DOCUMENT RESUME

ED 053 809 PS 004 908

AUTHOR Woolman, Myron; Grotberg, Edith H., Ed.TITLE Designs and Proposal for Early Childhood Research: A

New Look: A Systems Approach to Pre-SchoolEducation. (One in a Series of Six Papers).

INSTITUTION Office of Economic Opportunity, Washington, D.C.Office of Planning, Research, and Evaluation.

PUB DATE 71NOTE 152p

EDRS PRICE EDRS Price MF-$O.65 HC-$6.58DESCRIPTORS *Early Childhood Education, Educational Research,

Family Influence, Instructional Systems, Learning,*Preschool Education, *Research Design, *SystemsAnalysis, *Systems Approach, Tables (Data)

ABSTRACTSuggested is a learning systems approach designed to

be consistent with the objectives and scope of developmentalpsychology and yet sophisticated and appropriate enough to fit theunique demands of early childhood education. Identified are eightcontingencies for learning that any educational system must possess.Added to these are features considered distinctive to learningenvironments for preschool children. Among these features are thefamily as a learning system, a systems perspective, a psycho-socialtheory of classroom learning, the philosophic position, criteria andspecifications, the operational stage, evaluating and measuringlearning systems, an educational research superstructure, and theeducational research testing system. Eight figures specify designsand relationships important in conceptualizing and operationalizingthis approach. (WY)

U.S. DEPARTMENT DF HEALTH.EDUCATION & WELFAREOFFICE OF EDUCATION

THIS DOCUMENT HAS BEEN REPRO.DUCED EXACTLY AS RECEIVED FROMTHE PERSON OR ORGANIZATION ORIG-INATING IT. POINTS OF VIEW OR OPINIONS STATED DO NOT NECESSARILYREPRESENT OFFICIAL OFFICE OF EDU-CATION POSITION OR POLICY

DESIGNS AND PROPOSAL, FOR EARLY CHILDHOOD RESEARCH:A NEW LOOK: A SYSTEMS APPROACH TO PRE-SCHOOL EDUCATION

Myron Woolman

Edith H. Grotberg, Ph.D., Editor

One in a Series of Six Papers

OLFICE OF ECONOMIC OPPORTUNITYOFFICE OF PLANNING,

RESEARCH AND EVALUATION, PRE/R

1971

1

-1-

A Systems Approach to Pre-School Education

Myron Woolman, Ph.D.Myron Woolman, Inc.

New York City, New York

INTRODUCTION

Education in crisis implies a society in peril. Thegestational role of education as the key institution for pro-viding manpower skills and social stability imposes a require-ment for substantive educational change as a high rankingnational priority.

It is imperative that we identify, select, and use thosemethods which predict and control the improvement of learningrate, retention, conceptual and performance skills as well associal and human values. This paper represents an effort topresent one approach to the research on learning systems forpre-school children.

How and why behavior changes is basic to any educationalapproach. The improvability of human learning for education inthe pre-school area is dominated by the Developmental-Matura-tional position. Considerable attention is now focused onbehavioral specification and management of contingencies; thisview is largely derived from B.F. Skinner's work and thinkingand, therefore, will be termed The Skinnerian-Behavioral Position.Few analysts who have examined learning problems have limitedthemselves, absolutely, to one or the other position. However,for definitional purposes, the alternative positions will bestated in absolute terms.

The Absolute Maturational & Development Position

The maturational position, in its least sophisticated form,suggests a parallel between physical and intellectual growth. Inboth domains, the infant, largely controlled by genetic factors,is viewed as moving through successive stages, growing and

2

-2-

becoming more proficient over time. Cultural and environmentalfactors nurture and enrich both behavioral and physical growth,but are not the true determinants of personality, attitudes orintellectual capability. The child can attain only that level whichwas initially (genetically) possible. He toves through the successivestages of behavioral (and intellectual) development at a rate largelydetermined by his genetic potential. There is some modifiabilitybut, except in extreme ranges of environmental pauperism and enrich-ment, environment is believed to have little substantive effect

_

on ultimate performance level.

The Maturationist View

Rousseau in Emile saw the child as a two-stage learner: firstimagery, then ideation. His views are cogently expressed in thefollowing passages:

...Before the age of reason, the child does not receiveideas, but images; and there is this difference betweenthe one and the other, that images are only absolutedepictions of tangible objects, and that ideas are notionsabout objects, determined by certain relationships. Animage can be alone in the mind which conceives it; butevery idea presupposes others. When one imagines, onedoes nothing but see; when one conceives, one compares...(in Levitas, 1963; p.4).

His second view of the stages appears to equate sensationwith imagery and suggests that, as ideation emerges, ancientimages must be reordered:

All their knowledge is in sensation; nothing haspassed into the understanding. Their memory itselfis hardly more perfect than the other faculties, sinceit is almost always necessary, when they are grown-up,that they learn again the things for which they learnedthe words in childhood... (Ibid.)

Piaget, using a more modern and formal idiom, (with suchterms as reflexes, schemata, perceptions, etc.) has articulated

-3-

a position similar to Rousseau's but one considerably moredetailed and elaborate. The following conveys the essence ofhis position:

...between the sensorimotor intelligence which precedesthe advent of speech and the later practical intelligencewhich subsists under verbal and conceptual realities,there is not only a linear continuity but also thereare temporal displacements in extension, so that inthe presence of every truly new problem the same primi-tive processes of adaptation reappear, although diminish-ing in importance with age (Piaget, 1954; pp. 357-58).

Note that Piaget appears to reassess Rousseau's fundamentalview that sensation--Piaget uses the term "sensorimotor intelli-gence"--precedes speech as well as the later "practical intelli-gence", that is, the development of skills in making adequatechoices in daily life activities. However, the three stages(sensorimotor intelligence, speech and practical intelligency)are constantly recapitulated in situations where choices are notapparent and decision making skills must be learned. Piagetexpresses this process of three stage ideational evolution inthe following way:

...Thus it may be seen that thought in its variousaspects reproduces on its own plane the process ofevolution we have observed in the case of sensorimotorintelligence and the structure of the initial practicaluniverse. The development of reason, outlined on thesensorimotor level, follows the same laws, once sociallife and reflective thought have been formed. Confrontedby the obstacles which the advent of those new realitiesraises, at the beginning of this second period of intel-lectual evolution assimilation and accommodation againfind themselves in a situation through which they hadalready passed on the lower plane. But in proceedingfrom the purely individual state characteristic of sensori-motor intelligence to the cooperation which defines theplane on which thought will move henceforth, the child,after having overcome his egocentrism and the otherobstacles which impede this cooperation, receives from

74- ,

it the instruments necessary to extend the rational con-struction prepared during the first two years of life andto expand it into a system of logical relationships andadequate representations. (Ibid; pp. 385-86).

It may be worth noting that Piaget infers that the processof moving to the third "rational" stage involves a neo-Freudianconquest of egocentric obstacles. Only by overcoming infantilism,egocentrism and needs does the human infant move to the plane ofrational thought. The initial stage of this victory occurs duringthe first two years of life. Piaget explains the process thisway:

Moreover, gradually as objects, causality, spaceand time are elaborated, a coherent universe followsthe chaos of the initial egocentric perceptions. Whenin the second year of life representation completesaction by means of the progressive internalization ofbehavior patterns, one might therefore expect that thetotality of sensorimotor operations would merely passfrom the plane of action to that of language and thatthe organization of schemata would thus be directlyextended in a system of rational concepts. (Ibid; p.357).

A more radical maturational view is expressed by ArnoldGesell. Gesell considers the sequence and ordering of behaviorsas evolving events in time, largely controlled by muscles,nervous system development, skeletal readiness, etc. The child'semerging behaviors are essentially manifestations of physicalreadiness to perform those tasks we observe. There is a naturaland inevitable order of maturational unfolding:

...Every child is unique; but every child is also amember of one human species. Obedient to these speciescharacteristics there are growth sequences which arerarely or never circumvented. The motor control ofthe eyes precedes that of the fingers; head balanceprecedes body balance; palmar prehension precedesdigital prehension; voluntary grasp precedes voluntary

5

-5-;

release. Banging comes before poking; vertical andhorizontal hand movements before circular and oblique;crawling before creeping; creeping before uprightwalking; gestures before words; jargon before speech;nouns before prepositions; solitary play before social;perceptions before abstractions; practical before con-ceptual judgments. These are but a few simple examplesof the sequential order inherent in the structuralizationof child behavior, from its lowest to its highest mani-festations. (Gesell, 1949, pp.21-22.)

