CC) 03 q r-4 C\j -4 Ln t- C) -4 r- ON Ln O -P ^ Q N H H N l0 r- 4 V' tD 0) N O M (.; 1

0

N u M C 'd' ,4 ül CO M m N (ý CO N in

i 7 s J

Q Ln co Iq 0) r-I %I, co 04 N 2 Z '. O 0'% C% CO O 'S' CO

Ö M Co r4 H Ol N

W m ýD H C rn H H rn cD c-4 H rr H H

" M CV lD N CO M CO N H r-1 d' 0 r-1 l0 i0 C' ^ ß 0 N CO Q' CV H to (V C7 D Ql Zn CO N O N N " " " " " " " " " " " " " " " " "

'4N co [fl M N M M d M to M M M N d N M fýl C

Q ýD r- t- t- r-4 Ln r- r-4 tn W LO 0 -4 0 (\J 00 2 Z d' O a, tD co to co O O O ri V' %o . i' V'

Q'. N V' rH O CO N C1 pp N ri O Cl H M N ri r4 H H T-4 ri r-t r-i H r-i r-I

" 0 to M to . -i -4 tD ON r-I l0 CO N (V 1! ) l0 fa C1 H 0 N M c7' M d' ri S1' N LA L- U) ß1 H

"3' O'. VI N N V" d' :i CO In c0 M M M '4' N M N lt)

12 Q l"1 !N H N 01 to m CO N l0 hfl m N S In (ý 5

z z r-1 C' M H l0 O tD fM H 00 N º-I N q [1' N 4

W N L[1 (V H M ( 0t 0) N O O 01 4

4`1 H mot". r-I H-1 H H H r-i H r-l H r4 H

M ºn to CO v+ CO 0 1-0 d' '. e ON rn t4' CO a0 0 N 0 r H d? (fl in to u') N CO O) d' M r-f CO M N M ti' d' (0 N CO CO CO CV CO CO CO Cpl d'

NMN E

it Q N to iD In O M N 't' H '4' '4' l0 M H O

z e-1 to d' CV H 0 Ol 0 N W Ol M CD [- O N cý (il 0' N Co Vol N N Cl 0; (V (J') O O H N (, 4 . i'. rý eI r1 rI H H H rI r1 r-1 H

" a) a) M CO lC N 0 V H d' 40 t0 CO in 0 O ^ A m 0% (1 H tN r-I c0 co i. O U) OD d' H to in r1

U) (V H CO '4' in c1) (*1 M fV Cf c^1 (0 CO in NNH

L" Ch C:; l. 0 O? 0 *, ̂ V OD l. 0 C 0 cv Gt H Ci

M N d` M Ui O C3' O1 Q. N O') O 1-4 tV . F. H r-H H r-I H r-i H H rI r-1 r1

" N O\ LD LA 4 0 1.0 01 CO 01 N "ZI. (V fV (M H ß N ap pp CO In LO W [t' N r-1 01 In N N H Ol

n ()N C

" cn

" cl l

" A ý-

" CO

" " M M M (h N [I' (0

M C0 'CD

r-i 0\ W1 0 in N a

?. z (31 O d' N W N H O0 U1 '. D '. O co

6% U) H '. 0 0 01 N in CO tD 01 Q' 0) 01 H O) O r--1 dl H N

H H ""i '"H r"1 H r-4 H H

4 fly U W 0 X H 4 Z O C7 C7 C7 d

100

rn

w

H

t: H

0 V)

U, ca

w 0

CIO

r4

uä

s- w

0 ce U w w 0 x H a z 0 r ö ,

" Cl Ch %D rl H P H N 01 e-1 d N co tD In _ A N N Ol CO O H M l0 SD O' lD l0 M r-1 l0 O In M N Iq Cl) m c(1 lD M N M N

4 N M C4

+

C5 '. D Q .Z N (-) Ql CO N (') ri C) N M N O M Lt-) N N

2. (IN tN a) 0) %D O' N kD SD r-i QA (V Ol in c') ,.. -4 N ýD uA (3 O N .' o m O . -+ rn tz rn C' p

.ý r-I 1--1 1-4 . -i H H r-i -a

" ri i-i O C'') M CV CD CO N H 0) a) C" SD O\ "T A to r1 In Ql rl N M N M tD lD !I w t0 H r-1 d' r-1 U) M N S M V' N In ('M (V M N M C M M L

Wl r-1 t

O U1 il z Z -f CO N m [N (V 0) (V to 0 U'1 M co iN p) M v " " " " " " " " " " " " " " " "

CO tD ºn H O M M Ul 00 O H rn N O) N p

" rn 0 In rn 0 co r 00 un Vn C) M . -i to Ln Ln .. q tN ON Orn OD --i CO N r-$ Ln in 0 In N . -4 M r- en SD in

" U;

" C')

" r4

" F)

" M

" to

" CV

" to

" M

. N

. M

" M

. SD

. N

" N1 c" '1

+H ( ý

p Q 'V' N .O l0 (\J t+1 d'

' O1 O O N M CC) l0 C-)

Z co C"1 e-i SD '. D C` ) H t0 01 cV Q N QJ 00 N CO N N U') N N M

S " co

" 0: )

" O

" r-1

" m

" t-

" tý

" N

" Z 4 , H . -i H H

" '. D SP ß1 '-i N 0 LA (A to tN (V II) d' 0 0 . -I "" A d' ao O co Sr M '0 O to co 0) M rn in H N N C/) M H 4 M V' N LP M CV M CV S; N M M N

4m C") (V to (') l0 d' 0 CO M CC) u) 9 i--1 C'1 N M

2 Z 0 Ln tD O t0 N rn M 0 (NJ Nt' C in a% N ' v " " " " W Ip N U) l'41 M N M pp p: O O

. F. rl r-i H H H . -1 r-1 rI H

" ri Ul 10 H N tD SD PD C") LA Ol M lD u -f M p) q q O O a'. N ro q N co (7, N Ln M qr M 'p t-

d' N "

U) "

M "

N "

W "

M "

d' "

N "

tf1 "

M "

N "

M "

N "

V "

N "

M "

M "

.' V2 CV) N

1 fl 4 N

v, N CO H 00 d' M r-I Orn N d' M O 0) (l ýp c7 N % d' t0 tD M ' r-1 N '. O O M N N N O

" co

" N

" to

" c')

" d'

" H

" C") W" Cho O" O"

" Ol

" O\ Q" ý

" r-1

" N

'-4 r-J H ri H ri H r-I r-i

" C") . -1 tD CA S to CO r-i t0 (V C') CV N r-1 H O\ ^ Q t0 01 M Q1 In H to c) t0 O 01 Cp M N pý 00 1/1 SD

" VI

" M

" r-1

" V'

" C')

" SD

" M

" In

" C")

" N

" d'

" N

" SD

" M

" m

" en

" '

N

(i Qt' d' %D H U) to 0' cý Q' c"7 Cl co C') %D O ul N

A z D CO N Ul SD H H O U) M r-I M al to N

v 0 pp " N" vi c'') " " t0 " H SD " " W " O'. " H a1 " " Q1 " Ol " W " O " cM r-I e-1 r-I H ri H H ri

" .i t0 to tf1 N .Q

[T (V N l0 d' Ln tf1 N M 01 Q IN 'D O O M O h M m S 01 01 O) O' N r-4

ON M M r.. 1 ; el! d' M U-) M N m (\i «r (V N M N F-4 00

4 r-4 2 v SD N SD H co N Q rq en to z z SD % O co U) ttO M

v 0 Cl)

N LA N C' O d' N 0) O m C O' CO O "-i H H H H H { --i

7 x T 4 cA U w r4 C H 4 Z O C7

101

to In WOm

QQ

WÖ w

"

El a 0 U O- (0 x H 0 C U T

w

N a O O 6 N Ct' CV 0 r-! cl' lD r j e N I- O C U l7

O ip (n tD r r LA O LI) !n 0 N un C 0)

vl + r-; En r1 in W

t ý N V- M 0 C- N in 10 D N C C

Z 2 ý- { H In M r--! CO (. n ri N I

, r (\I 0) : M N N H H N M (V r ri

N N ý--! N N r-1 f-1 N N N

0 M r N CO -I ý '_1 N N n " I-- t0 (1) GO M UI t! '' r l: 7 CO N lo M r-:

,...! " +n ch r r r cr O ýn 4 in t uý rn En l0 r-1 U) r"! E4 I C9 to If

LO ý; r C\j ON r- ID 0) CO Lo cm H 0 (\I co Ln C) "o Z z '. o u H m '4 cv N C ' c-, 0 O H to r 17 O . -i O r tD CO r N N Cl t. 0 CV CN ri H r-{ (0 lV CV (V CV H H N N H r-1 CV CV (\i

co CY) N ko (n Lr) .. 1 0 '-f Lo .. " rn u) 1.0 d' O C v co O n u O % N O O N in

0 a

" to

" cn

" t9

" t0

" co

" LO

" 0

" in

. '3.

" to

" qr

" CO

" en

" to in

" ß1

tf1 N

k 0 11) p CO (ý

' Z Z t! 1 to CV , -H l0 d lD C' N N ri q H M v sý

S+] N d O d' M lp r h 'p N N O r Cl CO ý'

r{ H r) N N N N r1 H CV N N H H N N

M O " '' C H O) y t d' h t0 u1 M en N ko

IV N

A " i)

" cn h"

" c0

" r "

In . O ý. . .

ºn " 4' r " d' . in " in " C) ý

cn c! " rs' I Qcü

O M O d' to N h 0 0 0 C\j N Z Z,. N O H M CV W in h N H CO l O ri ." "

r ýn "

. -i in "

Ln "

ui "

r" r" r cM H O (0 O w M ""! <n N N N lV H H (y N H A N N (14

O C O CO ":: V 00 O op r O O C)" 0'% cV d'

r d' N

A "

" to

" en

. tD

" tD

" CO

" [0

" O

" u)

" in

" In

" C1' r"

" to lt) t! )

" co

MN Xl

p Q ccm 00 N CO r h Ö r O in rn 0 CO -o u_ rn '. o

Z Z tO C') (V l0 co I) N ri 0 Cn Ql In CO V) r N v h in H lD r M tO N CO M o Oý C;

M N N fV N H H N H H H CV (V N

ýp N pp rn d'

O to

r O

M M

cý i

N -%

V IT

" 1"

l0

Ln 1.4 h r CO ý O W r in

O

to

to d;

M

oo

H

in

(. D

u)

u1

'j rn VINr-4 Co 133 Q)

dn

N r N ('4 CO [N CO r'1 0 M Q'ý 00 c0 (D E-, C, Z z in 4` N N ff , v r7 Cl cc) tp [Y: tD tD Q .. "

h " " O r H H 06 h ý" e "

CAD "

O "

ri "

ON "

rq "

to { (ý N r) cV N ri H N r-i N N r-1 N N

in tD d' 0) O N M tV to " l0 Ul O W C11 d' d' h in N in N CO Ol M N

0% 0% 0 LA N [! h CU ºn O In N t/j co U)

r 5 p

H 0 to O d' H m !n r N cy, 'co z z co M O) 01 CO M h in CO M l0 d' to h m v

ao lq: r O in M O CO 11; Co N m O H-! r H Ln H "I en N M N N ri H N H N N rl N

.ý c4 U W W t9 x H .a O c Or d

102

C' w

a

54 H

C) u)

z H

V) r.

W

4'

a

v:

N'

c°.. '.

O L-4 s 0 + N cl) u cý a C' c>

N N c+ C- ca ,n ca (ý o O iC u, CC in In ý. cý N In O N O M c+ in (Ti c-4 q cD n N N " " " . " " " " " " " " " " " " " C Gý M N (n V' Q' c^ u1 (") M M N M N %Z

(1 iN v7

Q N v /C 0 C-' X) c, N O as r-I mot` to O al C-) :L F.. UN U a, C'- CO CO "-,

0 (N (Ti cn d' in W o`. [-

n 0 lD Q' a Cl M c-{ 0 (1* c+ N O N N M c-i ý. c-4 ci c-i " c-+ c-i r-I c-- , -+

" o in c+ m CO (Ti r, c- (Ti O C- w (q r) w ^ G1 O c4 (Ti O al (\J M in r-4 lf) a\ N c-I N Q) Q

v- Cn M N M 4

M M In M M Cl N fit' M c. 4

N ul IT) 0 N

(D Ln 11 Q In CO 0) C. ) co lU N N [! ' N N Ul M 0 Gl cI a.

Z 2 cf Cl) Q Cl CO 0A LD c", d" CO O H M 6 O IN " " . . " O ýo v C T (ý v c+ °Q c N co

stir' r+ c-1 ci* c-+ Cr C-1 -4 ` TA T-4 TA

" N co 0, co (n C) L' Ql [f' cI Crl in in (V o c, 4 Cl N N O ct N ct' Id) cI M Cp M al CO

Cm in M (V M M in M M M 4

d' (V M (N .4 wMM

n kn 0 Iq N N N 00 d N O 00 a Ci' to aN V' N vi (Ti

,Z z V4 0 M c') l- M M al to o) N in ID to N C

O N O O Cl N O N c+ c+ c-i pp c-I M ('n cI T"1 c-I c-I c{ c- (1 cI c-I rl c'? c-I C"1

" c-I l0 Ca c'r 00 0 fM 0 O N C1' N N c7' ID c-1 Q Ci O O in in cn in N (r N N in (Ti to 0 N

d' ) OD (V N 4 4

<Y' m to (4 4

M M C' N M M (n (n vM 7

(n 11 f^ to Cl L7 N N l0 N C) M [N al M 0 40 (V 0

Z Z c-I N N lC in to w ; i' a) d' (V v7' l0 N f- 0 ý e o [a? O N cf' O c-i N N c-( N N c-I ri w ri (NJ

t'4 c- cI cI c'H c? c-4 ci t"? cI r4 tI

" V' ci' u) c0 w 10 1T IH cO ON N cM t- r4 M G\ A N o 0 in N

" N. M o v' V O O d' 't' 0 in

C' Ov

to 1"1 (N In C' V ' M

ýri r) 4

ri 01 v

'ri ri r7 v

OD M U)

Q to 11 N CYN ýt Ln r- co J. 'r a\ Z Cl N co O O w IT M " c-I (N c-n C)

. eý W O [: C' e-4 C-) c4 N 1 CO N Q c-1 O) cqQ cr crl

rl Cr c+ c+ c-+ c'' " c1 r+ c'I t1 c+ c-+

" v in V' N a

t- lO N

c+ in

tD CD

N 't'

[N ýT

N I

" at 0 N CD

A "

0) "

Ol "

Cý "

c "

tý " " " " " "

c " f

M "

tp fM c-I ý' t. 0 U] N c-4 M fi't' i! ' M In m c"l M M CN "4

" M

" cal

" Lf

V] (v N

t' c i!

v " " " s o " a W Ol Iz

4 N t11 O f')

0 co N--4

a X-4 0 c en Z 'r-t r-4 c-4 ci H T-4 c-1 4 V cc-I

O c-i r-I

" (n N in 0 '7' 0 v1 M cfl C) v' t0 C) C' d' M ^ Q 01 O a) (V V" u1 UY) cu N to ci N (V 0 N S tf1 " " " " " " " " " " " " " " " "

C) 'C' in N N (n 4' M In M f+) M M 4

C-) M M [T V; ci N W1

n < T-4 M M co co t', Q0 p N to N c-I N Cl c1' cq d

Z z (, vi (Ti N ci CO 5 1.0 " V' N d' v to M

o ° n c , c c °ý 14 c ý TA IH ° T4 o 0 0

x 4 m u W 54 O H a : E: z O Ö Cf Ö C

6,

103

co ON

H

ei a r4

U, n

w w 4

Gr

cr)

äý

U ý7 «t, U x H :' z, ti O + N co '- Q a CY

"ý ;ý CO " O 01 00 N 0 N CO M N Q LA) N

O Co r- :R Co N V' Co M N CO Co N Co N ZIP CV M M C'

(ýT + c-4 f+I In

,t ul N 0 N U1 NN c1 co Co co UD N Co cc) co U) z Z W tt) H T-4 CC) 0 Co W Co Co Co co [- M !V . r)

of UD 0 O CO Co O Co Co 4 O N- m O fti

Co

" Ol O (V 'Y' 0 c' If) Ol C' U) t0 Co C) c-i '-D Q q7' c+ V' o? O 0) cP Co LI LI CO LI) u-, Co cr1 0

"Y O l7 Co N 4

Co mal' N CO CO N M N '4 N M Cl) tr c11 L^. (v W1 (D Ln U Q U) 0 to u') co O 0) V 0 CO (V O (V Cl) cM al O z Z rA U) CO (Y) 0 C' c1 CO H UD 0) 0) O U)

0) ko d' O C' M O co O c-I O 6+ ci c-I cý c-I %-I c-i c-I V-1 c-i

" Co p) M to co C4' cI 0 kD to N (C) CO V- C' to "" A u) Co c-I O CO O) M Ln '-D LD Cl) Cl) CO c-4 Co O M " " " " " " " " " " " " " " " " "

'P co U) Cl) c-i d' 'f' d' N l! 1 M N cM CV 'c2' (V CO C') ü1 W Ln

O tt) u 1y V' Co u) c-1 CO d' W d' LI LD tD O (A d' c+ ci 2 z ON N Co CN to it) O c cat c-+ v-4 N O O m

CO N 4'' O ('4 c0 1 O of O r+ c-I W N C\l c ( c 1 -1 Q c1 c-{ r+ c -1 i c-

t. 0 :Y 0 iI (V CO a cq M a) U-) I'D N c-I N . ý. A U) 0.? c-I ci to N co cM [b N QO l0 00 CO V' c-I to " " " " " " " " " " " " " " " " "

NP tn V) c4 c-i 'd' c}' 4

N CO Co CV Co N eM N f4 Co U) N'M

ý 4c ý1 N W cV ca 0 O" %D N O O O ai q; r rn O CO z z Co 'd' (n M CO Co r c-I t4' N CO 0) N Ln N (l

0) N d' 4

CO (V CO O 0) O O O CO 0) N C) c-A c-I Irl rl c-1 r-i c- ri %-i c-4 IH

" 0 O C) LI) O) 1S d' d' A ON CO cn CO ON l0 l0 N co t0 (X) 0) N M N

'Y O U) Co c-: Co d' '3' N Ul CO N M N .4 N M M 4

2MN

OA u a `\' O 0 ri W (n ' 0 ` ° z z w °n ) 0 (0 (0 N N öM CO

C') n

°' r o m o c m w TI c-i rq 1-4 c-i I V4

C\j N r-j C\J 0 u to Co A CO a) N CO O l O t0 13, 0 N Q IV

, 01 Ln L/! .