Gesell sees behavior as having definite points of maturationwhich erupt, much like a baby tooth, when the pre-existing behav-ioral or perceptual building blocks have manifested themselvesinto the necessary substructure to support new behavioral mani-festation. The skills for building a tower of blocks is viewedthis way:

...The final pattern of tower building is a condenseculmination of all the growth that went before. Thegradient begins with the comparatively simple patternof looking. With increasing maturity one refinementfollows another in lawful sequence: 1. Ocular focus2. Arm approach 3. Manual grasp 4. Finger grasp5. Release 6. Tower. At 15 months the infantunreels this sequence in a flash, but this skillfulflash of behavior is the patterned end-product of awhole year of constructive growth! (Ibid.)

Reading is viewed in similar terms:

All school skills have a similar pre-history of growth.They are always subject to the principle of developmentalreadiness. They are never the sole product of trainingor drill. For example, consider another simple six stepgradient in the field of Reading Behavior. The 15-monthold child who has just attained the sensorimotor skillof building a tower is also at the lower threshold ofreading. He can already help to turn the pages of a

6

picture book. He can definitely identify the circularhole in a circle-triangle-square form board. Surely thisis the growth rudiment of the capacity to recognize theround letter 0, which is the beginning of all reading!Moreover he can read some of the pictures of a bookto this extent: He pats a picture which he recognizes.Our illustrative growth gradient begins with that patternof behavior, - an elementary perception of a picture onthe printed page.

Reading Behavior

1) 15 months Pats identified picture in book.2) 18 months Points to an identified picture in book.3) 2 years Names 3 pictures in book.4) 3 years Identifies 4 printed geometric forms.5) 4 years Recognizes salient capital letters.6) 5-6 years Recognizes salient printed words (Ibid).

However, beneath these apparent cumulative building pro-cesses which combine to provide maturational readiness, Gesellsees another deeper pattern -- a genetic determinism. He unambig-uously expresses his conviction that personality patterns andlearning abilities are likewise woven into the maturational weband woof of life. The limits are quite inflexible; in fact, theinfant is largely "determined" at birth:

...Long range studies made in our clinic have demonstratedthat such traits as social responsiveness, readiness ofsmiling, self-dependence and motor agility tend to manifestthemselves and to persist under varying environmental con-ditions. Every child is born with a 'natural' which colorsand structures his experiences. The infant, to be sure, hasgreat plasticity, great powers of learning; but there arelawful limits to his conditionability. He has constitutionaltraits and tendencies, largely inborn, which determine how,what, and to some extent even when he will leaxA. (Ibid;Section I, pp.40-41).

In Gesell's teleological argument, both behavioral traitsand anatomy are inevitably determined by the seed. In man, as in

7

-7-

trees, the first moment of conception predestines most of the rest:

...These traits are both racial and familial. The racialtraits are those which are common to the whole humanspecies. The familial traits are the distinctive endow-ment which he inherited from his parents and a long lineof grandparents. The child comes into this double inheri-tance through an innate process of growth which we callmaturation. He comes into the social 'heritage' of culture,through a process of acculturation. These two processesinteract and interfuse, but the process of maturation ismost fundamental, - so fundamental that acculturationcan never transcend maturation. (Ibid.)

This ultimate position -- that behavior is an inevitableunfolding of "racial and familial endowment" -- places Gesell ina theoretical cul-de-sac. His psychology ultimately is reducedto tape, film and written recordings of events long predestined.In his hands psychological intervention becomes an absurditybecause all behavior has been pre-printed in the chromosomal gel.Each individual in turn is merely a combined reflection of twopredestined chromosome pairs who had joined inevitably, as had allgenerations of chromosomes before them.

Despite Gesell's fundamentalist view on maturation, hiswork reveals many ingenious devices which could serve to improvethe rate of development of those children he studied. His labora-tory techniques provided enrichment through novel and artful toys,activities, social and verbal interaction, sympathetic interest,and a sophisticated and empathetic research staff. Thus theactuality of Gesell's work involved the testing of developmentalprocesses within an enriched environment. What would have occurredthrough maturation alone is left untested in a laboratory sense,although Gesell makes his own faith clear enough.

Thus, Gesell, perhaps inadvertently, approached the develop-mental position which differs from the maturational in its freedomfrom genetic determinism, its faith in the effects of enrichment,and its emphasis on research methods rather than observation and

8

recording. Modern developmental psychology, though eclecticand uncommitted to any rigorous theoretical camp, holds that muchcan be done to improve human learning, performance, social inter-action and the educational process.

The Developmental Position

The zoologist-geneticist, Julian Huxley, whom one mightexpect would hold a strong genetic determinist view, speaks vigor-ously for the unlimited growth potential of the individual andhumanity:

The critical point in the evolution of man- thechange of state when wholly new properties emergedin evolving life- was when he acquired the use ofverbal concepts and could organize his experiencein a common pool. It was this which made human lifedifferent from that of all other organisms; and wecan now begin to grasp the nature of profundity ofthe difference. The development of animals is alwaysclosed; their evolution is always sooner or laterrestricted. Man's individual development, on theother hand, is potentially open. It continues through-out his life, and it can take place in all sorts ofdirections; while in animals there is only one normalpattern to be realized. This same sort of thing holdsfor man as a type - his pooled experience can be indef-initely added to, and it can be organized in an indefinitenumber of different ways. Animal types have limitedpossibilities, and sooner or later exhaust them: manhas an unlimited field of possibilities, and he cannever realize all of them. He has developed a newmethod of evolution: the transmission of organizedexperience by way of tradition, which supplements andlargely overrides the automatic process of natural selec-tion as the agency of change in the human phase (Huxley,1960, pp. 28-29) .

Clearly, in Huxley's view, behavior is infinitely variableonce man shifts (learns) to compress meanings into terms whichpermit the pooling of experiences and concepts in immediate and

9

past time. This experiential pool, once crystallized intolanguage may be used to classify events and relationships intoplans to achiave future goals and if organized into a researchstructure, to predict new events in future time.

Edward Zigler views the developmental position as theprovince of psychology, one which is largely concerned with theeffects of complex processes over extended periods of time, ascontrasted with studies of behavior involving narrowly focusedrespoimes to very limited ranges of stimuli. Further, the develop-mentalist also focuses on locating focal points at which changeoccurs. Both positions are advanced in Zigler's following state-ment:

...The developmentalist is interested in change, notas a function of time, but rather as a group of organ-ismic processes which take place over time. The develop-mentally-oriented psychologist has always been struckby the phenomena of growth and change and the orderliness,sequentiality, and apparent lawfulness of the transitiontaking place from the birth or conception of the organismto the attainment of maturity. The developmentalist'stheoretical task has been one of constructing principle,or constructs making such change comprehensible. Suchprinciples clearly have little to do with time and muchto do with those processes, involving the person and hisenvironment, which give rise to changes in behavior.

Stated in this way, developmental psychology becomesan extremely arbitrary subdivision of psychology. For ifthe concept of process is divorced from any particulardevelopmental psychology with the psychology of learning,the orthodox learning theorist would handle observednhanges in the effectiveness of his variables at differentages by introducing into his equations or predictionsdifferent values for the parameters of interest. Thedevelopmentalist is interested in a super-ordinate explan-ation of all such substitutions of values. While thelearning theorist may change the values of his parametersfrom age to age, the developmentalist is concerned with

10

-10-

discovering the transition rules for these changes, makingthe change itself, rather than the content of change, thecentral issue. (Zigler, 1963; pp. 344-45).

Zigler asserts that there is an important place for experi-mental research in developmental psychology. Further, the develop-mental research situation should provide a test condition whichis generalizable to a much broader theater of action than theresearch setting. The developmentalist must be ready to accepterror in his prediction due to the limited theoretical base; how-ever, the price is not too high because, without research, therecan be no significant movement forward:

...Only to the extent that we introduce some externalcriterion of validity can we choose or synthesize whatis useful in those conceptual principles or constructsinduced from naturalistic observation. To the authorsuch a criterion is provided by the experimental method.As indicated earlier, many developmentalists have notfully comprehended the role of experiment in the valida-tion process. In order to illuminate this relationship,a clear distinction must be drawn between two types ofexperiments. One type of experiment unquestionablyrepresents an effort to mimic nature. That is, natureis brought into the laboratory, and a number of controlsis exerted to discover the relationship between particularvariables. This type of experiment falls on the samecontinuum with natural observation, its superioritylyi:wg in its efficiency. Thus, armed with a thermometerand the knowledge of how far above sea level variouslocations were, we could through naturalistic observationsdiscover the temperature at which water boils and how airpressure affects this boiling point. But how much easiercan we discover these relationships in the controlledsituations provided by the experiment. However, manydevelopmentalists find this type of experiment inapplicableto developmental psychology, and their view is probablypartially correct. Much of the nature in which we areinterested either cannot or should not be mimicked in thelaboratory

A second type of experiment is more related to theproblem of validating theoretical systems. This type of

11

experiment does not mimic, but rather creates, nature.Here the phenomena of nature are not brought into thelaboratory, but instead phenomena are created which havelittle or no chance of being found outside of the labora-tory. At this point the experimenter can be conceptualizedas dealing with two distinct worlds, the world at largeand the 'world' which he has created. Explanatory systemswhich are advanced, be they in physics or developmentalpsychology, are established to explain the real world...

Experiments do not test theories in their entirety,because of that portion of theories which is given,assumed, and untestable. However, the elemental principlesor processes are tested, and the outcome of such testsdetermines the validity of the system. Thus, experimentsof the second type are invariably of the theory-testingvariety and involve the 'if such -- then such' paradigm.What the experimenter is saying is that if such and suchholds in the real world because of the principles expoundedin the particular theory under investigation, then suchand such should hold in the world which the experimenterhas created.

Developmental psychology does appear to contain uniqueinterests and concerns which make it a delimited domaindemanding approaches, methodologies, and theories havingparticular characteristics. What is required for suchdelimitation is not the definition that developmentalpsychology is concerned with change as a function of pro-cess, but rather that it is concerned with the changes inthe form or organization of responses over time as contrastedwith the change in the strength or accuracy of the response.What happens to a response after a prescribed number oftrials is 'learning.' What happens over five years is'development'. Thus, the developmentalist focuses onstructural changes in a response, changes which cannotbe defined simply in terms of changes which occur withsingle trials or stimulus presentations. (Ibid, p. 351-53).

!12-

Zigler indicates that there are substantive differencesbetween the developmental position and that of Gesell, Piagetand Freud:

...Thus, the actual difference between Gesell and Piagetis not as great as one might initially imagine. The pre-dictions derived from either system are limited almostentirely to sequential behaviors. Neither systemhas explicit concepts relating particular environ-mental factors to the nature of the sequence or tothe content observed. This is true for Gesellbecause he is a maturationist, and, his notion ofreciprocal interweaving to the contrary notwith-standing, such issues are viewed by him as pseudo-issues. It is true for Piaget because he is moreinterested in the general epistemological implicationof the sequences than he is in the psychology of humanbehavior. However, we do not mean to imply thatPiaget has not given us an extremely provocativesystem. Its prime value appears to rest not on itsmerit as a finished theory but rather on its suscep-tibility to translation, either in whole or in part,into more formally adequate theories or propositions...

...The tension-reduction, hedonistic approach to man,whether it stems from the Freudian or classic behav-ioral systems, has always been unacceptable to thedevelopmentalist, for such a view simply does notreflect what developmentalists have grossly observedin the child's development. (Ibid, pp. 362-63)

Learning theorists are viewed by Zigler as too limitedin their view of the human condition, and too overly concernedwith upholding their own school of thought among the manymicro-psychologies falling under the learning theory umbrella.However, Zigler is sympathetic towards efforts to test andexplore such theories, but cool towards efforts to use develop-mental psychology as a means of dramatizing an entrenched theo-retical position:

...Another factor that has made developmentalistswary of investigators with a learning theory approach

13

-13-

has been the reluctance of many of these investiga-tors to become truly involved with and receptive tothe total behavioral picture presented by the develop-ing child. Such investigators appear much more com-mitted to a particular learning theory than they areto the content area of developmental psychology.Their efforts have appeared to be directed more towardsthe demonstration of the applicability of a learningtheory principle than to the unraveling of the centralproblems of growth and development. Such efforts arebest exemplified in the numerous demonstrations thatthe Law of Effect does indeed hold for children andin that plethora of studies that attempt nothing morethan the mechanical application to children of researchdesigns originally employed with animal populations.Such investigations do little to refine the learningtheories from which they were derived, nor are theyof much assistance to the developmentalist attemptingto illumine the mysteries of human development.

This criticism, though a justifiable one, appearsto have been overextended by many developmentalists.For many years a group of developmental psychologists,though beginning with classical learning theory, havebeen genuinely committed to the investigation ofchildren's behavior and have been primarily concernedwith the development of a learning theory capable ofencompassing this behavior, rather than with the demon-stration that the original theoretical efforts werevalid. In almost every instance, the efforts of suchinvestigators have resulted in the enrichment of theoriginal theory employed, as well as adding to ourunderstanding of children's behavior. Though thefollowing listing is far from inclusive, such effortcan be seen in the earlier work of Child, Whiting, theNowlises, and the Searses, and in the more recent worksof Berlyne, Gewirtz, Stevenson, Kessen, and the Kend-lers. (Ibid; pp. 365-66).

The future of developmental psychology looks hopeful toZigler. He sees the field as needing a more fully articulated

14

-14-

theory; thus, he takes a positive view of the influx of learn-ing-theory based psychologists since they might possibly pro-vide some impetus toward theoretical development throughcrops-fertilization:

...Developmental psychology is not only ready butbadly in need of theories of the former sort, iffor no other reason than to move us beyond ourpresent state of knowledge. The construction ofsuch theories is insured to the extent that we haveinvestigators within developmental psychology whoare imbued with a heritage which emphasizes theimportance and value of theory construction and whoare sophisticated as to the rigorous requirementsof such an undertaking. The developmental psycholo-gists with a learning-theory orientation appear torepresent just such a heritage...(Ibid, p. 366.)

Baldwin (1967) makes the additional point that a psycho-logical theory must be testable in the sense that it must re-sult in predictions which can be established to be true orfalse in a way which communicates to others:

The properties of a good language and a good theoryare quite different. A theory should be testable,that is, falsifiable. A theory explains some behavior,but it declares that other behavior will not occur.A theory that can account for any conceivable behav-ior is untestable. A good language, on the otherhand, should be able to express any content, whetherit is true or not. It should be able to describeboth possible and impossible situations. For example,nobody necessarily believes ghosts exist because heuses the word 'ghost.' The language needs, however,to contain the word as long as people have commentsto make about ghosts, even if the comments only ex-press skepticism about their existence. (p. 587).

Recent investigations in the developmental area involvethe testing of key learning theory concepts related to theeffects of contingency on behavior. Work by Hava and Jacob

Gewirtz illustrates this type of microscopic behavioral ana-lysis. The central theme of the method is to attempt totrace and detail successive behavioral elements which occurbetween infant and parent. For example, if the infant initiatesa given class of response, what parental consequence occursand vice versa. This technique makes it possible to predicta given element of parental behavior, knowing a bit of infantbehavior. The following passage illustrates the Gerwitzframe of reference:

...A number of reasonable contingency patterns emergefrom an examination of behavior categories havingsubstantial incidences of occurrence. These arecharacterized by EE values like those which, in avariety of analyses, could serve as individualdifference measures (i.e., dependent variables).Some examples of patterns of adult responses tochild behavior follow: the cps that a child'svocalization will be followed by adult talkingin four Ss (one from each of our four environ-ments) range from .52 to .81; the cis that achild's vocalization will be followed by an adultsmile range from .21 to .42; the cis that a child'ssmile will be followed by an adult smile rangefrom .46 to .88; and the cps that his smile willbe followed by adult talking range from .25 to .51.Some examples of patterns of infant response toadult behavior for the same four Ss are the follow-ing: the ERE that an infant's smile will followan adult's approach range from .24 to .44; thecps that an infant's smile will follow an adultsmile range from .50 to .79; and that an infant'ssmile will follow an adult's talking range from .30to .57; and, lastly, the EEE that an infant's vowelvocalization will follow an adult's talking rangefrom .09 to .15.

In connection with adult-child contingencyclusters, our preliminary impressions is, first,that aside from visual orientation responses, thechild's predominant Lasponse to various adult

16

-16-

initiations (e.g., approaches, smiles, talks, hugs)is smiling, regardless of the 'modality' of theadult's evoking initiation. In contrast, theadult's predominant responses to the child's initia-tions appear more to be of the same 'modality' asthe initiator's, at least with respect to the child'ssmiles and vocalizations. Further analyses willdetermine whether such contingency clusters are afunction of the stimulation in the environment, ofdevelopmental level, or perhaps whether they reflectonly residual variation that is not to be explainedin terms of the independent variables of this study.(Gewirtz and Gewirtz, 1969; pp. 246-47; italics inthe original).