Co c-, -T V. '1 m LIl Co N M fV d' (1 M fM fit' u! N (V

U-: )

u < CV 0% v-I N ri '4' U'1 Co lV O Co ic, 'P r) t+7 C') G 2 c*) co w rn GN M Q: cri c+ 19 c1 cv c tCi cý .. Q " " " " " " " . " " " " . " " " W 01 lý ý ý ý Oý Oý O Oý Oý CD O

c ý ý c c -I

a' C° `-I CO N -% Q to co O O (0 O 0) 1N co N rn ON N En Co c0 N 14 N d' N CO CO d'

M Q r-4 Co N M c-I d' N N In 1qP c-i 0 rI m a\ O'. [-

v O d' M N N (0 0) U) (0 : Cl : CD U')

14 N m rn (N a) o 0; Co r4 -ý

O

4 0o u W w O x H a z z O N M ýt

a (Dr cr cr

104

rn rn

d Eý

ýi H

0 U) a

En G

w

U)

w U) x

cn T

Ü

U O" c O C'

t

. -. . Q w t.! ) O S Co in e lD ' N N lt) () l

L() Q ci tom- LA N ci N I() C) ý)' ri c-{ tv1 ri M OD ýfi +

i r N 4 O in C in d Co d' ýn 't o; W`

ý7tý 1f S N vi cU () s-I Co if) C) t- ) ci c'i ti)

a) N 0 G) O ýD fir) c LD c+ C O ý7 N - N s 2: vi c-i M cI N CV N N ri N N N vi N (\J N

.. " N 1- ra d' !: ) ' c-1 c-i N C) 61 Ul N Co CU G'+ rn

- Co in to C' co C' O

. c-4 c+ ul u'i Co c-i Co c-+ O

O N LC) (V

. ti)

" LI)

" M

" N

. lD lG 111

. Q

" lD

a V'

" in

. d'

. IN

. ' in

" lf)

" Gl

W1 c-f (D U) Il 4 u) N 0 ýj

Z Z c' ci cm v" U 1 O N r-4 O d' N . " " . " " . " " " " " " " "

m cn cl c7 O S ter' c-I I-D N M c- I ü7 c-+ to v' Fý c'1 c) N cV cV N N tV c-+ N C) N ca N N N

n " ßý O 01 in N t- M d' 1'' lD C) N fT O S C) N Q N Un N CV Co C) Ln M Q M in N to (r) O G " " " " " " " " " " " " " " " " " cv (n Ln Cl) Ln Co CN t- S O iD to in d ' oý d' in in a1

W j ýn u 'l CD «e C'ý fn W OD v-, CO Ln t-

Z Z c-1 M i ri In C) ill r) O N Co CO si N O .... " " " "

Co (

in N N N lp Q\ LA l0 CV N fV N !V c-i N C) N H vi (V N

.. " M tf1 ß1 N CV 'V' mot' Cr) N l0 N Co CA lq N A l0 M rI to c-t cý S V Co O In M in Co lD to

" f a " i " " " " " " " " " " M C U) to M r N CO in O li

4 in

i co in in W

' W to II Q (7 0) t0 l0 M l- N Q' *7' C- e cm co lQ N C) l-

2 d r r- O cv in CO to . -4 t- CO N s cn rn aý ¢ W d' 0ý N ýn in ýn ci tD N c{ N In O d in

V-i T-4 (V (V N N N N x-' N N N vi cu N N

.. " CO �0 to O N M

M N

Co N

o 0't Q O C") N Ol to

d p

" Ch 'M in

" 0

" " CD 0 0 O 0 CV V] If)

O

uý cn "

u-) "

M o0 CO 0 tQ LA in Qv CO in LA in Co Wi (9 Ln !i <N 9 c-I d' c-I if) c-' lV n ýp

Z Z z ßl CD d` N O Ol 7' to ON d' N c-) (V Co .W O

" " rn

" N

" N

" "'

" d'

" c

" ýo

" N

" O

" cN

" " O

"

cn "

I-o N cI N N N N N lV c-I N N N

ý N N N

c! ' A CO 0 Ln C' O 0) tN N cý in O S 2. U') " . . " " . . . . . " . " . . " W CV C, 1 Vl I) CA r! N CO N O 1D LO to d' to lt lU OD

a ., te 2 Z fry c- f Cr (TI N N O in [ý cc O 5..., it ;

{tj G7 tt) Oý% �z3' r, H l0 O g c Z ri [-) N N cA N cm (Ni c9 c -I N N Iri N N

n " c-I 0% d' 0 ln M s oo it M d' c-f c'+ lf) M LI)

ýo A to ri rn to O W N N e c) 5 c-+ O in N ýn ! v) c-+ N v) u) to in r co to O ýO er) to ý4 OD in to Oý Cw7 8 d n c 3 2 , E e N in

Z z to iD c- tO O M 0 ) 0 n O v cý

WWW " s " " " " s " O

a N r. '

N N M N N N Irl N %-4 N vl c'4 N

N

p'.

x Q m b W w H a X Z Q Cr) c ä

105

oz o

r Ww x

HH

CO

W4 c49

a

0 z

w A z

H

a En m

H

ö

H Q m u w w 0 x H a Z z O 1 O N 0

cM O

' O

tp N vi e In '. 0 In Lo N M w In Ol N 0 Ol I/l Vi l0 v1 Ci' '. 0 d' N d' O V-4 In O N de M 0 W+ . . . . . . " . . . . " . . . t7 In O O O O O O O O O O O O O O (D O Q ý0 I 1 I I 1 1 1

d' 0 M ýp d' Ol d cý N 0 M N In c, 01 In l0 t/j lp N d' tý di -ýZv In Ln 0 VA Ul cm In O N N

ON 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 QI "1 I 1 1 I 1 1 1

le M N dp lS d' 0 lO N CO N d' [N N t-4 00 N U] Ul N O N dý cY) N L2 N O d' CV r O ri cý

" " " " " " " " " " U1 O O O O O O O O O O O O O O O O

4 d' 1 I I I 1 1 I 1

d' 0. cf N O1 01 d' M N c-, N S-4 Co c1 e c-I O Ul de cm O v M v 0 v w N O N N c' 0 vq N

" " " " " " " " " " " " " " " "

(D Lt) O O O O O O O O O O O O O O (D O

d' t0 N 0) t0 d' In Lý W 0 Ol l0 (N to to In d' V! M M O O (1 c-) vi 0 Ln N O O N N O n N O tin c) O O O O O O O O O O O O O O O 4N I I I I I 1 I 1

Cr In M M N '. D d' d' cý N N N In CO N G1 N ü2 N N O O 0 ' M O _l O N H N 0 N d' Co

(D O O O O O O O O O O O O O O O O' Q t 1

m M In 00 () O W 01 CO OQ C% e u) M O N O (0 S1 (-i c-r N i

" " " " " . . c) Ö

O O O O O 0 0

O O O O I I I I I I

Q 4 w u w w C7 x H a z O

106

ýi 0 H

H

F4 a z 0 U)

U)

H

C9

w

H H Ha

CQ

m

z

w a z a

w

a a 0

FA Q W v w w (7 x H a O

Ol N co ul In d d M O O 0 W S l0 N M U1 vi in c- Ln In N c-{ l- c-, cm in ri cý cI d' N

U') O ö o O O O O O O O O O o O O O <0 1 1 1 1 1 1 1 1 1

c7 ip in 'O M O O to c-{ O O d' a) N O c" Uý lD N d' (V e I ci Lri M O d' (Y) c-i cI N O

OLa O O O O O O O O O O O O O O O O Ln 1 I 1 1 1 1 1 1

d' c cm N N In vi al In r -i d' O 01 O N c11 O C/1 Ln N c-ý M N V-f in (1) O N M cI c1 c-I CV

ü l O O O O O O O O O O O O O O O O

q d' 1 1 1 1 1 1 1 1 1

le ;e M N () N M to m 00 e to p1 pp G c-I to tq d' m c) vl co O N O ul N O O N c-I O O N

" " " " " " " " " " " " " " " " O

in O O O O O O O O O O O O O O O O

IK? Ln CD M c [- Ol in O O ri Co to to 0-, VI

d' 0 vi -1 N O -4 d' N ri O N O O M

ON O O O O O O O O O O O O O O O O

d' tO O Cý d m N O l0 l0 l0 (Y) N v! N cn N c-i M M M N N 0 vi cm Cl) O O e e

00 O O O O O O O O O O O O O O O O 4N 1 1 I 1 1 1 1 1 1

p ý ýp m M to U1 N N N O Cl) 01 t- e cm Irt ul m 0 N ri l Cl) Cl) N O O (1) e o

Ö to in

12 ll l' l' l' ' 9j O O O O O O O O O O O O O O O O

4 PO U W W C7 x H . -l Z 0 vi ri ; ei Fit

a cs

107

z 0

a

N 0 V: ci W

H

ä

ul

z

4 w a zd

H N z dH

4 ul W

ü

aO L7

H Q 4 w U W Ga 0 H r4 X 2 O N Q

CJ in M vq I d ri CO c-' 00 0 "O v CV lD 44 U') N M N c In d' O O O N d N M ýp W+ " " . " . " . . . . . . " . U in 0 O 0 O 0 O O O 0 O O 0 O O O O 4O 1 1 1 1 I I 1 1 1 i

0 N W v-ý 0 v cý N co c4 to l0 O OD O CO UI to CV ca 0 0 to N 0 In c-I 0 O M 0 IH i M

0 In O O O O O O O O O O O O O O O O Lf) 4 1 I 1 1 1

d' N 00 0 ßc'0 ý t- Ol u') 0 O N a1 N N CY) V) tll N O O r M C') 0 u1 O O O IC-4 O O M c

u ) CQ O O O O O O O O O O O O O O O O

4 d' 1 I I I 1 I 1

d' d' '

pp -1

pý 4

O -ý

ri CV

ri fV

t cv cn Cl) 0 M O) '

CO Ul C ) c v M c tf) 0 (NJ c-I CV ri 0 C ) N

U' En O O O O O O O O O O O O O O O O 4 C") I 1 t I 1 1 1

d' in O M to L0 CO 01 d' N tN to in un C N En m C'n CV CV to t1 0 O ul c-I M N d' O O r-

N O O O O O O O O O O O O O O O O

M 1p ap d' dI Cl) U0 N ri N co N rl CO (ON VI (V M CV 0 l0 t0 hfl C'') LI) M c-I

N ri ca

" " " " " " " " " " " " " " " " N o 0 0 0 0 0 0 o O o 0 0 0 o O o 4

p ul d' 00 O EI c-I o o CV in OD S c`') l C') 'r' r 00 Lo cy. ) %-I O LA O O O O O O O O O O O O O O O O 4 ri I I 1 1 ! I 1

4 CU U W P4 C7 x H º4 Z O N CY) ri

C), a

108

H H

ri) M

ca

H H,

4

ä

ci w MQ xw H£ H tx H

ÖQ

ö `ui w ö

Q Q 0 U w w C7 x H i-1 Z z O ON

Gl O 0) d' M M cI M M N O O Ol CO G\ V] M M N O X, 0 v' N T-1 O O d Cf' N N V in OD in O O O O O O O O O O O O O O O O Q 1 1 I I I I I I 1 I

vi d 'z in N CN d G" 0. ý M cº l0 M OD d' l0 in Ln N OD OD cm c-{ 0 O N c-I N 0 O M M N M ('') U) < in O O O O O O O O O O O O O O O O in I I 1 I 1 I 1

ci N O 0 in OD M N N c-1 ß N i O X-4 M N N N c-4 0 c -4 N N ( N Q in O O O O O O O O O O O O O O O O

d' CO C% N d' M OD CJ ýp lý

W d' CO O N d O O r cr c-I O O M O N M

OD in O O O O O O O O O O O O O O O O < cri 1 1 1 1 I 1 !

Co N Ul d' N dl Co r- CO cm D (n 0 UI M N

" N

" N

" in

" c

" N

" " If) N

- -i OD N 0 O cm

0' N

O O O O O O O "

O "

O "

O "

O "

O "

O "

O "

O "

c; 0

le l0 O x-' tý d' d' N [b in 0 N in CO N e M ü) N N (V v1 in in 'V' N in M M N ý( O CO O

0Ö U O O O O O O O O O O O O O O O < ýS I ! º i º I I

O Co t/1 IM IH O c-1 M l0 in N O p N r, m N c-º N 0

in O O O O O O O O O O O O O O O O Q 1 I I I

Ö

4 on u w w 0 x H a Z 0 ä ä öý

109

APPENDIX 10

ALTERNATEF0RMRELIABILI TIE S

0FTHEPRIMARYS0URCETRAITS

BYSEXAGEANDS0CIALC LAS S

110

1 ,)

C-.

CS

r.

.:.

c

W

ri) :y

! -i

Ca

1<

Q

c;. Ca

G+ 0

Q

cý

C

ii C\ e

if) if) '

L'') t

N '

CC M N II) Cu 1- t i 1 I .J i- l'; _

() t'7 M M M t'T C7 i) Cl1 C) (1)

i y ri N H Ui I IX7 N C: ) C-1 JI) C14

. L4 ko C) co Lr)

(n

s

C C) C? CO a1 Gl Cl Oý Cl Gl ON _ _. _ . _... _

p C1 -4

-I --t l0

{ -o v. M c. 1) . --I l^ C: ý /) CI . U r- -i N C c C-) ri O

rý G7 N M C) C+: Cl) M f y) N M cr M

(y) M

CO N

c. . -H

Cl W LI) Cý Ol M N

N N

G) N

O O 0 Ol CD C) C3 m 6

j Olt M O LO L') L.. ) y N O. C7 if) CC) to 'C I N Si . -1 M ýS N CA cV lO l0 {

H CO M 1-1 O\ lO N O in in co O Gl . -t O . -i N C7% N N CO Cl to

. c

. ul

. . tt1

. tll

. d'

. V'

. un

. to

. .

O, l7 ON r-q .O t Q N . -1 G1 O ri C' Ö

O Ö

H

M ý

N (`) CO T (N N N C) N (X) Al

i " aQ N . -ý N lO N f- ). O k. 0 I-T i

H r-1 r-1 'A ri r-4 r-C s-"1 -1

r-"f . -i

CO CO (0 OD N U) H Ol cf) Q CO N N 0) H C) CO llý t' CO N

.6 N N CJ N CO N N (N N N N

?ýr, K- in CO N N M ý) O C) N CO M rl N N Ol O N G1 ri N r-i r-ý -1 te -4 p r-q H 0 4 -q r

r 4 H r -i -q a

H r -i UT

(N '

CV N to O vý -'ZT to 0) ti In it) N N N . -I if) tt)

( CV "-i N

C N. IT Cd r-"I N If) N . --I (I) N t

S4 N

-4 M tN

-

"

lD N l)

O -i tf) N OO H 1 ý'

CO Cl

[ V

C C v

(0 'Ci"

N '7"

co M

X74 ý"

() G'

"" C'

6 t0 ýD N O c

co l0 O u, O r"1 to to to l0 o CO 0) %D -i' lo lJ N

t! ) 14 M

C4 M M

M c'1 cM M M M q

g . -i N co to . -i N

N C-1 Ln C) CC) C) -. LO N 01

: 0) co (Y) O1 r"-1 (N

r- GO Oý [- Cp tz C6 Ci7

C: ) 4 4 r1 m -4 %D C, 4 co m LO 0) "i H r-7 ) 0) 0) OCR N 17%

U) N M M C") M N N N C"; N

if) O ". -1 N 01 Ol O (fl Olt C) C N lD H CT) N O V N ýT t'

t; 1 C^ O'l Gl (n C') C) C) CO ) Cl

Z '' -I '. D t") Cl) O rl 00 LO C' lD I

CO O t- C) lfl N co CO r) in c" t`4 C' (N H r-1 N C') N

i1 00 f") r"1 CO N

1 to (n N 0) (\ C")

p) [- O) O 0 O (`d r4 C- co to v.