The Skinnerian behavioral position makes no assumptionsas to genetic limitations. Learning is viewed as a process ofimmense intricacy and complexity which takes place dynamicallyunder conditions of flux involving environmental stimulus,time, space and contingencies related to organismic drivesand goals. Learning is viewed as stemming from the conse-quences of response. The position is an optimistic one, sinceit holds that once the process is more fully understood,learning efficiency will increase. Further, just as buildingscannot be properly assembled by tossing rocks, concrete andglass in a random fashion into a pile, so humans cannot learnproperly unless there is a systematic and detailed engineeringof all aspects of a learning problem.

Sidney Bijou has recently argued that behavioral prin-ciples, based on Skinner's and Keller's positions, woulddramatically improve the rate at which children learn if theywere applied within the school setting. The fact that thisposition is held by a small minority of psychologists, Bijoucontends, is an obstacle to the rapid improvement of our edu-cational system. He presents these concepts quite optimisti-cally; behavioral analyses and theory are viewed as being readyto make major contributions now:

17

-17-

...Still another group of psychologists, which atpresent is only a small minority, responding tothe question of what psychology can contributeto education now would say: 'We can offer aset of concepts and principles derived exclusivelyfrom experimental research; we can offer a method-ology for applying these concepts and principlesdirectly to teaching practices; we can offer aresearch design which deals with changes in theindividual child (rather than inferring them fromgroup averages); and we can offer a philosophy ofscience which insists on observable accounts of therelationships between individual behavior and itsdetermining conditions.' (Bijou, 1970; p. 66).

Bijou lists five basic assumptions underlying the useof behavioral analysis to indicate the theoretical frame ofreference for this minority. These assumptions appear com-pletely general across psychology, except for Assumption#5, which seems to set up sub-assumptions and restrictionswhich might or might not have virtue in the: classroom context.Bijou's assumptions are:

The subject matter of psychology is theinteraction between the behavior of an integralorganism and environmental events. These inter-actions are analyzed in observable, measurable,and reproducible terms and therefore are amenableto scientific investigation.

2. The interactions between the behavior ofan individual and environmental events are lawful.Given an individual with his unique biologicalendowment, changes in his psychological behaviorare a function of his interactional history andthe current situation in which he is behaving.

3. As in all of the sciences, the subjectmatter of psychology exists in continuities.Continuities are assumed to exist in the stages

18

-18-

of development, in the rates of development(normal, retarded, and accelerated), in therelationships between normal and pathologicaldevelopment, in the problems of procedures ofbasic and applied research, and in the analysisof psychological phenomena from raw data to theo-retical formulation.

4. Complex interactions evolve from simple inter-actions and begin with the infant's initial relation-ships with people and objects. This does not meanthat complex behaviors are assumed to be sums ofsimple behaviors. How a specific form of complexbehavior, surh as mathematical problem solving, isestablished is a problem for experimental study.The final analysis of any class of complex behaviorwould probably involve many concepts and principlessuch as minute stimulus control, subtle variationsin setting conditions, and intricate schedules ofreinforcement.

5. A psychological theory and its technology areopen and flexible systems. That is, a new concept, anew principle, or a new technique may at any time beadded to the existing list, provided that it can dis-play the proper credentials: it must be tied unequi-vocally to observable events; it must be functional;and it must not overlap with the concepts, principles,or techniques already catalogued. (Ibid; pp. 66-67,italics in the original).

Assumption #5 appears to assert the sub-assumption thatwhat is learned should be atomistic and mutually exclusive.

Bijou also makes a statement on teaching-oriented appliedresearch which appears to be quite compatible with Zigler'sview that competing learning environments should be comparedwith each other experimentally:

...The strategy of teaching-oriented applied researchdoes not consist of designing a study to determine

-19-

whether Method A is better than Method B for theteaching of subject-matter X. It is, instead, a searchfor ways to engineer an educational environment so thateach child can learn specified tasks, and then, afterthat goal is attained, to compare achievement in thatengineered situation with achievement in some otherschool situation. (Ibid, p. 67).

Bijou's view of the behavioral analysis schema is neatlysummarized:

...Let us turn to the scheduling of stimulus materials.The fact that a school task can be learned with a min-imum of frustration and on the basis of positive rein-forcement via a program of differential reinforcementof successive approximations to the ultimate form ofa response (skill), or the desired response in theproper situation (knowledge), has led to an over-emphasison the role of teaching machines, and to a misconceptionabout the school subjects that can be properly programmed.Teaching machines, from the most primitive to the mostelaborate, are of value in teaching only insofar asthey assist the teacher in arranging the contingenciesthat expedite learning, i.e., aid the teacher in pre-senting the material properly, in providing for expli-cit responses, and in arranging for optimum timing ofeffective contingencies of reinforcement. The program-ming of any academic subject for a child is straight-forward: (1) state in objective terms the desired ter-minal or goal behavior, (2) assess the child's behavioralrepertory relevant to the task, (3) arrange in sequencestimulus material or behavioral criteria for reinforce-ment, (4) start the child on that unit in the sequenceto which he can respond correctly about 90% of the time,(5) manage the contingencies of reinforcement with theaid of teaching machines and other devices to strengthensuccessive approximatiuns to the terminal behavior andto build conditioned reinforcers that are intrinsic tothe task, and (6) keep records of the child's responsesas a basis for modifying the materials and teachingprocedures. (Ibid; p. 68).

20

-20-

To summarize Bijou's view of the idealized learningsituation: presentation of content should be organized intodiscrete and digestible units, combined with intensive sur-veillance and precise responses by the teacher to insure properreinforcement when, where, and under the conditions required.

The most powerful spokesman for the behavioral analysisposition is, of course, B.F. Skinner whose work on operantconditioning has profoundly affected mid-century psychology.His later interest in improving the classroom as a learningenvironment, heralded by his teaching machines and programs,can be expected to generate major repercussions in psycho-education in the years ahead. Ski-ner, in his meticulous way,describes contingencies of reinforcement in the following terms:

...The so-called 'contingencies of reinforcement' arean important feature of the independent variablesstudied in an experimental analysis. A few contin-gencies, such as conditioning, extinction, and delayof reinforcement are familiar.

...But many psycho.ogists are unaware of the complexityof the contingencies now commonly studied. In additionto many standard schedules of reinforcement, reinforce-ment may be contingent on rate of responding, rate ofchange in rate, or specific patterns of rate changesdetected by on-line computer analyses. Contingenciesmay involve several stimuli and responses interrelatedin various ways. Considerable skill may be needed todesign programs of instructinlal contingencies whichwill bring behavior under the control of complex term-inal contingencies of this sort. The importance of pro-gramming is, indeed, often completely overlooked. Forexample, the statement that a given type of organismor an organism of a given age 'cannot solve a given kindof problem' is meaningless until the speaker has speci-fied the programs which have been tried and consideredthe possibility that better ones may be designed.

Describing a set of contingencies in instructionsto the subject is no substitute for exposing the sub-ject to the contingencies, particularly when they need

21

-21-

to be programmed. Instructions have effects, of course,depending in part on the verbal history of the subject,but the behavior of a subject to whom an experimenterhas explained how a piece of apparatus work; will notnecessarily resemble one wno has come under the controlof the terminal contingencies established by that appar-atus.

Contingencies in reinforcement have been analyzedformally in theories of probability, decision-making,and games, but the theorist often has no way of knowing,aside from observation of his own behavior, what effectsa given set of contingencies will have or what kind ofprogram may be needed to make it effective. Certainassumptions -- for example, that an organism will behaverationally -- are sometimes used in lieu of observationsto complete a statement of contingencies. Formal state-ments of contingencies, like instructions, have theireffects and if detailed enough may supply rules whichfunction as prior stimuli to control behavior resemblingthat which would be generated by prolonged exposure tothe contingencies themselves. The two cases must, how-ever, be clearly distinguished...

The increasing power of an experimental analysishas made it possible to examine the effects of complexcontingencies to which an organism has traditionallybeen assumed to adjust only by exercising certain cog-nitive processes. It is sometimes obvious that suchprocesses have been invented simply to Account for thebehavior in the absence of any better information asto how the contingencies could generate it. The experi-menter has not been able to relate the behavior to thecontingencies, and he is forced to conclude that theorganism has somehow done so mentally. Supposed cog-nitive processes of this sort may be disregarded.Others, however, may be a sort of internalized versionof precurrent behavior -- behavior maintained by itseffects in maximizing the reinforcement of subsequentresponses. Precurrent behavior is part of the subjectmatter of an experimental analysis. It is usually

22

722-

studied in overt form though it may eventually dropto the covert level. In either case it is defined asbehavior which affects behavior rather than as mentalactivity. (Skinner, 1966; pp. 215-16).

Skinner then points out that his early animal work hadbeen criticized as too narrow, but that time is on the sideof those who hold fast to experimental analysis and carefullydeveloped data:

...Unlike hypotheses, theories, and models, togetherwith the statistical manipulations of data whichsupport them, a smooth curve showing a change inprobability of response as a function of a controlledvariable is a fact in the bag, and there is no needto worry about it as one goes in search of others.The shortcomings and exceptions will be accounted forin time. The strategy is supported by the history ofearly criticisms of the Behavior of Organisms. It wassaid that the book was not about organisms but aboutthe rat, and very small groups of rats at that. Howcould one be sure that other rats, let alone animalsof other species, would behave in the same way? Onlyfood and water were used as reinforcers, social reinfor-cers being conspicuously lacking. The stimuli--lightsand buzzers--were crude and poorly controlled. Twolevers should have been used so that the data wouldthrow light on behavior at a choice point. And,after all, could we be sure that the rat was not press-ing the lever simply because it had nothing else to do?These criticisms have all been answered without effortin the course of time simply as part of the normaldevelopment of the analysis.

Patience with respect to unexplored parts of a fieldis particularly important in a science of behavior because,as part of our own subject matter, we may be overwhelmedby the facts which remain to be explained. Subtle illu-sions, tricks of memory, the flashes which solve prob-lems--these are fascinating phenomena, but it may be

-23-

that genuine explanations within the framework of ascience of behavior, as distinguished from verbalprinciples or 'laws' or neurological hypotheses, areout of reach at the present time. To insist that ascience of behavior give a rigorous account of suchphenomena in its present state of knowledge is likeasking the Gilbert of 1600 to explain a magnetic am-plifier or the Faraday of 1840 to explain superconduc-tivity. Early physcial scientists enjoyed a naturalsimplification of their subject matters. Many of themost subtle phenomena were to come into existance onlythrough technical advances in the sciences themselves.Others, though occurring in nature, were not recognizedas parts of their fields. The behavioral scientistenjoys no such natural protection. He is faced withthe full range of the phenomena he studies. He musttherefore more explicitly resolve to put first thingsfirst, moving on to more difficult things only when thepower of his analysis permits.

A final distinction. Those who engage in the exper-imental analysis of behavior are usually conspicuousfor their enthusiasm. In a recent article Bixenstine(1964) attributes an unwarranted optimism in all behav-ioral science to the methodological position taken byexperimental analysts. This is perhaps to overestimatetheir influence. In any case, he points to the wrongcause. He suggests that the optimism springs fromrelease from the anxiety of theory construction. Thereis a more obvious explanation: the analysis works.(Ibid; pp. 217-18).

Skinner sees words as tools for imposing order and achiev-ing goals. The word functions to trigger a reward-related response,much like an animal presses a bar to receive a pellet. It is asignal to produce something of value to the utterer and justifiesthe expenditure of energy by simplifying the attaining of agoal object:

...The verbal response 'Tea, please' may have two con-sequences: it may produce a cup of tea for the speaker

-24-

and may injure a friend by depriving him of the chanceto ask for tea. These are Separate consequences,having different effects on the probability that theresponse 'Tea, please' will be emitted. Under betterconditions we could demonstrate this in a simple way.Insofar as the response is strong because it producesa cup of tea, we can alter its probability by makingtea more or less reinforcing. Thus, by depriving thespeaker of tea we can increase the probability thatthe response will occur, or by giving him a largequantity of tea before he speaks we can reduce the prob-ability of the response. Insofar as the response hasthe effect of injuring a friend, we can alter itsstrength by altering the speaker's tendency to workinjury. If we can persuade the friend to insult thespeaker, for example, or in some other way increasethe latter's tendency to take revenge, the probabilityof saying 'Tea, please' will rise. (In Hook, 1960;pp. 228-29) .

Skinner presents a framework or set of working ruleswhich suggest that a classroom can be organized as an elaborateprogrammed text or as a massive teaching machine. The teacherin this classroom operates as a high-speed program synthesizer,setting up contingencies, organizing sequences and reinforcingresponses with precision, virtuosity, and minimal latency.One problem in this approach appears to be the location andtraining of larger number of such teachers. In addition,the educational perspective appears, at this point, to belimited to verbal and psychomotor skills, and does not extendto the child as a total socialized human being.

Theory must be considered a flexible scientific toolwhich provides a resource for the researcher, so that he maybetter achieve a research objective. The wave and corpusculartheories of light serve complementary functions for solvinganalogous but different problems. Theory, in any field, psy-chology included, need not become unitary; rather, it canserve as a tool for helping man to serve man. Frank, in a

beautifully lucid passage, presents a functional view ofscientific theory which merits mention.

...In the same way that we enjoy the beauty and ele-gance of an airplane, we also enjoy the 'elegance' ofthe theory that makes the construction of the planepossible. In speaking about any actual machine, it ismeaningless to ask whether the machine is 'true' in thesense of its being 'perfect.' We can ask only whetherit is 'good' or sufficiently 'perfect' for a certainpurpose. If we require speed as our purpose, the'perfect' airplane will differ from one that is perfectfor the purpose of endurance. The result will bedifferent again if we choose safety, or fun, or con-venience for reading and sleeping as our purpose. Itis impossible to design an airplane that fulfills allthese purposes in a maximal way. We have to make somecompromises. But then, there is the question: Whichis more important, speed or safety, or fun or endurance?These questions cannot be answered by any agreementtaken from physical science. From the view point of'science proper' the purpose is arbitrary, and sciencecan teach us only how to construct a plane that achievesa specified speed with a specified degree of safety.There will be a debate, according to moral, political,and even religious lines, by which it will be determinedhow to produce the compromise. The policy-making author-ities, from the logical viewpoint, 'free' to make theirchoice of which type of plane should be put into produc-tion. However, if we look at the situation from theviewpoint of a unified science that includes both physi-cal and social science, we shall understand how thecompromise between speed and safety, between fun andendurance is determined by the social conditions thatproduce the conditioned reflexes of the volicy-makers.The conditioning may be achieved, for example, by letterswritten to congressmen. As a matter of fact, the build-ing of a scientific theory is not essentially differentfrom the building of an airplane. (Frank, 1956; pp. 13-14).

26

26-

Frank's view of theory holds that a theory should:

1. Provide assistance in planning how to attainend point objectives.

2. Provide a tool to obtain different objectivesdepending upon context and requirements.

3. Be capable of improvement based on evidence ofits own imperfections.

If we now consider the Maturational, Developmental andSkinnerian Behavioral positions in the light of the Frank posi-tion, we find:

(1) The Maturational position cannot assist inthe formulation of a final product as theproduct preexists in the chromosomes. Theoryis irrelevant as the ultimate behaviors inexor-ably mature.

(2) The Developmental Position has postulated afinal product and specified many broad char-actersitics involved in the development of awell-rounded personality with sufficient skilllevels to satisfy life requirements. Thereseems to be agreement among the developmentaliststhat, to attain these outcomes, there should beenriched and stimulating environment, empathicadults, opportunity for self-expression, varia-tion in experience, understanding and support,etc. This fairly cohesive viewpoint has notbeen articulated into a rigorous theory but maybe said to represent a "Developmental Approach".With reference to Frank's second point, therecan be no shift in theory to meet varying ob-jectives since the Developmental Approach hasnot been structured sufficiently to permitvariation in approach to meet differing learnerobjectives.

27

(3) The Skinnerian Behavioral position does articulatea framework or set of rules which is substantiallymore rigorous than the Developmental Approach.However, it fails to indicate the total productother than to suggest that learning qua learningwill take place more efficiently given applicationof the rules. With reference to the second re-quirement, adaptability, there is implicit inSkinner's work a periodic scheduling and somepotential in terms of modification with a suf-ficiently adroit and knowledgeable teacher.Finally, Skinner has always modified his positionsbased on data and thus should increasingly improvethe power of the behavioral paradigm in a class-room learning context.