. "i "

. cY'

. l0

. U

. Vý

. C

. C

. ý'

. to

. C

.

1 ý Ü

rU r-4 U

N U

[il Q

H O

CC) N C' ýO {

- I r1

ul tr) ºn to i F

(-A r

to .1 co

S N ýt' ýD Ui

-1n l

-- ti

v)

p4 P J

a a

ý L"7 L`+

O (Yi

E

j Q r; O OD C) O m H ra N { >w G^ Cý N N- co Gl C' 6' CD CO

.: L) M 61 c. ( w CO L') '-i t 'n CO

''-i U) C') ti' H

co I

N N .o Uc 16 N N '- N N N

`�yi O to N -4 L' N 01 C`'1 N O J-l a O O r-! CV O O ul O r-I r-1 ri

N (Nl CV N N N r-1 Ni N N cV

'

f

' C : i Lp Ln M v C ) r-i M CO H C; lo CO N Cl -i CO

iia C ? . N N lC lC N- N l0 l7 lG N- l7

z ý. M (n L7 I '

L) "-i N 61 O in co : [l Co N r) ) C) 1 Cl U) N- N

ýp c'ý r1( M <t h ý+ N V l0

in . -4 r-I O OD N CO co N 61 C) C' M N M

' %D co N co N- -4 N

M v' ý' v mot' r'1 M C) Cr

Ul l0 co l0 lý O0 lý N CC) ri O q OC) 01 ri CO l0 03 Ol CO O Ol

. -i 1 "-i N H .4 r-! H r1 N r1

C' O M rA . -1 N N C C) N "-1 H lt3 ýf' CO N N to ri N fit' O

to N . -! ý1ý N Ln

_.. Gl CO t-o l0 '-; m

O O H (") C) co 4 O O H O N N N C N; r--i r4 N N N N

rýr

ýy co ri tD in C) N Co 1444 (V C) co 'P 7 O ri In 00 C: 114, t- O l0 O W

ý w rV

[ý Cý l4 Ui Iý Iý l7 lD \O [ý lU

N 0) . `ý to t11 N N- %D in N r-4

ýy M M

ý 00 ýi -4 in ILO C) -, v C> Q0 P -i 'V" -T tO d' C" 61 CO U) .1 in

cM t0 V-1 CO co %0 C") ri m C N p co p) Ca N l0 CO O1 O G1 O O

f! ) r ri N N t-i 1-4 . -f N r- N N p

O N

in l0

-r r

N- N

'--1 ff

lD r-I

l0 e-i

r-ý l0

O U)

co ý'

N C')

V `. . [- 1- f- W 00 co r U l- 1- lam.

O N co O N N M O 00 M U) ON C1 O N H 0 "-1 O C1 (N H O

r1 N N N N N N H CV N N

- C) O N- Y C_) 4;

N M

tl tf)

in Cý c. CV N-

C C'

q C

. -i N

co N

( ) ri

co N ýo tý r- r

Z r r-1 l0 C"1 M CD '-i Co in C U I

cr1 O l0 C) 110 N 00 C (''') in IY' N %r N r-! H N M N M

O 00 O N- ul a1 v' co (N H v. }d %D O a1 l0 O1 Co N- co ' t0 (N 01

N .

ý" .

C .

Q' .

<N .

C .

N .

cr) .

q' .

-, v .

M .

`a n ' o o

w + rý jý "-i H H Lin in it-, th H . -4 N l0 . U)

117

ý1

cn

Ui f--t

Ql

-- ...

N0

a

r a

CIA

o H

H W

E

?, , ý.

Iý

a U H

8

U

ý

' U) H

`' C

M C')

rI C7

N N

O C)

I- O

L! 1 C1

M -!

7 t

1 1 t . 1-

L) LO N ') C) C

I7r ` In tI; Ill 4

(1 `: ' ý+ I H

i U) ` L. W ON O Lfl 1-4

Ol to ON

C4 C, %

ýý CO

cM Gl

rý C'.

rh { O 1

4-4 18

'O I I'D (N C7 410 ' "-4 CO N to c 1' C C) H C) cý ca: c- O

ef' !n t. ll ý(1 d' c`'S ri' V'

Z Gý r) CI LC) U) in '-f N Gl C) Ln CO t7 C I- c'_) t L'1 I-- N

r-4 C) c*+ N 'ý2` 1 C. ' to lCl

I i ri N r-1 N N O N N O CD N 0

N 1Si 'CF' lf)

al ý7'

l7 In

tD c7`

O Ul

0% C'

N 'cy'

N Ln

. -t Lrl

Q lD ON r-i if) N Nil N Cl Cl (N '3' cl lo r1 N C) In of lo CO rn m

vý cri C4 4 C4 Ci (, i C4 (n

lo OD

CO C'

M U1

W 6 -

V' Cl

00 lfl

N CO

N O r-S N N

l l N 61 Q7

n H -1 r-d H-{ r-i r-i H H '--I r-I

l0 C) N Cl N r-i v` O N Co H C) 0'. Ln N H CO C) H r o Q

M M C7 M M "IV (11 (+') mot' M ct'

l0 al H (V ()l lfl 0) Zi' 110 O M Ol CV OD M O '. 0 tN C 01, , -1

LI) (V mot' N N U) C) -zv l0 Cl un t. r3 N N N r-, l. o I. r) c'1 --t 1, - (N ýw N r-i N (M N H M N

00 O q I- t. l1 r-1 M OD O lfl y4 N co M O % CO N 910 co I- M

'

. . . . . . . . E

-

. . .

N O O r11 ('1 O Ä

p to O M

(') N

O N

CO Cl

O r-{

CO 00

to t-

O Q`

r-I l0

N ! f) M

Yý Cn to to

r--1 u1 rA

tp H H

U H H

Ln H r-i H

U) U. U)

[c], ti 7 {]

l0 CO

O al

d' O

to to

Ol O

G' iN

ü1 N

Mv O

-^- CO l7

d' 61

N ON

f C (.

5 C M M M c' M M (4

CO -i

CO V*

O O

C) 10

N U)

ül lD

( Ch

%D 1

l0 -

't' M

U) N

H Rý

r

tll tS) . f) L7 tf i Ul L()

9

Cý'

!

i) L') I f)

2 cý H ID cl cl O . -i CO in ;r 1-0 r`") N

C) Q'

'-0 C4

C) r-I

lD . -i

fV N

0p co

(-) N

C) H

if) M CV)

M CO LI) N 'V N -4 L0 rl iD 14 C H Ol 'd' ul V' v N (1) 'C'

Ln Ln r ul to M t. f) U)

H [ý ý1 Iiý

FH H

Vý. ý

--r "ý

r ýJ U U O

% 0 co z . i o 1

O w 0

F,, t t . -1

ui N

t mot'

L l0 CI1 co

r

113

GV

ýf=

W

:4 LA v

c..

üQ

b O

Q

N

T

II ýý

r-i rtn o NOOLn (w tnLn

N" a' l0 +

N M U)

N H %D

co (-I U)

m I-i in

UUU ý Ü Ü W ý L G

8 U)

I In I

C) 0 k. 0 CC)

Ho 400

IIA

}

00 0 N

c,

`4

ý_

ZS H

;f

U4

r

Ü w

H

U,

JJ F--t

r-1

4-a

Q

6

ýi

M rN to C) c, ", z;, O rn _t N i r . ýr r cý cý ýv to lzr e, ct,

. i

-: in r! C P-) LO O ýt O U1 N O C) cV cam: c, CO to ! r) r ri c

" -4

ö CI) c v' C') ' ' ci

rl cr M M If) N C) (') t' CJ Q rý - (N Cr) N N to c O C; t'

^r v ", ; 4

14 L`? 4

i' to L! V' I

M Ol CO tr) . --1 -A l0 LO c" t- N-

QJ to C'') U7 O : -i L r) H - N- O

(ý () to N O O' N ('1 N N N N N r -q r-4 H H H -1 r-l

H H H

m Cr1 O LI)

Ln Ln -i N 0) O Lr) z CO Ul a rn C"% -11 oll LO r- C-li

t" " CU H t`'l C N- ý' N lJ l0

N to CI) U) t co (N l0 co to O CV "--I LO H Co (V N -V r-i

i4 kD lfl Ln Ln l0 il- lO tO l0 lU lD

m c) C) . --l N N C) N O I tt) 6 (1) t. () Ol O N Ql ti' M N CN N

cn ýr vw m "I; r IV NT ýr c4 cV '44: r vl p - to -: 3' tO 00 r'1 . -i O LO CO to

r( N N O H En u) N N to O C ý rs tt7 d' N H M t'7

r"H r-1 f-i H ý--ý r-l ri r-i H r-i 1"1

CA to Co O LO U) ) En t- to (V W N O ý-1 .I M t[7 Co tO -I N

ýw 'gP "; Zr W to to 'r C' a' Lll tY'

. -t N O IV O N l0 al CO to ri C'1 V' CI) Ll in M v' CO H tD 01

In C-4 c) CY% CN Cl) IN -4 CY) r-1 H H 4 H r-i H H H r-i to N C' N N L)l O C ... __ _ _. _. ý.. _ ý.. _

z Ul LO CV N N- H t to CSI H f' N N H N cn N H C`) N

N O O n7 ('') d' N N to CO r-"f " 'ý

M (N >n Zr I O Q r-1 'Zr M C) W tD ýzS' l0 Zo [- In l0 to to to

, ca, d' O M tI? 'Zr CO al N CO f- 6 C t o oo . -i to r- In O 01 [-

'C ' cN C M tf) L11 vý cr Ul ý4' S1'

O C M H Ol 114V O C) '4 N LI) N- N 00 to N- En O to lfl H M

lO H O N 4 4 e+7 f et ý` i

N M M IS co to cn l7 [g O O Iý

oe. LA U; 4

1. n

ý! ý 1° C3

[ý Ch

C) In

U) r

lO Ul

co "Zt'

O (1

tf) lD

N q! '

t- [l-

lO co

ßl r-i

P tt1 N O 0) ! N N N H N N r"ý r--i rý r-'i -i r{ r" 1 1-4 --1 H

z -i to CD -i co in kD

M O tO O l0 N W c U) w N C N rl H N In N H rq m

N CO N N G- Ul tO -1 M M N- CO C- M m N H Cr CO U) C') to

. U)

. U)

. C

. tD

. N-

. lo

. to

. to

. to

. to

.

}U. }ý y.

ýý in

"ý

l

" Ct Coil

G C1

ý ý ý ý f 5+ ?ý >+

c1 Ü Ü Ä H

o tD CO

ü i o co t l t tn ri N c' vý CO

115

0 0 iI

tS. N

I

ä

LI

ti

U) ! -i

! S: C1

A

D

G

V

ýi

C7 a Gl lD CO -f N l^ Cl

cl

O .

b 03 t- N ri m m GZ .

N N N C) r N N

co 1.0 C\) CO : I'

f, 4 " N Lý (\ Lr) I) C i O (1) II) + ..

L r1 N ('1 t1 N N N CV

L^_ r-i r-4 H r-1 r-I H ri r-1 r--1

kD N l4 l0 t- O CO N, tý ' rq'} ri G co r- Ni i: 1 l11

l

N M r-i C) [- N NY' U V' co C lO Cl lD ri .ý U Gl 1

O N M %'l N N CV H ri r-i ri H A r-1 ri r-"! H -1

Z ßl m C) U) Ili Iß'1 H (14 C1 Q LI) a: ) in c3N N hl M ýYý dl LO iý N

CO C' ri m r- '3+ CV 7 lU

M CO N- N H cV cý . -f N al M

cri "-i to '- tip CV to CO co N. M $4 `V' C' M M to Cfl mot' M ýt V`

tý C) co to c, N r- - (j cc) 00 al Cl .-I CO CO Ol C t0 C)

I v Ql Co N N N N M N N N

rý mod' pl

Oh N

M Om

(n M

to L') co H_A

r4 (V ch M Cl M N N N N (") w .c f. _{ H H H H r-1 r-i H H . -t H

C) CO N N fit' O 1-1 l0 M v r-1 O ft lD 10 (N N LC) N

Q NT M M N M C4 M M

a4 N CO H to -4 -[r U') ei

r. -C' QO ', ' C') O CO O 1 4

co

ý to

4 ß r ý r"i - H C t'{ i -1 r{ r1 r{ e'{ l e{

- N cV N Ul O C 'a' lfl Ol L') In N N N i-"I lD i-fl

t ý'' N wow N H N Cl N ) N

to (N 00 N co m Cl lfl C") C) U)

C') C'r) In I11 T (`') Ln C

pi N r-1 N N C" N CA %D p M 00 C) M (N m N O r ""i QO

f7 C' N C1 M M N N M M N N C)

C U) C, O e-1 C ') ON r-i C> N cfl G

G H H H H [sa H H

" -1 H

r -1 ' -"1 T

.-! H

r -i r H

4') to Cl CO 00 C) M r-1 U) co (N Cý "l r N N O co to I, CO N. N

N M l'''1 M M -: r t4 M M M M

y r-1 N 6'. %D U) O N O CT f"') . -i O "q' N Cl N NI' O ül "--1 fý

M M N N N N N

H to M M O . -{ CO to lD M O l0 O ID N 00 CC) C") fl l

CZL- I- -4 M .a c ̂

O Ln . -i 111 ý1 f7 CO

M N O Lf 1 tf1 N

In "

M C") LI) LI 'd' - C In IM

N

-4 CN E4

--w Iýq t. 0 co N 'C I

i- p -4 -f d) Ln th En O %O Ott

.I N -%' O N

116

)

0 ci

51

H

Z-

G

t..

Ca

cn ý+7 E

El) H

all

CQ a

L=+

(U

x

M cl 00 Cý; -4 CD ^" ý _

` .

s L) 10 ifl .n to

-' M M m N tc ýt: H . CO le 10 t. f. C G7

ýn - N N [ý1 CD r) N H (N CD N

-4 n i 1

tYi . _. _... _....

L Ll-

ý ý` ` to to C) to } (N f'- -V n to " Cam' : _r-

t. n Q () ]1) ! sý c r . -ý cr c """ ý. fr in : tl Ifs lf'i Lf) Ul U) LO til 1ý1 Lý'T

ý-1 ul a CO c -i Co (N co c 1r Cý r7 N N . -? CO N N r-i N (N N

Cl C) CD i'1 2f :" 1 rt . -i (14 Cý C- II) i; ') lf) 61 N CD C1 c« { lf? h- N

M c". .ý ., w t r.

CO d' r-1 CO to -4 O to V al N 14 (N U) C' ct' l0 W N N CO CD -1

In a1 U^. 10 CD 0) G} 10 N- (D r- Q CO N ýD N O r-i N Ol N

(n In !n Ln Ln lD U) U) cri lei Ln En p 110 co N 0 V' 0 -I N O H M ri 61 ßl O tý r- 1-4 lqr Ol C1 10

O ß r4 Cý M N H r -1 . --1 H --1 i -i "-i ri H ý--1

"" .. .. N m Q (1) H co 10 117 1n N (N

-" M r l0 lD N ri CO C' to CO Q

Ln ul to Ln in ul Lo W Y) Ln to W Fý n

CO In . -H C) 1-1 CO co fl- . --t --1 uy N rl O r1 M 0! to N

r-I r-4 r-q ri M N r-I -4 N r-1 ', ý r-1 . --1 r-i "-i ri H H H r-i "-I 1"1

u7 (N -zr (d N 1t) CU if) LO N N N. . -t 10 if) N ý-1 [-- N cC t .I .t N c` 1 N 1 <h N

.i to co to [- O kD "V co Q) I-W t CO l0 N '-0 1- to N l7 CD N (11

[v) d' N 10 ll) N C) N to N lz: r l0 tl t. [) rd C) lD U) M rý 7

V) 1; UA t1l ID In to 1. n to Ul Ln Ln

CO N to l.: v' -: CO co O ON co N LO O N mot` Ul p) p)

-4 -1 rI -i

co mot' c" C d' lfl C% M N u1 O un to ul U) t[) tr) to C4 L'

CO - 6- CO (' 1 1ý DO 00 `d H

cr M N N CD D C N ce t1 cy -4 'A

rte --- Yd'

. -i lD M ý

C-*) V-1 '

to Cm C) W C) to N oD c. ) M to N C' (N r-i . --{ (N CO N -4 (0 t-)

(0 "$ GA N 03 G1 N H i-I "'7` C. ) c" fV to afl CO to M N N I` ýO N- N to to [ý- N lG

1 cil C'ý1 ý

ý v

ý "ý

G . -t u N u

W Q O O

00 11 ( (ý

N I

a ' ü

%D o 8 `ý

tD -4

co t E-4

4 ID

117

\ý

r! /ý v

`-4 T",

H

S H C d

ý f" i

V

ýi

C'

iW

r

c, I E-ý

G

Q

:u 0

H

W

i LI) C)

: r, VD c

co cl t ý; M C

ii ri ýý rJ rJ cý c) c -t r; rý r"

ý" n C) 6ý } p () In O C'7 D) C"

(n H r-1 HI r-4 O r-A

Al C; H

C H

Gl

.

Qý 1

1

_ c; 1 C3 O N Cl O fý r-t i` C"1

-.. F-. !

rT in M C [') t. 'M r) ("ý f. 7 fß 7

}""r. I

c= '"---+ t-ý

h N

r-I

h ý7

r-i . -i

O O

ri -4

. -! Ol

O

r-i

«J Co

r--! r-1

O in

-! H

"'D co

O

. -1

M Cv

r-i H

C Cl

H

C)

H

Cl Co

M in

O G1

U) }tt

N} LI) M

to ()

N t1

c` O N

U) (N

ON

cJ

I-

V`

C)

r1 (

r-

M

lfl

v

1-1

t

. -i

V' f

in

Cti

CO

in

M If) Co N tN N . --! M N d' If) () in in in (ft in C, Ul U) (ft L'l

(N r-

01 co

N in

L") M

O O

O' N

(1) I- I, -

cU O

CO N co 6

N N N N M N (N N M N N

0.1 C:

I- CO

N N

O O

O in

c' Cc

N O

%D H

H H

N CO

(Y) LI)

(N CO

Q 0. O r1 N N N -{ . -i y

Q1 . -1 0) M N C` M co Ln co in tý fý CO Co 61 O1 N lJ CO I- Q

N N N N N N N N N N N

t- N

in to

m N

H O

H O

m C) , -1 O

In O

v to

h 11

m O

w1 M M .r.

m r

m i- i

f r1 rI

w. . -i

M '1 'A

M r-i

M r-i

M '--0

M rH

`. t I! ) in

N to

C ' (V

N t-

(N I-

In -4

O l0

-tv to fV

ko ' -i

Cl N

(13 c' N H N in (N . -1 in N

1{ M (V

N- ý4i

N . --1

ý4 N

'V in

t- al

u1 CO

o a'

t. I M

M C)

CO O

07 C' r) N M t'') cV . -i N M M

.4 m (7% r-i

M (V

N Co

't ul

. -4 in r -A r-I

O O

c" U1

r) N N

CA M f7 M M M M M M (l7 ('7 (l7

V. Ri V' Ol

N N

CO q

H t-

l0 O

in in

t- ßl

N CO

C)

. Ol

CO M

W :,

A t- cz CO 01 r-: Cp O

Q C IN

c' tf)

oll ri

O O

CO tip : f)

. -i t"')

O 'i

-q M

M "

< M t'7 c"7 m m m ('7 c7 M M M

co q: v 61 cN N Q1 CO N r- O r-i O km N tom- N N o

'H

in C'. in C1 Cl 61 co* 1l 0l Cl 0)

C' M . -ý O

l0 %D

C, O

M l0

CD N .H Co

CO r1

ul (C)

v` U)

ko ;: v

N (V r-1 N M ('4 "--!

N r- Lo

co -4 -4 t

C4 `

co r-

t- C)

N co

I- oll

00 rl

. -1 CV

C' C C') L7 LI) U') '

C C' 'C

F] >+ ä >+ >-I a

Ü Ü Q o o k

C! ) N v' ý !

to to N

in C

t+ l7

C) LI)

Ei in co rl

}. 0

6 H 1-4

x

1111

2 Ca

in

z r

ý5

Ii

0 ý` M r-i CV c' O . -i CO u; c- n tn ,n co r r- cr. f rn i ' N :' N N cam' N (mot iJ iv cam; r.

Gi CO N N . -t cý CO . -i f O CO N a C CO N Cý t9 CO

to C1 CO CO CO W Cý C0 q CC os q

p q V7 ýv rA Lrl M LO Cam) %, O M c:

r' FCC M M M M' M M M t^1 t; M ti

N CS. r4 N CO C tD (N N CO O 61 Cl ri lý CO C7 N -N M CN

Cl t`- tý CO Iý r- c CO O) CC CO

n tom) O CO CO t11 "-4 N cl O CO

CO CO Cý N M M V' Gl If) V- N

N CO -i 61 CO N to r-i 4 CO CO 14 CO O CJ O c-, lf) )`N to CO O OJ

th M .

N .

r1 .

M .

M .

N .

N .

N .

M .

N .

c r- "

r- r- -P M N r- '

CO to CD to U ) to CO N to l0 ýY lJ l0 l9

N N N N N CR N N N

t9 N N CO C) "; zr O O 7 M r-(

_

to (71 N N )- to N to N . -1 3

-1 CO N N CO N N O C) CD 0 CO

*

r-4 lqr O lO CO CO co CO N to N

Q CO M M C. <. ' CO tU fit' 3 r- L f)

a, m (q m M M N M C'1 M M M

CO to

CO C

N )n

O (

to O

NY' r-1

M O

tO ). f)

N V7

M C

O Ol

Q

. 2 00 r r- r- N N C C) N c- r- : U1 N ' N N CO C)

1 IV d to m

Ui Lfi . z"

N N N ri 10 CO LO CIA . -i CO N

CO lqr O " a) CO OD (Y) ý 'IQ, CN

14 -q , N ' Vi l I

N 0) t- -4 C) U) CO M V M N N 0) M N

N ä tN

ue co n co O co In s -

N N N C tV N N N N N

Ly ýj ct' co CO CJ . -t [V N r"i cl CO

O "Cr O N (N N ON a% (V D

ß1 Oý týU 40 Ol. . 01 00 c G'; G1 C6

rl N N -_-

'd' -: Is

co O

N U)

01 U)

a% C

co N

O t- r-1

rd) N try

Q M C') eß'1 M CO CO CO r') M M CO

co U') -f CO to r-q m F )m

to d'

c, CO CO Cl

. t+')

O )t) )-

N Q1 N N Ln

, co '00 .3 0. ý 0ý Q' c as 00

r-i XD C') CO O . -4 co U) C+ tO CO C) lO O l0 N CO c') CO CO -rr r4 `; r N r-) ri (N M (N CO M

t "

C) . -i 0 O 'i' to CO (r) CO O if ) M N N

14 (11 .

N .

N .

9-4 .

N . .

N

. N

. N

. . .

is W ru

U U CQ cl ý d N2, C' m ýq %D co

n n o L ) t to to . co

N eT %D (n

119

)

M r4

90

x v)

co p

q

rz R

ö

U H

8

z a

w

CO N .O

M C) C' en

ee

M M M M r r') M (") i'1 ii rte;

' Ul N

- c-

ri Li

. -i W''

C7 (n

' to -V

. fl r-t

- L:

" C"

b'+ . ý1 O C) CU CS r-; O ß1 QS rJ O CO

L) r3 l0 lC co co o r c, 3 - Lii Q Lil lJ C" t-O U) 'tS` C`. ýn *f)

«-3 ti r' Iý M O O to co Ln N

M r'l

Iý N

"c' CO

3++ Q (N M N rt 14 N N .-1 ý r-4

Z CO n Cl G N

t! 1 to -t N CT) C) 11) N

r^

fNýl rr C)

SO " N N NNE U-) O M M r-1 r-t M M r-1 .

(N N

N r''3 N

O 110 ý

to N m r14 M om '

M 1 to ON r- V l0 lý

q U? M M cr) M m ei ri Cr) c- M Cr;

N O

$0 Cl

N- ri

Co º.

N- 1'

lO Cl., . -1

L1 . -t ll

co N ,

CJ 2 C> Ch cl CN r-4 C) CS O O 01

ý-+ r-i ý"-1 ri

9 O W

ßl r-

(n N

t") U7

m Oa

%. 0 W

C) Cl)

. -i M

[-- i- G'l N Ln C-

W Q VJ M tý M t7 M ýýT M f`) fy; (4 M

y C14 -W m 1-4 rý co 114V to C%3 f- % o sL' L P" dl a1 ti) N N O1 00 co , d'

4 r4 r4 r4 M N r4 1- 4

r4

Z ý

Zn N N N LO

r-i c- "o

N N . N N '' N r--i N Cr) N . -t Cr) N

v' M .i "d' l0 tý . -1 Cý M lfl }4 r 00 CO .N O LO Cl) M CO m O Cr) .

Cr) .

N'-O .. .

Cl) S

N N M M M

Li co ý-l

(A CO in

-4 U) C4 N rOi in

m C) N 110 Ics, r

* r- CO t y-

uj C4 M "M c4 .4 C' C v) ch M M

", ak 'm N -4 O tf1 N CO CO . -d 151 4 -V L ý N N tfý Ln N t- N

h e-t r-I rt r4 rf 4-1 ra O p

i .'

cC O to

M M

CS M M

N tn (In l : " "

ci' m . (`1 M Cr) Cr) Cr) Cr) Cr) C7 M Cr (Ml

co CD lqr . 00

N rt r-j N N N -i r4 ' . -1 r -i

C) O LD O tO N co co

Cr) l-r Ln l0

N d' N . r-1 . -i N CI) N Ln M C)

, -4 cq r-i r-4 h S ß ý C) 00 cn CO Cr N r-4 c i r -1 t N lD "--l $4 N N r4 N N r-i O . -r N N

` U U Ä ° r ` c w L o ca c

H

120

U)

0

Q

i:. Q

W it 3, 'sý' R. a

w

W

S

Q CO Ln 00 Ln ý j - CO ` y

N Cý: C3 (N M M tV N Cr) r3 LV

n ri . -1 N N O H v'1 CO Cr) M Gl v M ýr

. Cr) ) 0)

u) `' rn ö ri r{

ö c ö ö c cd

J) C) r-I -4 O CO CN (N CV Ol C rH N (ý N ch M M N M fr3 r1 cat rZ rý M

N r-t %D "-t v0 Cr CO w co L! ) 2 G; N Ol co N O 'SO r--i Kr 00 01 (N

"; Q O O r-! N - O rd r{ O O ra

e-i r-i ri º-ý rl r1 . -I r-! r-i rý ri

G1 m O Lr) LA ul . -! cV m Cl BIS 0.7 (A 0% C- P1 Cr) -" m Lr) Cr N -P CO V' -i M N C' N to

-" N O 00 N CO M tO lO N ON $4 N

N .

r-4 N

.

M N

.

N O

.

0 r-1

ON P`1

CV C'1

.

ON H

.

H N

.

C> M

.

.4 N

.

M- to . -1 O ß1 co G1 00 N tff CO CO to to t1 0 C LA tO O r- l0

N N (N N N N N M N N

(j jLý to O CO H C) 'q' r-i N M tO

a 1' O rn N to u1 ra to o

H 0 H c) C; H H r -1 r -ß H H H H "-1-1

N ýa aý +n c in m ON . -i rn Ä ri C) O CO H N H at H r^{ r-1 M (4 M N M M 14 N M M Cry

O r- sý

CO in

rl CO

r! LA

M N

ýi' 10

M O

N N

C' to

00 (1

rI ri CV N r-4 r-i r-i r-1 r-i r-i r-H r-i H r-i H H H r-i

to N --" N N (A C 4" cP to O'% U,

_

N H (V

M N

(N M N

O co N M to (%A N to . --1 N O M N H N

N C) W N to N

M Gl

-I N N .

N .

rl . .

() .

N .

r-i .

N . . .

sT d' M N N Q ON co C31 O 01 O\ N M O CO

U) N N N N M N N <'t C4 N

U) O N M C) co C) U) C- N ( eA N

. N rn C " N H co

C) C) C) 0 M 81 lO V7 N 4

" 01 C3 CO r-I r i CO OD Ö

Ö

N " M " N N M M M C'4 nj M . th.

( CO %D ON CC) l

in 00 '

(7% O C) lx) Q (V N N L O -G CO

. ü1 C tQ

Ol O r4

ri ri

r4 ri

O ri

O ri

G) ri

O ri

_^ - ..

O . -i

O t-

C rý

. 7. C H to M (A O . -1 co ._

U) %r lG (A O tO O %D N CO M. M to *4 Al C' N ri '"'1 N M (NJ H C) M

O to ri U) (A O U) ON oD )f) u) . -0' N

tC r-4

O N

LO O H

O ',; r

CO . -i

H N

0' H

c7 N

th N

dl ý V V

Cl 9

a sr sr yý ' co

u j om tn l l H tD co ri ' l D U]

121-

x

) w

Z-

uk

p D

Q

W P4

ý LO) H a

Hý 8 w

" co GJ Ln r- cc r- in N- Cv +^J ri

LF r--

p N

. r-i

gyn. " C)

Z., f

(N CV C3 (4 M M N Cat Ci (1! N

r-i H N N Co ri . ̂"1 CO -4 vs M 1 (, y In v" C" a% G1 t; D re) Q' M M M

LA a Co ra r-H Co C' Ca Co O O CD

.. ... ... ._ ._ F f7 Cl C. (N co O {` r C C) v) [! ri ri Co CO CV C14 N Cl C ri N

rs V? ( C"1 (Ii M N M rn M N M rr) M

[-" rt vý -4 -. O C) Iýw CO to co LA .j

Z;; c N ON co N O w r-i "zr co V7 (N L fU O O r! N H 6

-4 H O O H . c. rH r-1 H H r-4 H H r-i r-i . -i

% c m Q N 10 in N at r

Co Hfl CO LA O N- CO M a\ L ) CN N Cr CO --S' H CO C t' v" (N to to

"

(. N O CO N CO CO to to N ON

I{I{ N N

C-4 N

C) N

N O

01 r-i

ß\ CO

N N

o H

H N

O M

H CN

M in e-i 0 c71 Co Gl Co N t! 1 co Cfl tD to If) rn G1 Ul l0 O 1ý tD

N N (N IN N N N N M N N

Ln Co co 4 O "Cf' r-i (N e2' CO to

H H H r0-i r-4 H r-i . O4

r0-4 rH-i

A lam- ri

l0 O

.

Ol O

.

1n co

.

H C-4

"

O N

U) r{

t31 (n

.

rlti rl

. -4 r-{

O\ rl

U3 M M M N CO M r`'n N M M M

(D ßr1

H CO

H in

M N

C' CO

fM C)

N- N

C' tO

Co ü1

A .4 CV N A 14 A N r-1 ri ri

Z tn Ln

N 111

C' (V

N N-

N r-

Ln s-S

C %J

Ilzr to C, 4

to H

Ch - r

.r r1 ( cn N . -i C) (V

O co N M LO lD 'N c, to H (v

Cpl .

N .

N

.

H

.

N

. (n

.

N

. rr- i

.

En

. N

M

. r-4

.

Q Ol Co DO 0) XT ) 0) N C''1 00) CN- CO

Cl C)

%D co

CA N N N N M M N N M M N

N O N- M O CO l1

(7% Ul N N- 1

M M N N CA (n w

; N lqp . - CO

; C C G= M

CA C7) O

to Cý

V co C3 rH1 "i Co

m Ö Ö

N M N N M M (r1 Cpl N M . C+').

cn 00 to a CO LA co '

C'. Co C) '

to

4; v,! t3! N N N- O C CO

. U) mot W

" m O 14 4 O O G; O O O

W M C O . -4 t0 t!? '4' '0

C) N

O d'

to N

O e-1

a ri

(N N

co f"1

m N

M . -i

to M

lqt' C`)

O to ri tf LA O tt C\ CA In Ln %T N

to CA .O N

in O

to H

O V

Co r-i

H CV

CSl rl

M N

C') N

Q 1-4 U

cv

V Ä ý' Cl *9 "4r ý co 8 co , -i N c ýo in El) H

121

/

)'. 0

ui V4

O

c2

N

I,

'

CO N N

C1 co q lýD z

,J -- I , V' N O it "Y' .c N. M N -,

roc 41o c . -i (1) CO C: ) 14 r-4 co H H H O O O H H H r4 O

Lo H H H H-i H r"t r-i r-1 H ri

Ii Co O C' N Lfl V' C N. (N to 'C' U) Q M Co M N N N M r -1 C> \ t''

r LJG 4 v c' V' d CY V d v V C

Q ti Ol Co

N Co

O Co Ol . -{

Ol N

CD Ln

Ci mot'

. -i r `

-c7' ()

O M N V ,

H H . -4 O O O H N H ri H

Q1 ("1 O Co ff) M C

N ON C N Lf)

d' co - H Co N. v N t! ) V

14 CO al

tO H

'ct' N

Co Co

to O

N O

L; l H

Co Co

Co CO

Ui N

00 N

U1 .

l0 .

U .

tD .

lD .

lJ .

l0 .

U) .

to .

lO .

Ln .