The remainder of this paper is largely an exposition ofthe writer's personal positions vis-a-vis theory, systems build-ing, and evaluation in education. Its objectives and scope are,by and large, consistent with those of developmental psychology.Though generally inconsistent with the maturational positionsof Gesell, Piaget's work is viewed as a botanization of languagedevelopment which offers little of value to the educator, despitethe evident care of the execution. It will be clear to manyreaders that much is owed to Thorndike, Hull, Skinner, Keller,Symonds and Lange. There are clear and obvious differences asto some aspects of the human learning process. The writer'sposition, in essence, argues that learning is fundamentally aconsequence of shifts in affect. Operations such as deprivationand shock are viewed as producing negative affect shifts, as isanxiety, apprehension, uncertainty, etc. The change in affectivestate which occurs following a response is the crucial conditionfor learning. If the shift is positive, learning takes place;the response tends to recur with increased frequency in thepresence of the stimulus. In addition, the writer holds thataffective relationships in humans fundamentally involve relation-ships with others. Man's psychological space is, for the most part,man. Thus, the development of affect shift contingencies,necessary for learning, can be obtained most efficiently ina social setting designed to provide high rates of response

28

and unlimited social rewards to the learners. In short, myposition holds that schools, when viewed as learning societiesin which all members are upwardly mobile, can be more efficientas learning systems, more satisfying to learners and staff,and provide greater variety, opportunity and freedom than isnow the case.

29

STRUCTURING THE CONTINGENCIES FOR LEARNING1

Learning is fundamentally a matter of gratification(satisfaction, drive reduction, etc.) which occurs as a con-sequence of a response (positive affect shift). A vital functionof any learning device or classroom method is to insure a highrate of positive affect shifts at appropriate times to increaselearning efficiency. A corollary to this point is that thegreater the affective investment by the learner in producinghis response (everything else being equal), the greater hisgratification when the response succeeds. A piece of educationalequipment or a procedure designed so that the learner is emotion-ally involved in completing something meaningful--an equation,a map, or anatomical drawing, etc.--should be more efficientthan one which merely requires him to press a button to indicatethe most acceptable of several pre-selected options. The veryenergy invested for a meaningful response completion is animportant consideration in the perceived gratification of successor the felt frustration of failure. Beyond this, the feedbackfollowing the learner response should be meaningful and function-ally related to the problem or test item. The use of buzzers,lights, and other signals as feedback, though useful, fails tosupply the learner with either the correct response or methodsof overcoming error when his response is inadequate. Similarly,a test which provides the learner no information except scoresmarked wrong or right, fails to define either the proper responseor the method of overcoming whatever errors exist.

Learning Contingencies:

Maslow has recently indicated that shifts in responses

1. This summary overview of learning contingencies emphasizesaffective consequences of response. The differences fromthe more generally accepted drive reduction and reinforcementpositions reflect the viewpoint of this writer.

30

in the course of growth and self-realization are largely amatter of the affective consequences of response. He presentshis views of the process in the following twelve points:

(1) The healthily spontaneous child, in his spon-taneity, from within out, in response to hisown inner Being, reaches out to the environmentin wonder and interest, and expresses whatevershills he has,

(2) To the extent that he is not crippled by fear,to the extent that he feels safe enough to dare.

(3) In this process, that which gives him the delight-experience is fortuitously encountered, or isoffered to him by helpers.

(4) He must be safe and self-accepting enough to beabout to choose and prefer these delights, insteadof being frightened by them.

(5) If he can choose these experiences which arevalidated by the experience of delight, thenhe can return to the experience, repeat it, savorit to the point of repletion, satiation or boredom.

(6) At this point, he shows the tendency to go on tomore complex, richer experiences and accomplish-ments in the same sector (again, if he feels safeenough to dare.)

(7) Such experiences not only mean moving on, but havea feedback effect on the Self, in the feeling ofcertainty ('This I like; that I don't for sure');of capability, mastery, self-trust, self-esteem.

(8) In this never ending series of choices of whichlife consists, the choice may generally be schema-tized as between safety (or, more broadly, defensive-ness) and growth, and since only that child doesn't

31

need safety who already has it, we may expect thegrowth choice to be made by the safety-need grati-fied child. Only he can afford to be bold.

(9) In order to be able to choose in accord with hisown nature and to develop it, the child must bepermitted to retain the subjective experiences ofdelight and boredom, as the criteria of the correctchoice for him. The alternative criterion is makingthe choice in terms of the wish of another person.The Self is lost when this happens. Also thisconstitutes restricting the choice to safety alone,since the child will give up trust in his owndelight-criterion out of fear (of losing protection,love, etc.).

(10) If the choice is really a free one, and if the childis not crippled, then we may expect him ordinarilyto choose progression forward.

(11) The evidence indicates thatchild, what tastes good forfrequently than not, 'best'far goals as perceivable by

what delights the healthyhim, is also, morefor him in terms ofthe spectator.

(12) In this process the environment (parents, therapists,teachers) is important in various ways, even thoughthe ultimate choice must be made by the child:

a. it can gratify his basic needs fr,r safety,belongingness, love and respect, so thathe can feel unthreatened, autonomous,interested and spontaneous and thus dareto choose the unknown;

b. it can help by making the growth choice pos-itively attractive and less dangerous, andby making the regressive choice less attract-ive and more costly. (Maslow, 1968; pp. 57-59;

italics in the original.)

Some of the key elements needed for an effective learningsystem are schematically presented in Figure 1, The LearningContingency Lattice. Cell A-2 asserts that whatever thelearning system objectives (Cell A-1), the learner's previousexperience must be used as the base. Learning is organismicand interdependent, and any new required learnthg product mustbase itself on what the learner perceives as personally importantas learning goals (Cell A-3). The learner's involvement orreadiness (Cell B-3) in the institutional objectives (A-1) isa primary consideration in designing a 7.earning system. Thesystem must accept his existing learning goals (A-3) and moveto the learning system objectives (Cell A-1) and the stageachieved by the learner in his learning development (Cells A-2and B-3). A sequential organization of learning materials(C1.11 A-3) is necessary to simplify the acquisition of skillsand knowledge. Cell A-4 involves the development of an informa-tion network, either a lattice or a matrix so arranged that thecomplexity and difficulty of the items presented in the programand the success probability of the responses to them may beplanned as part of the learning system specifications. Theseinitial planning methods function as the backbone of tha learn-ing system and control (as does an architectural blueprint) theultimate form of the system. The lattice permits the simultaneousexposure of all aspects of the system, permits control overthe development of the materials to be presented, provides thebasis for planning th learning sequence and response difficultyand is the basic tool for later evaluation of the system. Aboveall, it provides a method for controlling the rate of learnererror (Cell C-4) which is the vital component in determiningthe learner's affective reaction to the system.

When errors occur with great frequency, the learner'sresulting frustration may shift him ote:side the immediate learningcontext and produce various impulsive avoidance responses, suchas noise-making, clowning, throwing and a host of anti-socialbehaviors. Conversely, when the learning system uses presenta-tions constructed to minimize the possibility of error, theprogram is often perceived as monotonous or fatiguing; thus,affective investment of the learner is minimal, and the affectivereaction (gratification or frustration) is sharply reduced (F-7).

33

F

csti

EB

LEA

RN

ING

SY

ST

EM

MO

DIF

ICA

TIO

N R

EQ

UIR

EM

EN

TS

ASY

STE

MS A

PPRO

AC

H T

OPR

E- SC

HO

OL

ED

UC

AT

ION

TH

E LE

AR

NIN

G C

ON

TIN

GE

NC

Y LA

TT

ICE

LearningA

System

Objectives

LearningH

istory2

FIGU

RE

I

LearningR

eadiness

LearningG

oals3

Probability

of Response

Adequacy

Response

Investment

LearningV

alue ofR

esponse

40.Level ofG

ratification

Frustration

Intens ty

a4

Order of

Presentation

of Stim

uli

4

Meaning-

fulness ofR

equiredR

esponse

5

LifeR

elevanceR

equiredR

esponse

6

I+

Feedback Loop Indica ing

Response A

dequacy

7

Respor,sivity

Adequate

Response

Frequency

8

Evaluation

of LearningS

ystemO

bjectivesa

Evaluation

Methods

9

Fig. 1

This lattice indicates that an educational sv.tem

must carefully control the probability of resF

onse error and zit the same tim

e must insure m

eaningfulness ofresponse in

order to obtain responses which have re:atively Ligh learning value for the learner.

-33-

Response requirements should be meaningful (A-5) and per-ceived by the learner as relevant to his own goals (A-3 and A-6).These aspects (A-3, A-5 and A-6) of the system requirementsdetermine the degree to which the learner invests himself inthe consequences of a response (Cell D-6). Thus, Cell D-6 isthe fulcrum on which learning hinges; it is the index of affec-tive involvement. It determines the magnitude of the learner'spositive or negative affect shift following a response perceivedas successful or unsuccessful. Response investment may be termed"motivation to learn." Learning systems are most effectivewhen they produce "motivation" through controlling the conditionsunderlying the affective investment of the learners. Obviously,a high level of response investment produces greater frustration(Cell E-7) than a lower emotional commitment whenever an endeavorresults in error. When, however, the learner makes an adequateresponse as determined by the feedback loop (Cell A-7), thelearning value of a response (Cell G-7) is directly related tothe level of gratification (Cell F-7) which is, in turn, deter-mined by the level of response investment (Cell D-6).

The educational technologist who organizes learning systemsis primarily concerned with maintaining the learning value ofadequate responses by insuring that the learner perceives suc-cessive correct responses as evidence of achievement, ratherthan evidence that the work is too simple and a waste of time.The perception of achievement arises not only from the learningmaterials, which may or may not be viewed as important to thelearner, but even more from the learner's personal values, inter-ests and needs and the degree to which these are involved in thetotal learning context. An effective educational system is onewhere, regardless of prior learning history (Cell A-2) andinitial learning goals (Cell A-3), the learner develops the skillsrequired and is motivated to attain learning system objectives(Cell A-1) .

A high response frequency alone has been shown repeatedlyto be insufficient to improve learning performance when thelearner does not know the results of his efforts. On the otherhand, when the learner makes greater numbers of responses withina well-sequenced learning situation which delivers rapid feed-back, response frequency (Cell A-8) makes a substantial contri-bution to the amount and depth of the learning outcomes. Thus,

35

-34i;

high response rates are trivial or even counter-procuctive inthe absence of other key factors in learning; however, whenfactors such as feedback, spacing, etc., are present, a highresponse rate is highly productive and indicates an efficientlearning system (Cell H-8).

Once a learning system has been designed and put intooperation, it should be monitored both in terms of its totalfunction and its key components. The primary evaluation of thesystem involves the degree to which it is achieving its ownlearning objectives as specified in Cell A-1. This evaluationmay be performed in absolute terms. Suppose, for example, thatthe system objectives were limited to (1) a 9th grade readinglevel and (2) command of pre-algebraic arithmetic. Severalappropriate measuring devices could be used on samples of learners,employing equivalent but differing sets of items. The scorescould be set up diagnostically across each area to expose areasof system weakness. The rate of acceleration of the meanswould be an important index of system effectiveness, as wouldmeasures of dispersion, forms of the distribution, etc.

In addition, measures could be taken of such key sub-components as error rate (Cell C-4), learner reactions to mean-ingfulness (Cell A-5) and relevancy (Cell A-6) and the meanand variance of the response rate (Cell A-8). Once all theinformation had been collected, data would be analyzed to exposeareas of weakness of the learning system for purposes of revision.Where necessary, this data could lead to substantive systemmodification. In turn, this might result in changes in systemobjectives and components. (See the system modification loopflowing from system evaluation (Cell 1-9) back to systemobjectives (Cell A-1)). In addition to evaluations to upgradea learning system, evaluations should also be made to comparethe relative effectiveness of different learning systems whereeach seeks to attain similar learning objectives.

Evaluation instruments must measure the learner productfollowing graduation from a learning system. A pre-school sys-tem should be measured in part by ability to perform effectivelyin elementary school. Post-system performance and adaptationcould then be combined with the within-system information to

3G

provide criteria for system modification as well as new systemobjectives. New learning standards would evolve for the pre-school system as changes occurred in the elementary systemrequirements. In a complex, evolving society such as ours,educational technology must also evolve to concurrently conformwith and support changes in the technology of the total society.

Further, education cannot bypass its own moral obliga-tions. Amorality in education can be as destructive as condon-ing immorality. Childs describes the moral responsibility ofthe educator extremely well:

...The moral nature of education stems from the factthat schools are organized and maintained by adults,not by the children who attend them. Adults engagein deliberate education because they are concerned todirect the processes by which their children matureand learn to become participating members of theirsociety. A manifestation of preference for certainpatterns of living as opposed to others is thereforeinherent in every program of deliberate education.Schools always exhibit in their purposes and programsof study that which the adults of a sociw.:17 have cometo prize in their experience and deeply desire to nur-ture in their own children. Hence the curriculum of aschool is an index to the values of the particular humangroup that founds the school. It is because some con-ception of what is humanly significant and desirableis implicit in all nurture of the young that we may saywithout exaggeration that each program of deliberateeducation is, by nature, a moral undertaking. (Childs,1960; p. 96) .

37

-36-

THE FAMILY AS A LEARNING SYSTEM

The family is assumed to be a learning system which setsup contingencies for establishing and strengthening the twospectra of responses which we term personality, that is, thosecovering the spectra of habituated performance and verbal be-haviors. Initial infant skills, formed through relative ratesof success in achieving goals, become deeply imbedded and formthe basis for an interlocking set of related skills. The childwho is successful in attaining goals through language as aninstrumental tool develops an elaborate language network toincrease the range, speed and specificity of the goals he canacquire. Where goal success is based on psychomotor activities,these too proliferate. Finally, where skills are developedwhich annoy, frighten and obstruct adults as a device for attain-ing goals, obstructive and annoyance behaviors multiply.

In our society the home, church, and community have beeninundated with insoluble problems of mobility, technology,pollution, crime and violence at home and war abroad. Theeducational system appears to be the only institutional struc-ture which has the basic stability, time, personnel and equip-ment to prepare a large proportion of American youngsters foradulthood. Historically, of course, the educational systemhas been responsible for insuring adult readiness. However,in our epoch, educational problems are an even more crucialsocial imperative which directly affect our viability.

In addition, learning problems are increasingly moredemanding for students. At the same time, the resources ofthe home, community, and church provide appreciably lessbuoyancy and support to assist the educators in their moredifficult task.

If the school entry point is considered as kindergarten(or as first grade, as in many locales), we are consideringa human being with about 44,000 hours of life of which approx-imately 30,000 have been spent awake. During this period,the child has made literally millions of responses to achieve

38

'-37-

an enormous spectrun of goals. Some of his goal directedresponses are personal and direct; they involve pulling, grab-bing, reaching, grasping, holding, crawling, climbing, walk-ing, and other such behaviors where the goal is perceived andthen attained by self-effort. However, the great bulk of thegoal seeking responses, particularly in the early life stages,requires adult intervention. This indirect method of goalattainment is dependent upon the use of signals which indicatethe existence of need for the goal object. For the very youngchild, this entails a type of psycho-motor signal, epitomizedby pointing, gesturing, change in facial expression (e a0,*,smiling or frowning), reorientation of body position, holdingup both arms to be lifted, etc. Here the adult goal supplieris required to make an interpretation of the signal value ofthe expression or posture and respond with the appropriategoal associated stimulus, if possible. In addition to attain-ment of goals through direct and indirect psycho-motor means,the child has a battery of sounds which are interpreted assignals requiring adult intervention. At an early stage, theseinclude such sounds as crying, laughing, screaming, cooing,babbling, gurgling, and speech mimicry. In addition to thesevoice-box sound emissions, there are body sounds which serveas signals indicating interventional requirements such as clean-ing and diaper changes.

The central point is that, at the earliest life stage,there is a psycho-motor mode and also a sounding mode for goalattainment. The pattern of adult response to these modes,varies, of course, from family to family. The infant behavioroccurs in a cultural matrix involving physical space, a varietyof objects, and the interaction of a family social system which,regardless of its previous behavioral equilibrium, must adaptto some extent in accommodation to the new family member. Forpurposes of this analysis, it is assumed that families andindividual members differ along four major dimensions: (1)

Environmental Richness; (2) Symbolic Exchange; (3) ParentalResponse Thresholds; and, (4) Parental Responsitivity. Eachof these four dimensions is presented along a five point scalein Table 1 (The Family as a Learning System). We may, usingthis scale, evaluate the environmental matrix of the infant bypatterns such as (A-4, B-4, C-5 and D-.1 (A-1, B-2, C-1, D-2),

etc. The scales have been designed so that the higher the num-ber, the greater the value for accelerating the development of

39

VALUES

ENVIRONMENTAL

RICHNESS

(Levels of

Stimulation)

SYMBOLIC EXCHANGE

(Meaningfulness)

PARENTAL RESPONSE

THRESHOLDS

(Reactivity Modes)

Level 1

HOME AS CAPSULE

Family is isolated

and does not inter-

act with outsiders.