Q C') N

Lii to

' tf1

N O ! ll M

Co 'tom

d' M

to co

lfl M

O l7

d' mot' d' LO

N N

Lii If!

tO C1

M 11

M N

N Co Cl

O M r-1 N

tO H

CO ßt1

("') N N -r4 N N N Co N Co N rt r-1 . -4 H H 1-4 -1 H r-i

---- ------ --- C-) C1

(D O

m r-1

a% O ß1

CC) O

C) 01

00 CO

CC) co

M O

Lo O

(Y; ýy d. Vi

N O

N ..

to ýi'

CO l0

G% "

to al

co GO

N t11

Ol Co

H O

CO V'

M N N H H H N Co cV N N r-H ri H ri H H -/ r-1

Ln Co

N Co

-T' N

N N

N N-

Co

r-i c,

t. 0 cv

in c' (

lC H

a%

N M N -4 N

Lii O

N O

C1 Co

('4 N

M Lii

N b

O N-

Cil ýS'

to Co

to W

M M

Ul tD Ul kil Lo \O 111 U") l0 to tf?

Q lO co

Co d'

O Co

T to

N [`N

Q' to

Co M

C) to

N t.! )

rn to

N d'

ýr ýr a dý

L=. N CO

d' m

N Ul

lD r-1

H cl'

O rf

to M

Co . -I

N l0

C ý2'

Co dý

w %" 431 Cl C1 C1 ON c c1 O d1 C% CA

M O " C-i n CO c) r- i r- Ln G1 ( 4 ( O Co H G l Co H

" "' co V" sip d' qw Cr d" f c

y O t0

N H

C H

Wo Co

Co "-i

U1 O

O N

M Co

O Co

d` a1

LO c')

Q ý O' - 0' C7 ý + o o C) C1 Ql O .4 O %

. ý+ C; r M . r-1

O to to

M Cl

Co tO

O N

f 00

oo (-)

ul M

d' U)

to e7'

N 'a' N H N Co N . -"i Co Co

CO 01

w to N

CO N

ON fV in

Q1 Ln

N rr co

r-! (I) Ln

14 LO .

u) . Ln .

in . Ln

. . . Lr) . .

Ln

. t

.

}ý Gý . -t N H s7" 'd' C

9 ko CO

ý ül t ' Lr, 8 H li ,O O) U r-4 N n st' t0 N

122

-3

T-4 1110 101

ti Q

i-I

r

r

w

C7

ä

UJ

H

r_ p

b A

+ý

Q

r

b

Cý

ä ý3

-4 Cl r- tC) C O CO in

Cý N rl 14 N N M cýl

Q ýý lJ ý1 t0 ()

-. Co co r 01 co

r yý Q r-4 in G1 r M Ol '-i C) r! ko

N I 'ý G1 co t- r- Gl w CO CO CO co 00

<-'l fV in N- in CN in r--1 in r-i w M o c t7 LJ7 lß"1 '-t ýI' C N

w tt- N C) in N CO C' ' c-. M Lr) L-. O

O Q' r in O C'i () i N -1 QL:

O M CO r oO 41 C) ON CT; C r-i

r N co in CO

Ql r L ? -!. in

H G1

' Lr)

O N-

+n : ý:

"" c-J t 1 r)

I, - N- N N -r c, G1 in in ri O to to co to H CA ON C) r d' lii N N O H M M r-1 N H N N

H -. f' H "W CO H CN ;W m C) <D CO in C). co Ol

M N N N M N tV N in N N

gy N . -1 1-4 tý) 1-4 M C O H r-4 CO a sý

to W r-1 d' i in C) Ol O :

14' N

ºý+ " co co Cp. al co of r- r- .0; U)

-W O' in H r in CD VW O N N r . -1 6A r to M b O 00. M r-1 O

M N N N M C+ M C' M C4

y H O M in O U) N H t0 N r-1 i^=. ý

in 01 tD tO 01 0p N Li) O M O }ý

+ a% r N- tO N- t- OD co CC) co CO

in N N N in C) 'C. cr ýO 61 T ul in cri r r H ko U) N . -i N N in (N . -t t r5 N

in in in CV M o0 in t- r-S in 0% CO -4 M M N N

. . º-i

. . . . . rA

.

Ö

. . N

N

.

N co co '9' ) lo " (V al tO in d' N co in rl O O

in N N (V M M C'V N M M M

-rr c: > cl, M 00 C, 4 -4 in th lD t0 N ON co Ol M ' t' M

QQ a1 00 t- t- 01 a1 QO Op 00 co co

A ri N Ol N N r-4 O to M (VI N M M C4 M ch

M , I* N N- H in

O M

0' 1

O M in (n Co N C 4 N in . -i 5%

i ö cri rn ö ö ö ý- 1-4 -4

ON 0 -4 r-i

rn

i 9' -tzr H Cl) C U H co U) M O t O to N co, C'-) M ºn

N H H N M N H M M

. - M O M in in W -t N m r4

- Co N co H ý N

.

N

. O

. r-

. . . r -I

.

N

. -t . .

N

.

N U tU UUU W

ý W '

Oj q N (r]

C) C> N

o ) U) in t in r-i N l2' W U)

123

)

r-

ix.

94

A

Cý Q

A

rz-:

F r . -t ýý

ä x

w

Cf 1 {M M tr M M t"a f'} M

i o co gyn. co N o o LO C) o cv

ý! 4 ý"t 'C Lri M Ln LI; in l7 LI)

UN . z. CO CC) uö c, CO cc m to Ci co LZ

N ttl CO (V 01% O N O 'd' -a' Gl ^ to <f' N M iý l0 ST C Sý Ui c")

x N Cl CD M M M t-) M M CD CD CD

^i N r-A r- N ^- N N (4 N llýr "+ Q4 - in U) [`N N LA CM in O CD

C3 . -i -i N r4 rt r_4 r^f N r--1 r-i r-i 1 -4 r-1 H H H H r-t H

Gil t'7 C Lrl ; J) If) . --1 N .T d 1-1) c0 in Gl ['- " rI) C mot' m LO tN (N

ý, _ (") t' (N w N l0 \D

! ") H M O co -' U'1 N co CO al $t Cl Lo 0% co \U VA N ON 0 -4 -4

ý'1 .

C .

r*1 .

N .

ýf` .

V .

ý7' .

N . . . .

0) N in CO M CO OJ lid r-1 C7 to A 0% N O O '. ' : 'I 0) 0) H CV O

N M M M M (11 N N r'7 l) f")

Co N CN [I) r

In N

N O

m N

r-# 0

CO GU

U) M

t0 N 1

CO CO C0 Q) CD oQ o? 0) CJ CO q

(tý -W O O CO N N r-1 Ot l0 C) N U) L') CD v'' d1 tt) M to CO "C *`

O M (M D cD) M to M M M M CD

Q (I) Ov O N (7 00 to ON %D N l0 lU

'C in CO . -1 N . -i in l) r1 d Ö

(ý O ri r N H H ri H .-! i _i 4

. e. Zr) N Q' N N LO C) l0 C% LID LID N N l'- H %D In N H N N C' tV r-1 (N M 'N H M N

M 00 co [' "Cr l0 1-1 '4t: r rýr O CD N co CO O ui M H o". co C' %D M cf' I! CM LA ti' C' N LA C M

r-I %D 0) C) UO N 0% N 'O 110 lO

. -i M C' -w Q7 V' H e-1 N N (14

Cý f; C14 114 C4 C4 C4 C' H

M Rir 0X

CO lD

if) 11)

C) 00

N 0%

01 lC

1-4 W

01 H

N 00

C) r

CD co a7 . 01 W OJ CO 00 01 06 C0

Q Co to U1 C co 01 to . -4 q r 4n N M Ln ko Ln H in %JD m M M M M C4 r'; (VI C4 C4 Cý C4 CN in

h ao

N CO

N t-

) CO u"1 to

03 u1

a1 c7

u) N

W to

P, .9 C, Cý -4 4 r4

d' H lQ M co O H CC) '

LID d' lD (ID O %0 4 to (N CO f 7 M to `cr N l7 N i-I H N M N H M M

c l0 M Z CO co ý L .0 C-4 0, 0 O al 0) L f)

'i Cý

ý}a6i ý }nýjj

ý ý ý"-Y y}

rý""

` r-t U cV t a ý` 10 C' ýr >4 U o

#, o CO g

i I `_ Li)

vt un L n

8 to CO

r-I N "G' %D U)

124

ýr,, ý: ýJ

N

l

. -1

ti x :n

ate.

fl

64

ti

cv)

G1 UC) F C)

ýT' M rl

(N C

M tý tý

c ýT M ýJ ,

H) P--3 LI) r) M rl c) r

C) O N

N LI

Ol (N

C) N

C\ GJ

O t0

lD -1

N LI)

fam. l0

M co

O t ri

. -. N -4

M N -4

r; -1

-4 -1

ri H H H .4

. --i r -t

L^. q M .I Ol co lý O r-{ (N [) fv

Ef) M M N N M M: M M M C) M

^ lD tD N

ßl 'IV

C' O co q

O N

M ri

M (

r-- N

J 1 4 N

W O N M N N N N ' r-i N N

1 Z ON co

M LA

O Gl

1-11 N-

_4

f) M

fn Li) M

-4 d"

N C

Ol to

C) I-

tf) N

.4 N- v

ß1 N

M -4

N kD

M d'

61 Ul

d1 01

C to N CO

r- N C v ý' M l0 Hfl M V' U' to

Q to '' Cl CD M' 00 O N r1 r 1 0 0 M

q U) M m co (V m M M M

m rr) M

G1 to [- .

N r- N N CD CN co ti l0 N N r-I N Li O CC) r-1 O

G+ (n .4 N M H H . -i O -I ri 4

,A .4 . -i .4 H H r-I r-i . -i r-I r-i

t0 I- co a'. to N CO l0 V` (V C) " Q CV .

0) .

IN .

C N 0'. O M N O ýy I

Q ch N N N .

M M N M. C. j

+ c""1 ý

r! 00 N

ul CO ý n oll ý

li'

w O . -i

-4 r-1

M M N . -{ r-1 r-i N

Z LO U7 C-4 N N

.4 t. 0 U) ('4 -O CN

N ýi' N H N M (V .4

M (N

34 N 00

N N

N N

N ,

N ßl

lo C)

O C)

Cr) V'

N N

a1 a%

03 Cl

cn .

-i' .

M .

d .

r .

In .

M .

v .

C .

M .

ýS' .

6 tLi c(N

N rr-4 9 to

M M M C

6 C4 M th M M M M M M ; M

fü L~ti Ln to (7, r4 co

N co "0 N ýi L) C) O

W N

C) rt

. O 1 N C4 N N .4 C4 C4 Cý W . I r .4 . -4 r-1 -1 r -1 , .1 H V-4 -4

A M C14 N

(14 O

rn a% n

El- '. O

co O

m N ýMj

-4 M

U) r7 t_; (4 N M M M Cß'1 C4 M

N O

co tD

LA N

(0 U1

(") r-S

O f,

I- fý

N U) ()

N C'.

r. ) , ( -: Cý h N th N

cn N C) "W to N .4 1.0 -A " N CO M cy) CV

M

-4 n M (1) ('ý1

:J M N N M O O u1 (- M r-

tf) to V' M N ko w mot' to t! 1 ill

Uýl ý 'ý ý ?+ flT CQ r-t N W

ý ý

Ä ýr ý ýý-1

ý a U U G7 w co (%j ICV to 0

1ý + Q lý rt rt

,n v Ell H co

)

.,

H

1ý + ( ý

ä" ý

cr

r

%-4 X N 14

£+ 8

Cý ~ r"ý

U r3

ti s-ý

w i

ul

Ott E0

2

1 $4 1ý

a

w

[i AN

q

iý m 2, -9

z

$-i

(D -

1-0 t Co , --; 1- t: . -i ci

C') 00 C ^' C'S LV (*) CJ

MNO

. -i -4 -4 '--!

-4 O [- ON. Cl c`1 C) CO

O ON w cr. G1 _j CO

c') MN

ON o 'LO (--; in Qi i c3' CO '1' --A

l! 7 CQ 01 E*- C\

lD lfl .Oi. '

O V' r1 N

rt tn to o0 4 4 4 4

M O . -i ;: f' Li 1O 03 Ol

U) C M -i F-4 r-i

to C N i M ttl lfl Iý

O to M CO . --4 fV

S C Ifl

. If)

. ýP M N

. --I

t11 CV ý"' N Lf1 U) N N N C N

I- O C") CO co O c') '9 mot' LO to li 1

r-I . Cw) O C O Cl c!

L! 4

hfl fit'

M O O IN O 00 N O

C) r-t to to r-{ O OJ G) .

. -i -co 1fß M 0) Ol N

ri rý r-1 ri

e1ý r4 l7 M Cl) Cl to O N C N r-1

tn r o co 0

w 1- to to

2 2 2 2 a cr "; r N V ýO

I # u'1 ul N

r-i N "; r to

3 lÖ C' ) ý' C C) X31

Mv cý n N f--i CC)

CV C\: rl M 1 . -i . --1 -4

Lr) M co N 0

Ln 4c4

if Cl rý N

NNM -4 r-1 r-t . -4

LO 1. -1 -4 " fý) MC Ol

N C\ to N G1 Q' to M

tT OM al N r- 'c1' CYl 4

'4 fit' 4

[") Ur lJ .

M cr ýt) H r-I H r-i

N Cd aN C *% UN

CD tl- 1: 31 03 DO d' Cil

CV Ui OC N . -i to U'1 . -i N (1) lei

LO 07 [N c7' Nm two t! 1 '0 Ul Ui ! f)

M . -+ "r o 'O o CO r1 ºn v. ui

t- rn O cr1 ý-1 MN

CO .-M fV O 0) a1

CO NN CV W CO 41 Cp

r-! 4

C> OD t0 N 00 M r-i C'S (Y) N

I ß1 mot' N

Iý l0 lD \D

Ü Ü Q

8 U)

12E

ý? G ri

N C} Cv . --1 , --f . --i

Cl Cpl lý CD G : -

Ll v; ": z

ON C", r- o co rri ýýi cv

c Co in in r- N CV lD .D

C Co C' al lD C lfl tD o

co LO ýT ko

Ln r1--

110

444 - CC)

ýr ýr cri

C--, en U) ra ýý Un aý ýr ri N O' Ol

ri 1 r1

ý: ' lD Ol

OGN cn ILO lJ Ul U')

O r-i O

U7 II) Lf)

O co O U,

e-1 N +-1

co Mq

dý to dý

MO U) ill C) O

N ri C, r-; r4 . -1

Ln 'KV t. 0 M LO d' r-I MM

U) CN

C> C: ) ri -, A r-li -A Hw co

lb

APPENDIX 12

ITEMANALYSISDATA0NTHEHSPQ

127

'0 2I Z

NN J

CD

ä

0 U-

J

a 6 22 w a J

O N Tq

W

H

* 7_ <

WS Q 2 1- < - üc of (1) -<H W- O - M O - I cV 00 - - In In 'CT ON In N - O ý- 00

-M of z ci L. 1- - OX to

2W

1ý v ao p r O r N in r ýr ýt - ko o) p-, co "t 00 0'

I0' IO N

I M r

I r M M

I IN 1°' M %0 N r

ý J Xt

I 0

Cy O O M N K) 0 (N - O cV - 0 M O O O -

. qr co co C, 4 0 14'% I I N 1 I- - I I 1 I O

I I I

d - O N N N - N O O (N N - N 0 O

U) (14 It'l EK) Iin '

M rv

«) %0

20 I ao o

I p r

Iý O I j q% Orn Nr

r jr- I- Ißt W GO. N 0 O O O

N _ ö O

J () M M til ýt M ON r- r- in %b %0 /p O% O ýD cf N in CO U) O M

Itfl I Iv IN r %0

I I 00 Ln

'rrl in

IO Orn

Ir It11 N

IO In r

I O O - N N O O O O O N -

p O Z

M r Oý p N r- M to " CO u"a M - %0 (V - O Cc Z -ý

I I I p I J I

N I Icv M O O O - O O O O O O 0O O M -

0 -

W lo ýp IIn -

tin in I Oý N

O I r Iý lc r O\ 'ct N1 Q% I M - 'D O I '. O - 0D O'.

o O O M - O O O - O O

IO O

M N in pN '. O - º^ M 0 M ON u1 N M Cp t+1 %0 0 3 N Irr R- I IN I IN I I I

- O - N O - O - - O M O Ö

N - L J O CC) N ul N N In n %0 O' M t0 M 0 O CO CO a O

Iý N -gr

I 00 M r-

I O n Ln

I - 01)

IN %D

I- N) ý - 00

I- I

O - O N O M O O N - O O - U. j V)

W ON pr) - rn 0

I- - 00 I N N CO WN

I I- "ýt

I 9: r U'%

I C%

I (7%

I trN I IN 04

ca Ö

O O N 1- O O O O O - O N

O O - w

w N

' v I 1ý Cl) in fit r - 1t 0D IÖ 0

00 00 %0

I %0 (N

I r 00

N Ul

i(V -

tV N IN N I O O N - O 10 O O O O O - O IO

o N 1

O O M I I I

r- Nr I

r- (N 1 Iý CO

r- I

ko n I

N 0 Ü O N - N N N 0 N N O

N -

p !M

_ O V

N Lrl I ý Iý O

I I I r

I : zr 04

I O N O 0 M N - - ct O - IN IN O

0% wi CO Ö

I0 N # o 0

co 0 to

I CD 0

V) I 2 I l O 0 0 O O i0

0 O

0 f-

N Ö o

M

C14 I I I 0 I0 t- M

I r i jLj-j I IN I0 I 0 fn I O O O O O O M O N 0

O W

a J W Q

ONJ Gý W-Q c/)

Öa Z '0 O O "D O '. O - M '. O v C)N (yý

LY L M

N - N - 0

N O N N

00 N

In Ö

O cn Of x-

p M N - - - - - S U) S U

OQ .i Ou Q

Ü Ui-Ui N

5 Z N ;t Cc v' to 0 N ON a% (0 N '. O 0 N N 0 c% M %D M ul Öm g L. r O O in 'i 4 r, a 00 r, 0 N U-% O N - M N O N O N - - - - - -

Üýö -

CL 3

W OD OD tt in (X) 'Kt in cT ON 0 CO u1 0 0) N Il In CO 0 0 O in in Kl d C in In In 'd' M 't in 143. Ill vt In cY ct V . Zt %0

x W6 1-

N Nl 'q7 if) %0 IN 0C) ON 0 -

- N M IT In %D IN OD ON O - tV - - - - N N N

128

Z ...

CD Q

w

Z- 27 -1 oQ

1) w

r F- J 4

iQ

W 0.

J

'r 0 6

W J Co 4 H

*Z cr QJ

:uMQ U' I- -Q 1- W

N - N O - - d N N O O O O - N N - - V U-) O

ZýZtn L - }- -

OXm zw

D N N N N n 00 M CO CO t- cr V 00 M ' 1- co 0' I I I IN I %D N

I I I 0

C>1 O O O O O O - O - N O O O N

111) n q 0 N N 0' N O M -t ü1 "O t[l Iý M 0 rn

I o D 'O

r UD

ýin co N i-

tO ý Iv I'

III I"

d Io O O O N N 10 O N O O 0 N I p Q

ýf tN - N Ul 4 O M 'O Nl N N ;; t N ON C) 10 CO CO Cl) N V

I N CO il Kl 1.0 in O'

I 0 M 0\ \O ül

I 'O

1 CO y- -

J d - 0 0 0 10 0 I- 0 - O O - 0 0 O - - O O

C) 1- N "D s O M 00 oo r n O L(1 ct co - N rr i N r, O N N M

I kO

I O t0

I l N O CO ct

I I` - N N

I %D (N

j f- -

O O - - M O - - - N O - - - N - N O 0 Z

r- 0 f` M in - - d to _ N ýf w) N n (-4 M z M rn 8 -

I Ö p C14 M Q 00 4- K)

I 1 Ict LA

- O 0 Q - O O O w

- i N co N zO C) M p' in M Kl O' ýt 0 - N

I ( I

U'N I O 0

Iý IM N teý %a V)

I ý p

I

n N O O N O J - O - O O O O

;; ý S (7%

I CO

I 1 I IOS I I ýO

IM IN rn 0

- O N - O O O N p ý

L J 5 0 in ON f-- N - N CO CO CC) OD N - N 0 O n Orn rt < M N U) o0

I M ýr

( O O M O O

I rn ' d %T

1 N U1

I -

N -ý O O O O M - O N O M N 0 cr- w N Q N in N M in CO _ in Q (7\ o\ U1 . i- M co oo in

U- IN C N ýt N ý1'N N M t[, \ tT (

N 1 -

I N q: r

I 00

I CO N

l M

co i0 O O O - O - - - - O O (14 v O O N O O 0 w ( "t W OD CD O - . '. T M Q GAD M CO N - M -zt ýt I

M N J

0 I I"ý

00 NT I I Iý I I I

ö - - O - N O O O O O O

O N (N i r- N IN IN r- in 1! -\ ap O (: )N O c* in ý2

O j C7, CY\ N CA

j I-*- (N til

I - -IT rn

I I 1 -

I 1 = O O O O N (O O O O O O O N N -,. T N O V O "It

U I Iý L

%D C) - O I I0 I I ! I (i

N O O O O - O fle) O O - N i O

co n (X) in i- T d - in ýo M M P- M

CD M l I I I I- I I

0 I- IM I

O 8 O O O O O O O O O O

0\ co rn - N CO in -; t n ýt in 1 - in in O rn

Q l I 0 O I

C, 4 I

0 N I

- 0 ä ö 0 0 i0 - ý - - ! 1* - - 0

w Q J Ü

0N -J w - 4 u g

w is

O Z ' O N tom- 0' N of - in m in '0 0' O in p' N `t O V O' M N N M N CN r- M3 Cr 0N 0 O M O M d 0 C `N r-4 N O N

_ ÖÜ ' m x

Cam- 1-- O

pQ J

w- 4 0U)EX v w-w U) G_)

N Z in M CO N Hfl Co in N N ON - CO N in N, - in - [t O 00

L. M M M 0 r ý 0O 0 O

_ O N O N N - M Q NN N N N - - N O

ax i- " OQZ z 0-- F- 00 a 3

F- lZJ Oý i W N +7 N ýh CO

N CO ( CO q: r mot' in in IN M U) in

N-

W Kl et ßc1 % r 0) 0" 0 N Nl 10 un 0 t- 00 0' 0 - N M

O N N N N N (V N M M M M M M M in IN in -Z

129

.. w M

a

z

J

0

0 U-

Z cn

cn W

rv,, t- zo

z !J

0 a W

J w

04 J

mw I--

H-

Q J _ w

S F- <- (D oý O N

Ui - N M d- Oý N O - N - Ln M PC) 0 - -ýf (N 'Z- pr) 0 O 11 ZV) S L 1 - OX u]

Zw

I00 6+ d IN in ul i- a' a0 N N CO N c, 4 CO %. o /O I I- I0 I I 0 0

1 in

1Ö

O1 O IÖ N

- O 0 M O O O O Ö Ö

N n (N CC) n tf) I I IN I I- - I I 1 I- r

I 1

4 O M N O N - O O - N - O O O

N rn d O O 0 0 rn O' C- N O i t' r r ýl V) N 'D %0

I I a.,

I in

( co

I O - Lf'l 0 0 N OA - of O N

fi

J O" O O O O - N M O 0 O - O M - N N O O N - 0

U r (, 4 O n O /o d v r /o IN O (n ýr oo O

Irn 1 O 0

Iv ao

I(N Ir Iw ýr

( N a' O r

I N co

Its Ip f K) O (N O O O N 0 0 0 O M O p

O Z

V a' O in 0 O' M N N o c0 M a) (V rýl tp u\ N a) --, j co

I Pr) CN

I 0

I- I I ,

O !O - O 1 O O O O O 0 - N 0 O 0 - 0 I -

LU O %0 'ch - p\ in t_. O\ N- M 1 O u1 CO O N tO W

- I - l O t 0 IO I- I - - o o o - o to o - - g o in _ - 1ý tfý - O N r Oý V N d (N M d \O O\ O O' d

3 1 - - I -- I I - - f I 1 N In %. 0 l - 0 0 0 O N O O O d 0 0 L

\p co O IN M r M If 'O - O M ý %o a' N 0 w V 00

I I IN I

O I IN I

0 W) in N r - Ln oN O I

d d - M O M - - O - O O N O O N N - - O - - O --

w M i1 1 M 0 ý1 M O D O --t r (V in

U- O to I - d O Ico lqt I Lr, 0' iI , ( ö c I IN ý ý 1 tr% CC) 0 I co I CO O C'4 N N N O 0 N

- N M O

O O Ui cV ! fý Ip o (V "o O ýO c:, m r '. p N r O N try a'

V) w I

N I IÖ IN I

Ö 0 I

Ö I IO IÖ

Ö ý O O - -

Ö - N

Ö

Cl) C) 0

IN Iý r - a,.

IN I 0 CO 'rr

I w'T - (N

I IO C"i O

is

0 p O O - - M N N O N U _ O C11.1 . zr -; r ct

U , ` I

O N r f I I

0 f I I

O N - N O re) O O N (N 0 (N N - O O O

LCI Eft r M M . Ml 1r) r M O\ a CO cr 0 p\ CO , a' 1 I I l I I ý * O O - _ m O O 0 0 N O O O O O O

CO r. GO U'\ I; r

I m I N r-

I ý N

In I_ I ` I_ I I_ qT UN

0 O O N N N O O NJ N C) 0

w O J 0 'j, J

Ü

W-< t/) - V c r Z rn r 0 r 0 v co in r CA CO in OD 0 ao '. p ao tO O r - L to N d - C\ v 0 IT 0 N %0 -t - rn in co o N WO O - N - M N N N N t'1 N N M N N

z 0 co U- F-

O -

J w O< J W-

OtoM U W W (1) R ÜO-

z N in in to M d \O N O d' to to in in in - to in - M WHmL. 3 N- - kO M co O r '. D in OA 00 - to u1 . - (]\ in Ll - 0 M

_ 0 O -- - - O N N - N - O - - N N - N - - (N -- OUz S

_ Vo0

a 3

W N- N N in - N O r in M %p M - 0c) N O\ O r O 'ý F- 1 CD in to tt to in -3' M ýt in to 't u1 in qt In i in 0 OD co

W" d in ý N 0) O' 0 M . in to r op a' O - N M ýt 1- O - v J v v Ln LA Ln in in Ln in in in in tO to w .0

130

M

Z

J

(D

X7 0 U-

FS

ui w M

J QZ

W

N

1- _

v. J Q

*Z <J

W 7Z < 2Q- 0Wa U)

Q F-- W - - 0 - (N - O O 0 N C' 0% ýt O O O M: Of Zcn i- - H- -- OX co ZW

0 O MO - O - O N N N ,; t rf1 , ýp +ý 10 rn N In In of

ýO Kl ct et 0' 0

Is (N

1 I- IO w t-

Ico

O r1 N O O I! M O O O N N N O O 0

- 0 [- M Q% -

I IO i0 N

I I0 I Co In - (O M

I In

I 1 N

1 U-N Ct

It In

I ýp N O N N 0 O O 0 Ill O - N

co -gr M U1 n t0 ON ;; t '7 N I O O ko N N V N 1tß N N N I- S -

I - I- I -zt ON - O% N d N d O 00

(

- N I J O- - O N N N O O O - - O - - - O N O IO

0

U tko CO N - C-4 ON --; I- 00 O %0 co - O ON N c. n et M V) O (I'`l -, t %, o

I d n

1 M

( N

I ,t In N

( co

I

- O\ I i

co 0 ( M N N N O O N O O O N O O O O N U

Lrl (, q 10 o) '*', C)

z U"N I

0 - ýo rn 1

M l

Ö I I

- I I - I

O O O Ö O O O O O O O N - O

W N O II UD ý N d - '° O I C\ C)N (N n . I O

I I I I I I I I I

O 0 - y- - N O 0 O 0 N 0 O O O O S

to o oo n N 1N C' O - %D - Wl ýD In IN In O r- In ()N 3 M: I l IC14 I I I J j

N O - N 0 0 - O - O 0 O - 0 O - O O -

M M K) N In U\ O 1N U*) N In 0 O - C) N IC' I

_ I IO

CO I to it

I u 1 _

I O O

IO 1 O co 0 0 N

I Ißt OD N In N 0 N C) N N O 0 N O O - - N

W V) N C4 %; t In n p n - lL

IN In N

jUlt j 2 ý:

IN i°

I IO N

I IO Ia) p- ff)

IN

CLJ - C14 - - -

O Lij co n C, 4 r) In

W 0 I I I (rl

%D (1 I I I

- (

C14 I

N 0

- N O O O - N N 0 O O N - 0 0 1

, - u*, M '0 N 'O Qrn 00 ' 0 In N O

,O I r-

I

I l I

It p I I I FN

CN I

N r' - - O O O O N O N fn N O 0 O O

O U

I I1

( I I- Lr)

I Ö

Id IO I N

1 ( 10

- I O

O - N O - - N ý

O

co O I

O I O FN I I

- - I- I ln

O O N - O O - O O O O O

n le) cl ; - 0 I O

I I M N

I 1 0

I I N

I I I I 1

O 0 N O 0 O N In - N 0 -

O N

-'

W

W Q 0uJ C) w-< cn UO cr Z ' cT to - -, Zr N co d kO -2' N I 0 In CO 0 'O In p' co

WL 3 'D (T - N N- N CO O M In n C*- r- Pei Cj\ n ON N NJ -tt N O O M N M 0 N Ul N ^1 N 0 0 N C' N 0 M M

e: U- F-

O

-1 W Cl < J W

Ov Ü

W- W V) g ÜON

Kº - t i Z - N N In '0 'O ý 00 0 In % In n ON co t4) 'O 00 kO W 1- n L. '

3 C' M In - - Q% - M In N - hol in N In O' N O\ d' '0 I d O 0 N N N O N M N N M - - O O - - - O N N N - OZ I U

2 F- 0Ö

Q.. 3

W c " F- LU

T N - IT 0' 0 T N

0 In

00 O\ In co (V N In In In ; -t %t N In Nr u1 N I 1.0 'O N In ýY N 'O I -r 'O N '0

W Z In %0

\0 '0

f- '0

CO %

ON 0

0 r

- N- N N

M N

I*' N

I N

% N

N- N

CO o 0 cc)

- 00

N ca

M In o N N co cc) c

131

'0 M

Z

U, J

0

Q

w c

z z 0 F- N w

Q

W N

1- Z

d

J

O NQ

N J

mW

*z a4J WI< I F- Q- 0 ol: of (n F- W O N rn O O O N 0 O M N C' M - to O N rý 0 0 -

OX Q] ZW

fl tý M Oý M M tý Iý - Iý M 'D to r ýt kO (- p I I f t I ` 0

I I I I- ` I

d M O O 10 0 0 INN O 0 N O 1 O O O N rn

N it1 t ßr1 O ýO d ýO ýD n N- r- N to CO ct co M O

I 0'

I N %0

I I - d' N- In

I N d - CD I ct N 0 N n

ý

O" M N M - M N O - - O - - O O O O - - -

CO co co c"j (, 4

1 I- l rn

l I I I I l

W O. c) M - O H i- - 0 0 JO - O N M O 0 J

U In CO in co N O CO M LC\ 00 1ý CO I- 0 -c* 0 O N to U) O in N-1 v

Ißt I O

IN o0

IM (N IQ 9

In IM O' O

I r-

1 IC: )

II I r,

N - O O N 0 N O C O O (N O O N O 0 Z

CO In N I co tp I N M N O N in in I ! kp to

M tpý I 0

I

- O N M O O O O O O N - O O O O J- O O O

I - I I I I I

%. 0 !

O I Io (

O 0 - N - 0 O 0 0 O - O N - N O O 0 I ~

CN --T N- 00 00 N tl Ul Q' M ul 0 O - 00 _

I ( I I- N

IO N 0%

I 1 ko %. D CC) CN O OD

I

N - IN N O O I- N O - to O N - L

Iý v ` ýO M

°ý O

0 U) I.; r

ö tO

° '' ao

8 'D

ö 'O N-

ö 0 lö r

Pr) in

0 O

- }in N) ; 04 - ö O --

cr -

lL M Ö IÖ IN Iý

M C% Iý

O 1 IN

O N Iý IO

Co in U) - tri i o

Ö ö

w co -4T r- r- qzr 0 Cc) co w

I I l I I I IM I-- IO I O O O O O O N O O N O N - O O O

CO r4 M u\ %, 0 M N- d' Kl to - in %o d' to " O

IM I I -

I I I ICO = U

O M O O O - N O O - O - O

O U

ý I IO

0 O I I I I IN I

N N O O N O - O O O O O

r- r- Ln u-1 O g ' N in %D

- N

IO O

I 0 0 N

I 0 a

I I I N

I

, -t l

j 0 - - - 0

I _ N

I I0 I IN I O

IO co CC)

I 0

O - N iO O N - - N O O O O

W O -J W Q

O cn J U W- Q l/)

sý ÜOa Z O' W o' O CO U'% M - ^ M kO in - to O In O' in in N- CO

}-- W L. In qt O\ M

Ot M

N - t

00 N

r M

- "T

0 -

N N

ct O

O O

d O - M -

M - in ý7

cr- = Cl) O M N O - 0 0 M M N

OU cn m

O

J W O Q J W- Q

Ocncr U W-W U) N ÜO

L Z N-

ý ýn pý N- in CO

M - -

- co N

N 00

N- in M - N-

M CO co

N - M 0' It

to -q

in - W F- m 3 t O N1 o - %0 O :f ýp p\ O N N O M M (V N M O O O N O N N N -

U OO

~ d 3

W %0 O\ N O in N- CO O to i - T 00 d- N - in N N

t-- Lei In -t qr qt 4 C: r N in tO Vt N in in 't In 00 in -T \O N tO

Ui O to

c N O 0 - N

+ M '. 4r In to N- co N

O M O

in to o co C D 00 at 0 O rn at at at at at rn O O O O O -Z - - - - - - -

1142

., ýo

Z

N J

C7

ä

0 LL.

W

N

J Qz

N

W IL

in N

S ýp

., -

*Z QJ

WI S l- Q L^ ýý to -Q F- W - CO - qzl- N V u1 0 U -N N u N r - - - N N S 2' Z v) L

F- - OX co ZW

oo - Nil tn

I cr,

l (14

I r- V)

I 0

I V) 1,0

- OD OD (14

I Iý tel

I I

O" i O O - O 0 0 - - O O 0 O N

Lr) l I- ý N

I 1 IN IN O

I0 r

O" O o M - N t 0 Jo - 0 i

V, N r ul C O M N o a' (N - a' M M O O r V)

I I N

, ö v

, co NJ- Ö

I M O

I Id -ý ß

- - O - M J o

J

L) U)

co I r- I 10%%0 ( I I I IO I I I

t-- I

O 104

O - -

ý'`, - O O - O N N

N O O 0 Kl (D Z

O %0 r N O N co rn v . am .,: T Q CT

Z 1N

0 I J

- I I I I I

0 O - O f 4N v 0 - O O - N r'n 0 O

C-4 0 , zr . 0 ON q cc) r re) co I 1 1 (N

I I J I- I J

N N 0 O O 0 0 O - O O - 0 0 O =

- N Uff co N R a' M oo M N n r O r QO t') r "-% N

I O

I Iý I l \O N In N

j ý - In O O

M 0 O 11 º^ o o M O O O M L

co N r cV in "t 00 O t tzr UO U I r Iý

N - r 0 N rt- O O l0 V U)

r IN IN

- J

a' i (N E

- O O - - O M - - O {O M

co

o 1 i !ö 1 \ l jr- 0 - - M m U- - ö l ý j S 0 \. O O iQ (N O ý t! l I d M r r tD ' r r M ýO t0 U) O U) U) N Cl) w N

Iö ö

Iö 1 1ö Iö o

I ö

I k ° 0 o

I H J - - 1 -

o o M Ul 'D M t m rt M O r - u-, U-\

p Ln O p I I Ir

r- i o-. CO

ý T h co co 0\ N Ist in '0 co -r

I rn in II N N O O O O O f M O O D

U _ p M O 00 00 (7, O , 4) ý r- (N r o o V M M t

U ý

CD CO 1 ý

r I

O 1 ý

N ýt N I C+0

t U r

l l M ö

- C14 C14 M O O O C, O d (N O

ý ` 43 a 0 0 N 0 N

- I p p

0 O - O O Q

0 O

t11 r r a rn r V) c, U) - r pp 0

I f M

I U

I - CD r- CD

1 cc)

O N O O O N O O O 0 0 0 N

W Q J LLJ

CU) -J Ü

W-< U) Z , C14 M \O d' UO M d r In - d co U ýY N W

F- WL 3 . OD N

- t0 co d' M O U) 00 - O N - cc) c) o 'C) O cl) O - - - N o

OL`m S U- F- , Cý;

J (L pQ J W- <

Du U W-W N ÜN

'o \0 . 4t in In M M - Oý N N W I-- U) L

2 3

M -

- N O

N In

O\ N N - UO CO V' - in 01 -

'D a' r N - CO pý N

O O - - - N - N - N O - - O - O - N i O - OÜ

Z 2

_ _ V OO - a 3

ý H t M 'ct in V

in M ' '

M I 10

U u s n Q t t0 V l

WO 0 N M "I* In U) r U) rn O -" Z

O -

O -

O - -

- - - - - - - -

N N N N -

N -

N -

133

%0 2

Z

N J

CD

ö

z z 0

J

F5 ä- w a

w N T-1

I-

*Z a<J U1 M< SH<-

Q I-

WN O - U O 1,0 N N d- N e S2 Zcn L.

OX 0D ZW

in - - 10 0 ö l- IN X 0 0 i - lö l G. 0 1 ttI O ýC M '2' O N Il> M 0 N M Co N O fO

laD Ir I - %0 M 00 -

I M M

Ißt N t0

ý M

O> O O O N O O N O N N 1110

to N 0 00 _ 1

M J

p l'r, 8

Ul n O I M

M N %D w O" O O 0 O 0

S O O - N N O - O

J Ü

00 %. O e '. O a% ýO 'o M ý rn MD '. O

- (f) 0 1 l0 0 I- N I 1g ä lö iN - 0 0 0 - Co CO C'4 N O M 00 QD M M N O O N z -

1 I 0 r- r- 0 0 Ö ý

- - w

M V u-% - N U1 �O N O '. o o0 N %0 I I I l N

i 1 - Ö

N - N O O O O 0 - - -

(V N

lo OD

N OD

O IM lu-ý -

- -t

M (

p', r-

l t0 I -, 0 t

I

O O - O O c" i O O O M

J r- ` N Q ° N

Iý Oý

lw i ý

Iý M

1 Q M

ý -- -

IM M m

' O

O 0 0 O N M O w N %0 %0 in N co Co M -e . Ki- OD N - O !O aO

LL. O l

- m %D 100

- O 0

1 N

- pý

1Pr)

I p O Co 0 - - 0 O O N

LLJ U1 00 in 1.0 M ch O% a% 00 %0 00 W W qt O r-

l n iT M O% - O O N

O O N N d ml M N '-O rn ýt O M u1

O p in 0 N IK I

I\ M iý I

Ö O O I N - O N N U

O l n

U i l IN I0

M N I

U i I

Kt !

0 p O N M1 - O N O - M

CD i lö ý - _ 0 - 0 0 0 2 8 0 l0 0 , I 0 Im 1 N

jriÖ M

1 ( l i i lÖ

0 8 O O 0 O N 0 0 N O O

w p J

_ O ccc i Ü w-¢ cn

_ ÜO2 Z rn %0 N N - OD N O v d rn in in '. O in cm

w F- W L3 e N d - ý ý - M N (V - N N Co O - cn O N N O - N O N N - - - N M O V N

On 2 C> - I--

O

J w

w Oc ii0 Ü

gN OO Z

L CD 1N fN M U, 1 N O O' CO N- 2 o Co

O m p O+ r M

O N

O - - - 0

O N N O M

- z 8U

O

v c2 O C

w 0% N tV - 0 N 0 N CO m - O N CO M O'

I- ct M UA st M N (A n M Ln Ul ýt '. O

- CO N M frý M r, CO u 1- N (V N in M M M M

M in K l Z

134

ýö ý0

Z ...

J Q

C9

Q

0

w

z

I

r J

W CL

_J 81

N vi

W J co

Q J W d` Q. ' O 0

aQs wTO =ý- Q- OWi Lr) -< F- f- -- - - N O - in N o' N ct kD n 7 M O tý C I tY ZUL - F- - OW Z0

C

ýp d tý V N- N o' M M1 00 r- Co M O o' V OD N

iO' CO

I 1-

I Nt Co N

I N O

{ co

I %D

I dT Co ý' O tf\

J O" 0 0 - N - N O N - O - - - O M O 0 O

p\ N1 - C, % M M n N O in - O\ M ff-I d M E p Kl C

I Co - N ON 'O N r r p\

{ 0'% ýo

I Ul N

I in

{ ON

O- - O O N O O - O - - O N O O N - - - O

N 0 O0 -4- 0' N N et n

- t0 MO ON O ul h "0

ý 1 a - o o ö - I - I V - - 0 - t O - 0 1 0 F O Oý m N n N N o - ý0 V o En O c itl co t+1 Z O d N-

1 N

1 I I UO N M d N1

1 N

1 i

1 c7 f`r) 197

1 N N

I 0 ";: r

w O 0 - C - - O - O 0 - 0 - O 0 - O N - :E V) ui z J too N N -- O \ ý' O\ 'mot O\ ION h 'D E M M

N i O I

(O Ö

{ O (

0 0 0 0 IO

IÖ 0

Ö t4rt {O

M Ü (j) O U' i- (Co fi { IN IýO (00 Iip IO\

0 ON ýfi 1 (Co N IN N m- 1 Co M I

O O O O 0 N O - 0 O O - 0 0 O 0 aZ 0Q W S

til O

N- N

N -

(14 -

0 d

00 %0

Iff) O

N O

O\ -

Co N-

co O

CO 0

I f- N

., t O

O' '-

Iz II

I ul 00 { d' tN

cn W O N - - O - O O -

1 M O - -- W

Ö= -, r - N IO GQ o' ' N ,t C' d1 N- O ' M o M r- u-

H-- c9 1 rn O

I0 I S N 0

I 0

( "CT (IN

I 0

- J f - - - i v - - - - I ' 3 Ü

-L M En r N 1` M in iz Co M N o' O\ C7% -,: I- N t- w

{ O

l I I -

I V I I 1 O 1 H 0

- O - Ö

- O O Ö Ö 0 Ö

N O

1O

Co N` co -, t N N- N n -,: r -. -t L)\ in W O

{ 0

I { I { CO t Ö `0 - - O 0

O O O O Ö

- O O Ö

CO %0 lzr CF\ t. 0 r Co O

( I ( I { I- I (N

{ IN Co

0 - N - N - N O 04 0 N O 0

04 V) t %. 0 U

t H {

O Ißt

:T t-rN i I Un

IN { (N

1 0

Ia) Im±'t O 0 O 0 N N

O

11 11) r C

0 o

o " o r- o Q O Q 10 ISO

0 N O

In

O N

- V 0 0 O

I n

Iý {Co IM N

I ICA (En 0 00

l

O V O N O - N O 0

I- w OZ -j

LLI - w 0 vý g ÜO$

Z M in CO - N- %0 rn NT C; \ O ko r- M M - \0 Co N- a) W F- F- L3 qzt N. M rn O\ OD co to in in 0 ul v O\ 'T %0 -- ON N

Oj 0 N - - tv - O - O N - - N - - - 04 0 0

OUQ U-0 h

LY_ - }- :L 3

Z J D

Li ü Ü

O I- - cn E 7- I- t) N M - N in Co in r- CO Co n M Co

a I- L3 O N-

Co N in M 4 N CO in tO in 0 N -' N ch tO in r- 0V 0 -gr M M in -; r M %r M ýt M M :r zt in -t M M ct M N

p Ü0 U _ D ä ß

E-- a. 3 2 0 ul d u1 Co CO oD ref Co O\ O\ - - - - M N- 1ý S

LU < N s- in N- CO - IN 00 CO - 00 0' "O to 0 00 Co CO 8 O -

F N in J in to N- CO ON 0 N M -i' in to N ý 0\ 0 - o N

-Z

135

Co N

w J

H

0 %0 tu)

v

N J

C7

d

2

wu ff ý_

Z

NC

J

T5

W O.

I -- z z u, Ci c WS0 SQ

n < F- F- - - N O M N - O - - O O v' O O - N rn Of zU ý- - -

O Z U-) 0

a-

00 N r- d 0 N ýh I-- 0 M oo 0 O to /D N 00 N c? co (V

J O'

1 U1 U"% 0-, (, 4

I N ýt

( l co V O ON -: t N

I CO

I

O" O O O 0 - M O - O O - O - - O O O

M N C-4 - V O' LC) o c\l , - M ON IN O M to u-N QO a) M O

I O

ý (IN co

I N

Ir l UD

1 Lr) "D M M ,t

Icy IV l 't M

IN

O " O 0 O O 0 O O N 0 O O O N

N 0 - 00 N M ' ý ; O N "4- ýt r- C14

I I ( I (M I I I I

d O O O O O O O O O O - O O 0 -

z 7 - N 0 0 N N N %D n N N (14 N

I 1 IN N I O I I I IO N I O 1 M 0 O O 0 0 - ý - 0 N J

l

o I I0 I I I

co ÜN o O O O O O C) NN O O O - O N O O

C) 0 - N N CC) 0 00 C'4 0 (-4

0 ON 0 00 I I

r- I C'4

IN I - %0

I- 1 m ON %10

-

az - O O O O O O O O O

Q Z: w M: "t r-- co l %D r- N o-, N U)

r

UN 8 I-

N p 0 I I I I- I l

N I

O O - ý ý

N O N

I I J I I L ( I 110 co

3 O O - O N O O O O O N O O N (-

o ~ W

I IN M

J O

f- I I I 1

- O 0 0 0 - ,p O O N rn ---; 0

-7 M - 0O CO to o-, N r- .t C o' N N N W a0

I V N CO /O N '. O

l M '. O

j M N

I 0 co

1 N

1 ON

1 O

1 00

1

O O 0 0 0 0 0 - 0 N - O - 0 0 - 0

-zr r- rn O a-% vt ko d c0 M O ko O d n %0 O 0 0 M 0

IM IO' N N M

( O

IN t'"%

III (V ct

I N

I '

III

- O O O O O N O O O O O O N

00 (A n r- U

I I IN 18 r- 0 r-

J - :t

I OD

I 00

( 0 fr)

Iý

O O O O O - - O O N O O N ý0 0 0 D M

CE) ö

tN ö O o 1

N 1

M u1 o N o ct t - I OD 00 I ýO I N I o o , o O 0 0 0 0 - O O O CO %0 N 0 C)

0D a\ Uý

ý - Q %0 IM %C 04 - CO 2 1 l0' I. ý I I O 0 0 0 0 0 0 - O O 2 O -- O O M M O O

Z M

J

N Ü

0 O$ z ul M M - ON N O' %7 - 0 0' N N (V CO - Kl to (V i U L - 11, tf1 ýt - M ! fl M kO N M ýi M U-\ N N N

OU ciý O - O 0 ' C7 ff) N M p n; CV N N Lr OUO =

F-a. 3

F- W Q Z -i 0W Q

ui V) Ü N WOt Z 00 ul r' (V - N - n O N LL N a) N C-4 M N

ý- ý- L 3 '. O N 0 - O M N N It M C\ M N N U1 O\ O\ fý ci

=j O M - - M M ul M M M n ;t Nl Ln N '; r 'Ch Ct t- ýt 000 0

--

F- 0- 3

< N N "t %0 co M kO VJ 7 m O1 O\ r - W C' O' C' 0' O V - 0\ ir1 d -t. d DO kO O ý E- lei _

. .

N N Lr\ to F- Z t N N N N N (V OM M M M M

M M

M K l M

'0 M

ON N ri

Z z il l Of ¢ w= 0 T 1- Q li. C7 lY I N

-< f- E-- - M O - W' - Lf) OD N O O d' O O - M O - SIrZU L - -

f- -ý OX0 ZW0

S

ul r O' 0 N- ZT 0 M cy% V N N M ü\ 0 N I- I 0

I I I 1 O 1

O, O IO O N - O O O N 0 O O O O O

o N N r1 M N N- O d' O -- O Ln N

týl l o 1C I

ýo I- I - rn { VD l KN . - OD = K) I - C) In co

ö. ) N o - - o - - , N Oý 00 , 41) "; ', N O ko to - o vo Co I- K 1 1 I r I 8 tn

O- O O O O 0 O - (4 04 0 0 0 O -

M "') O rn ý % 2ý O O N) 'O O

I I (- - M

1 0

( I0 (N

' S 0 0

O O 0

O -- - O Z g

cn r1 un p\ M ' O rn N N Ln O jO O ko co N- OW M 1 M 1 M ul M I - N

r L11 M M t0 Cp n

1 ct ul N

Q 0 0 0 o o - o I o I- o Io 0 0 o - F-U UN r ýp u1 M n O N t` N O! t') LC% 0 - -- I I IN 1 ( I

D C9 ý O O 0 - O 0 - O - O O o LZ C

11) I Ipr) ( I 1 IN

D¢ O M

O O O O O N O O O 1- O W

I- r r- LrN O O

IO Co

I I I j i 1 I(N Ir 0S

N O 0 O - - O O 0 N 0 0 H 1- L3

0 0 Co N - - 00 N- o' - O N M r) a\ U U- N - N 0 1

I 00 - n n M L[ 1 1: T I I N M

I

L N O O O N - N -

N - - - N

~ N 'V DO 0 ý O r- ct N - oý - 0 -4- M

l ý I o j 0 O 0 IN o { O 1 I 0 0

10 0 J - - J 0

0 0 ö M 1 N N N- O M ko Ln c0 - 0\ O O M

I: I Iý 1 -

I I 1 N N

1 N

N O O 0 O O N 0 O 0

o\ cc) Ln N Q o - N r- rn C4 O

IO N 0

`? (I l O

( Io I

0 O - - O - O O O O

m 0 0 1 0 - ö ° 0 0 0 N - 0 0 0

0 , 0 0 0 N M N N N d' CEO O\ Ln N a0 00 C' N N O o' I - -

I I IO O

O O O - N O O - - N O N

az J 0 (I)

0

i± 9 L' 001 Z 1- _ 00 d N N O - O' O\ 0 M 1- O Co N M N ui ý- L - - to

O0 O N O O N - - - N N N O - N N J 0 =

U-C I- - Of Q F- U 3

DZ -j aWW < WCXM U ÜR

W01 Z U OD to N co ul N ýr 'o 8 - O\ U Co - - L1 M L - ON M ' 0 00 Co ul ul IN O' N -- co Ü

O N N

M ýt M M un n Z N M ýt l11

U Cl m Z

Otr2 1-- i- CL 3

0' 1- U1 1N N O, N CO p N tn co W< - U 1- CO N 0' - N N

O '- N- O' M O C) 0 CO CO -.

W - N M V Ln to N aD 0 HO v q- - . v v wt ý- - U U-1 U U-\ ir ,n Lc) u1 U. z -

137 I

% M

zw wSO SýCZ

<H j_ M N M M 0 0 N O 0 N - O N - 0 N ,: t co co 22ZU ý- -

0 ü 0 ZZ

0

in t0 N ý- a) ý- a' t- M M M u1 O O N I I IN I

C), - O O O N N O O 0 O N O O O - -

r O N N N N M N CO N 0 N CO N O L, -, I I I O 0

I0 1

d i- O IO O 10 - N - v - N N - - 00

o C o a. M u\ - N ON n l- O' 6\ m0 ul a) M N N in _

I M

I 00 ON M - ül

i o O' in - N 0 in N Nl

(> N 0 O - 0 O O 0 0 - - - O 0 0 - - O

%0 N N t- ' O N N C)% .* N 't Oh ,J O O' 1 I- I I ( Ir, 1

Z O N O O O N N N 0 I- O O 0 O W OW N { N M CD aO n O "I' ill N Oo M a0 N - Lr%

-' J 0 0 ö 0 0 0 I 0 0 I0

lö 0 - -- 1 -- 0 1-- 0 0 N

ul CO - N ON d N I\ CO CN p I- co M C c I I I I I0 1 1 I

OZ - O N O O - O O O O - N 0 0 O O

dZ d ("4 O f- w M N r- - ýO 1' ) ,, O V l

2: I J I I J ILn IN I J

N I

W O M

N 0 0 O - O O - 0 O O 0 U) W w

o' in 0 ON in N Cs /O r- a0 in (7% ý Ul N CO O - CO (D I: r

- 0 0 - l - J 0 0 0 l ß 0 0 - - 0 0 0 1 1 1 - , 0 0 O-

kc) - 10 L I I IM I I {aD % O 104 I N o% to I 1ON I 0 N O O O o I- - ö I-

ýO V N in O' N ul t11 CO L[l O L" N I- N W

IN I I 1 ( I 1 I

0 0 0 0 - - 0 O O o - - - 0 0 O

(7, (N 0" N (CO 0 O

1 S

I lo Iý 1 I

N N O - N - ,0 O O O N

f'f CO CC)

0 'D IN IN

co N I- IO ( ( 1

N O (- I0 I- IN

O O 0 N 0 O C) 0 0

n Ul O N Ln - r- V) ` r- m )

i ! (

I N -

n CO Kl I r-- 'O ýf N OD

J co - ;t ý ý

O O O - O O O O 10 O O O -

rn in in N N ---t M ül N ý Q, 0. ') a)

< Ir

rn C: ) tfN I

0 1

00 "t O\ I

r- I

r- I

V) I 1

M O O O O - N - 0 O O N N -

J OZ

O u) Z C) Q N

- M U-1 N et Q\ M N ; Zt -4t O N N O L- r N N O Nl aD CO N O - N O in N N _ u1 N

OU O N

N - N N N O N V N N M O O r_ T ZD I 0-, C) _ v-o Cl:

I- a s

OZ J W w W <

u '1 ý F- - cn WOl

Z 0 M N - CO in ýY' N M C - in N O\ M CO in CO in (N (N v in rn - in N rn in co O 'o O -

Ix OU O ýt v in Pr1 V' N t ul M V ýY in in N OÜ0

_ Z-0 F-

w F- a 3

i ýt N K: r ct O M X11 N N N CO N O U-\ N N ;t O' W< 0' M - in CO CO N ul - Nl N 0 ýO in in ýh ý* t0 N

-W F

-i - - - - - - - - - W O\ 0 - N M "It in 0 N OD O\ 0 - N M V in ýO N

O in %0 CO %0 %0 \0 '0 "O %o tO v0 N N N tN N N N N Z -

138

''o M

Z

N J

0

O- z

ýS 21 Z }- OL

r J

F5 U, c w a J

N1

W J

H

z w

w20 Uc1 cn -Q I- H- - - v v N O O O U 0 N G' O O O O O O M N S Of zU L

oxO Zw0

(if d

p OQ N 1N M OD N IN o N M CO C, M C M N tp a%

I CD

I CC)

' 0 %D Ln 4 N

ý0 ýO N 0 o

' O

d O - - 10 0 0 - J - O 0 N

0 %D v co O %, o N n OO %0 O 'r M 0 co N Wl 0 V

' N

I M ON

I I ýY N - 00 N - I-) Lf

I p'

J

d - O - 0 O N N N N O M - 0 - - O O

V) rn -qr N ff) C, (14 ON CYN 14)

I IN S Cýj

IO C14 -, zzr

I I

N r- 00

IO r- -

I0 IN 0 0

d 0 O O O O 1 -

O 0

I- 1(ON I ( IN 5 I J

Z 0

0 10 IO 0 N N - 0 O 0 0 N O 0 LLJ

U) N ,.: o - OD 0

(Y% NO ko

- N p M L

o' V)

co OD " r

0 I o p o

-- 0 0 0 n

I0 0 0 0 Q

- o o 0 ,0 - J0 O

U to ' ,O O N ii ul 0

O' 0

C' 0

IO CO Ian co OO 1kD OD r-

co 0 pýO

I 0

M 0

M 0 v

C7 O o 0 i0 0

- O I oz

_ a- Z - 0 N N N o-, v D %, u1 co OD (7% N ßt1 a: 0% q N

1 r- 0

I- I M

ýI ý

0< O O O O 10 O - N O O 0

Lu M: LLJ "'

- (X O v o r- cc) N M v 0o oo ( rn OD v N m (0 N to M

I O t[1

ý M O' N n OD

I 1`

l n OD %D Ln i

I

O N O O - O N - O - - O N O - m

U Li. 100 I

M to O ýT it11

O N I IN

Ln CO l

r- 1(14

10 M r M N

I- ~

OO OD d' ()N ý OD try kp M %0 o 00 C O aý w - ko

j - U-% fn %. 0 n r-

I rn r- co

1 C-4

I- C-4

1 p

( I

N O O O O O O O O O O O JO N O O

Q ill %0 o' 00 't CO N M ul 00 N N O 0

k4 I I I I I N

O - 10 O O - I - O O - o

00 -

U O I '

- J J

- - N

J O O 10 N - O O N O - O O N

(7% % N "o NT tn M M m N

f 1 O \

1 d' - p\ O OD If\ lsl

O O O O O O - O O - - O 0 O O

Cq 10 r- V) tl'\ r- r V) qr kl) -ac %0

I cc)

I i N

I 0

1 0 N

I0 ISO

- O 10 0 0 - O O 0 0 JO

w az .j - Nw < D

w- v) Ü1 Z N (V \O co U-1 OD CO M tD O - M - N t`N u1 OO

w 1- F- L" N \o O O 0D v tO to rn - ao ao M 00 rn co n N 0 of O= 0 - - N - 04 N N - N N 0 N C N1 N N K) 0 N (Y M U L0 a

F- a 3

t- ui Z

Cl W

W wN U WOI

Z %D N ul N O 0 n Nl M W% %D O co 0' U'\ rýl %1' ix F- F- -3 '0 - tll 'D CO CC) N O 0 - N 00 N N \D C' tO N 00 Ö20 O M st M 'q7 ct Ul 'rY U, \ (V ' n w 'Cl- t Q' X11 Pr) M

L) L) a m

i-- a 3 M M h M N 00 N n %D OD r i

ULJ O% OD 0\ N - N

0 N N 0 O N l W

w CO 0' 0 - N M d" 1[1 tO N co D 1-- Z 1ý N- CO OD 00 CO 00 CO CO CC) CO O D O\ O\ O\ 0'

0 O\ O'

139

'o M

a

z

N J

0- Li- F 2 W4

-C Z1

L

i-- c

t- J

W CL

J

S U

F-

~ i

w

cr-, c Li M:

oQ1N -< I- 1-- - ON M (14 CO 0 tO M 0 0 - 0 - 0 vt ýY 0 in SZU L - - - ox o Zw0

Of

00 Ln V-1 CO I ( I (M I I I I ( I (

O" 0 I- O 0 O O - N 0 O - - ( O O 0

- - t1- co IO co

I r-

I 0 00

_ 1 re) r-

I co _

{

O" O O O O - o lO _ - O

IÖ 0

I IC I IÖ 1 I Ö

1-. t (

IÖ (Ö in

I , Zr C).

0 C) N Ö 0 - -

p -

v co v co mit ý u- 'r N O O cc) to Ito un - in O I 1 Ip I(

pý I

C14 K) (

1: T I

00 I

N I- iO - 0 0 0 N 0 N O - - - O

0 0

Z LLJ

(J) M 0 N N -t to O N to - in N N O O

r ý

O

IO

O

I -

O

I IO I IO I

Q O O - O N O

o 0\ 0 0 CO r- ON N 0% - 0

I- I

US r- r- I I-- I 1 1

- I

U' 0 - O O O O i- c'1 0 O O O 0 OZ -

a- n co 1.0 04 r- a\ Lf. ý cc) N N 1 1 I- - In

{ fl-

I 0 rn

I \0

I

OQ 0 N 0 M O N O O W

0 0

I I0 1 I I I M

O2 p O O Ö

10 O O O O O M

10 O O 0-

cc) rn cc) c\ fn W

I0 I I 1 lCA I I IQ I1

L O N - O - N O 0 O - O O O N ý

~ N a\ 00 O h

g 00 M h o\ ¢ CC) N

w ko GO

ö IM I N M ö co pel 1 I (14 fn I N I U-\ I I I f O O O O - N 0

0\ "0 c) 0 tO - 0 ö ) IM I o Iý IN Iý 0 0 IQ 1 - N 1 n ý0 co 0 N V

i 1 I- 0

I I N

I 1

- 0 O p O O O J- N N M O 0 O

tf1 - N cy% in Q M 0 r-- M 'O N m

ý N CO jr- O

I I - co

( I j r- U-1 N O O 0 10 'O - 0 O O 0

Ö O

Ö O O O O

o m pn 00 I (7% 0

I 00 %0 Ln

{ r-

I r-

1 M

IN rn ('4

I

JO o O 0 O O - 0 N 0 0 M

. 10 N

C) Z J

Eii N O

Ü

_ U) 001

L Z - - N ao

M O+ h

O' N 14-

to T N

0 0 to

N CO

M - CO dr 0 O O N Ü Ö p M - N C N N tp co Q WO0 0 O O p N O N N N O N N N N 0 7

p0O C)

Of Qý I- a- 3

WZ O

Q

w N L) Ü U, WO1 Z il 0% M p' M 'i' %0 00 - p' N h tp p\ %0 1-- in -. Co

3 F- 1- L M OD N co - 'IT to 0 'I' in to I- N O\ h - h 0U p M N N (V ý7 M "T in ý' M M M1 qt ct M d M

M p ÜÖ U Z

F- 3

co I- in in in L1 CO M M p N tp N (V co 0 zr - ýt u W o% 0' r-- O' 1- a0 0 to -3- in to in 00 O h N cV M p . . . . . . . . . . . . . .

W6 I-

1- CO aý S - O N M

0 ýt 0 Lr\ 0 to r, co N M T in

m at c7% O O O O - - - - -ý

140

0 M

zI Z

WX0 =Qs

N < I--- I-- - M u1 N 'D M M - N -ýZt O - C M 0 M M d M N

YCz c) L OXQ ZW0

tY

.;: r O ul N jo' O CO ' 1,0 C) 0 ct -

CO -k ON t ICO r- M jN 'IT co rr-, co r ý M N b 1 'ct ý I is N Jr ' V N I

\p -CI- zr j co tt1

0 0 - - o {N 0 O O N - o 0 - {o o - - (N iO u O N rn c o N pr) u N IN a' IIl co o N N

T 41 I I I I

O J I

% I

ff) I t

ON I; zr l(N

N O' !O - O (, -1 N O O - O O O - 0 - 0 - N

0 rn K) 0 N N 0 o co o O }O - (N

(O tr% d N I

I d

I o

1 vý)

I K) OD

J ý pll to d

l 1(`J N )t0 N

O O O N M 0 0 0 O lO 0 0 O O `O N

r - V) N <1 n M d %0 M d O "') Im ICO N IOD O d' 0

j 1;; CO M - n

ý N 0

1 I- p CO

ý CO CV

I O 10 M to

- O N N - O 0 O M - O O O O O - I N - 0

p' ill S? N O N O %0 M CO u1 n - O co -zt

-' I I N . Zt o I o 0 O I 0 I

o o - jo ö o N o -- J iO o o 0 I- W M CO N Ul CO 0 N M N d N N CO M - - N N

V) - n O N N 1 I I

d' N I

N I

kO I

M I

to N N OD OW O O O O O O - O O O O CN - - O - O N 0

1Q 10 M o% - N m CO - I- \0 CO 00 d- M M t- %o N- U

2 ý l

N O N K1 N I

'14 d- O a0 M ,

O m I

V) M 1 N

ý

-T Q N O O N - O O O N O - O O O

O z

Ix - V, % CO 0\ 04 lzT CO oo aZ cD do qt

I I d ao

1 N v

` %D N

I CO O

l IN OD f

M I

co J

r- rn O O 0 N - N IO O N O 0 O - N O ) 0Q cW

W 1'' O N co M ýt 0 N 0 00 N N co O C" N N N

-a W ý 1ull I

~ - (

r- ro

O O O O - O O O - O O N O . 0=

I-- ý- O- N V) CO t0 - d a' . 't a0 r- M N

:L 3-1 W 1 0 J

o l - - N - - I- 0 0 0 10 10 0 0 -0 - 00 N 00 of co tn - m CO 'CO oo 111) Ln N - r- 0

J- I l I I I

O O N I- M O O D - O O - O - D -

LCN CY, (N (D Lil 0 U

IN IK) I(N I 1 ,. T (CO I I I I

00 I

N I

0 M N O O O O O O - 0 - N -

CO M M M QO Oý d r C, CO N 00 N 0' n N N N ý I I I I

Q O O O O - O O O O O O - O O - O O O

ao N ý O i M r Ö "f) N i t l = ý

8 _ I

r %. o -

uý O I r 1 It) Ln

I ý

c) 0 0 O 0 O - O - - O O -

h- W oz --j Ü

O(1) W N ÜO aI

Z CO UN - kO a M n M N tN 0 QD ýD M - N - %0 N F- I- L3 M M CO N O\ 00 O CO 00 N O CO 'r -t 0'% - - CO CO O W O0 O - - O - O - N N O - N - O - - - N -

O C-) - IU-c. Cc - H a-

I s

I- W z -j OWW <

WNM 0

t5 iý O U) i

Wol z N - 0' u M rn V) ul N Lr\ ct '0 co N oD cr r` M N N F- F- L3 - a' O% O N lfl O

U- N M N N 'O '0 N - 0

- ui I- ÖZ0 Q - "Y ýt M d t1 ýt M M tit M d' M M

000 S

(L 3

iZ - %D - CO N 00 O O CO Jt - OD - CO v 8 N m ýcl to W¢ -- M - N1 If) Qý - d lrý - OD If) OD N O D Oý CO d' %D I- W

W ýO N co o' O - N M d' If) %o TN co a' O - N M d' n 1- Q - - - - N N N N N N N N N N M M M M M M -Z - - - - -- - - - - - -- - - - - - - -

141

0 M

z v

c9

dý O w

w

Z Z O_

H N

} F-

3 N

W a_

141 cn ri w J

z ZW

w 0 = S F- Qi

Of ý - I- t- < L rý - to O SQZU L

x O O

ZW0 Q' 0-

ts ", C' 04

lo -

d N .

0 UD pr) 0

- O

0 I

O ' 0

co T Nl N M

I M tý

d - O - O

0 CC) CD N O O O

H

- - - ý-; r (j) 0 0 w J

0 1-) 0 C) U) O p - - - 0 (. D

N a- Z N N - 0 0

O C9 N 0

( N

W 0 O O 0=

co oo ýzr u, \ 7 0- U- O lPf)

0 N ; N

Ö - O

-L w o fn to F- W - N " s. 0 N SO O O

p n N - M O N - M - O N N

co 1 U 1-

l - N N C

1 C)

O M O N

OD M - 'zt -

a) co 1 I

(N 0 0 0 0 - I Ic l° (ö lö lö

o o

aZ J <

ON W-O U) 1 T 50 0, r, co ") r- co

L - OU 0 N

N O M - 00

0 7- C) -0 F-

m cy- F- O_ 3

U-i az -J W CW)

Ü F- -O cn ý WO 1 z - - M Wa N U)

it C) C. ) C) in

000 2

F- CL

Z 1Sl N W¢ \0 - U) C' - Ul

w

w r- co M F- O Ma M

142

a critical cattell's model of personality - brunel university research

Documents