VERY LITTLE MEANING

Verbalizations rare

and responses are

brief and often mono-

syllabic.

VERY LOW THRESHOLD

Parents provide de-

sired goal stimulus

immediately rather

than tolerate noise

or tension-producing

activity.

Level 2

0ao

CD

NEIGHBORHOOD

CAPSULE

Family interacts

only with itself

and other families

within narrow

social and geo-

graphical area.

HIGH AMBIGUITY

Verbalization is

guarded.

Meaning is

opaque and listener

must guess at mean-

ing and emotional

content.

Speech

interactions very

rare.

Level 3

COMMUNITY LEVEL

MINIMAL TOLERANCE

Parents usually pro-

vide desired goal

stimulus rather than

tolerate noise or

tension-producing

activity.

PARENTAL RESPONSIVITY

(Feedback Modes)

NO FEEDBACK

Parent behavior unloving

and self-determined.

Overt behavior of child

has no impact on parent

attitude or behavior.

ACCEPTING-UNRESPONSIVE

Parent behavior is loving

but self-determined and

child has little impact.

Rigid parent behaviors

are only slightly

affected by child's

responses.

PARTIAL AMBIGUITY

Family interacts

with other families

and individuals

within the commun-

ity area.

Verbalization is

often guarded but

has wide range.

Interpretation is

sometimes given spon-

taneously but usually

in response to direct

question.

MODERATE TOLERANCE

Parents often provide

desired goal stimulus

but child often fails

to achieve it because

of noise or tension-

producing activity.

DUTIFUL-FLEXIBLE

Parent behavior main-

tains warmth and is

responsible to child.

Standards are reason-

able and child is aware

of own success or

failure.

Level 4

INTER-COMMUNITY

LEVEL

FULL COMMUNICATION

HIGH THRESHOLD

ACCEPTING-FLEXIBLE

Family interacts

Speech is informa-

Parents usually re-

Parents attentive and

with other families

tive.

Language has

sist producing goal

responsive.

Respond

and individuals in

wide range of tone

stimulus under duress

selectively to needs

other geographic

and content.

Sus-

of noise and tension-

quickly based on child's

locales.

tained responsive

interactions in

speech.

producing activity

and tend to respond

when child is socially

adaptive and non-

destructive.

response.

Flexibly

adapt as child matures.

',

Level 5

MOBILE INTER-SOCIAL

INTERACTIVE

FLEXIBLE THRESHOLD

RESPONSIVE

,nW

COMMUNICATION

DEVELOPMENTAL

,

...-.4

Family has wide

Speech is informed

a. Parents resist

Parents are loving and

social interactions

and has high symbol-

response to ten-

supportive and highly

across groups and

is content.

Great

sion-producing

responsive to child

in varying locales

facility with lan-

activity.

and apply shifting

and cultures.

guage, both common

and literary. Re-

sponsive to shadings

of meanings insure

comprehension.

b. Parents provide

alternative,

socially accepted

modes for obtain-

ing goal.

c. Parents systemati-

cally provide goals

for verbal and

adaptive behaviors.

standards to increase

skill levels. Planning

done to insure growth

in abilities under con-

ditions of adequate

success probability.

Verbal interactions

are sustained.

-39-

language and social skills. A comparison between two FamilyLearning Systems is made below.

42

t40-

COMPARISON BETWEEN TWO FAMILY LEARNING SYSTEMSTable 2

Scale Family AValue

A-4 Inter-Community Level

Family interacts withother families andindividuals in othergeographic locales.

B-4 Full Communication

Speech is informative.Language has wide rangeof tone and content.Sustained responsiveinteractions in speech.

C-5 Flexible Threshold

a. Parents resist responseto tension producing acti-vity.b. Parents provide alter-native, socially acceptedmodes for obtaining goal.c. Parents systematicallyprovide goals for verbaland adaptive behaviors.Verbal interactions aresustained.

D-4 Accepting-Flexible

Parents attentive andresponsive. Respondselectively to needsquickly based on child'sresponse. Flexibly adaptas child matures.

43

Family B

Home as Capsule

Family is isolated anddoes not interact withoutsiders.

ScaleValue

A-1

High Ambiguity B-2

Verbalization is guarded.Meaning is opaque andlistener must guess atmeaning and emotionalcontent. Speech inter-actions very rare.

Very Low Threshold C-1

Parents provide desired,goal stimulus immediatelyrather than tolerate noiseor tension-producing activity.

Accepting-Unresponsive D-2

Paren behavior is loving butself-determined and child haslittle impact. Rigid behaviorsare only slightly affected bychild's responses.

-41-

The child from Family A is in an enriched physical en-vironment, stimulated by variations in objects, colors, sounds,etc., contacts a wide range of other types of people, and hearsmany different vocal patterns (A-4).

The child from Family B, on the other hand, is largelylimited to the same four walls with little variation in form,color and sound patterns. The family members largely representrepetitive stimuli in terms of physical appearance, voice soundsand manner, etc. Child A is in an ever varying environmentwhich can provide him both stimulation and gratification. Themany objects he encounters have a variety of terms identifyingthem which, if learned, provide the basis for an expanding voca-bulary. Child A is also more advantaged in terms of the specifi-city of the language used in the home. Further, Family A usesinteractive speech whereas the speech patterns in Family B tendto involve only delivery statements and rarely include sustainedexchanges involving information delivery, receipt, delivery,etc. (speech chains). Thus, the Family A child has an immenselygreater opportunity to learn speech and to utilize language asa tool to obtain goal objects, social interactions, and to makeknown those wants which can only be satisfied by appropriateparental behaviors.

The D scale relates to the immediacy of the satisfactionobtained by the child as a consequence of a goal related response.In Family A, the parents are attentive to the child and providegoals quickly and selectively, taking into account the child'sneeds and state of development (Level D-4). In family B, onthe other hand, the child is given little attention; his respon-ses go unheeded, and his parents tend to his needs on the basisof parental convenience or some pre-determined schedule. Thus,the goal attainment level for any given sequence of responsewill tend to be considerably higher for the child in Family Athan for the child in Family B who obtains little or no benefitin terms of goal attainment from goal-related responses he makesto parents.

:42-

Finally, in Family A, the parents are selective as towhich kinds of responses produce regard. Crying, pounding,screaming, shrieking, complaining, and sulking responses are givenlittle attention unless there is some indication of pain ordistress. They tend to respond to facial expressions, bodymovements, voice sounds, and actual speech, to wait for moremeaningful types of behavior, and to be tolerant of highnoise levels. However, in Family B, the child, unable toobtain goal objects by other means, may increase his noiselevel, aggressive behavior (or withdrawal behaviors) to thepoint where the parents must take notice. If pareritL,, have a

low threshold, physical activity and mere noise may producegoal objects, social involvement, and gratifying events moreefficiently than expression, gestures, speech signals orlanguage.

Family A then, lay be viewed as a social system inequilibrium in which the new member learns how to maximizegoal attainment with minimal energy expenditure. To obtain hisgoals, the child in Family A learns to signal his parents in avariety of ways, and speech becomes a functional life tool.Child B, however, may become relatively nonverbal, aggressiveand impulsive since these expressions may be the most functionalmeans of goal attainment in his family.

Where speech is more functional as a goal generatingtool, it evolves and is elaborated; however, tha languageoriented child may construct elaborate fantasies to produce abroader variety of goals than the parents can or will deliver.The central point is, of course, that both children havedeveloped efficient response repertoires in terms of maximizinggoal achievement per unit of energy expended.

In a more precise convention then, we may simple statethat, all other things being equal, learning (L) always takesplace in the direction of maximal yield (M) per unit of energyexpended (E)

L = ME

where yield is considered to be the perceived goal. requirements

45

of the learner. As the child moves across his availableresponse spectrum, he tends to emphasize and accentuate thoseresponses which generally yield the greater return for theeffort made. The parental patterns that exist in the initialinfant learning environment generate the initial infant responsepattern which, over time, is elaborated into a complex responsenetwork. However, the kinds of responses which tend todo urinate are determined almost totally by the response tenden-cies of significant adults during the first years of life.

If we asrnme that the i'atra-familial learning contingenciessketched 'plat in Table 1 operate as a learning system, then itmay well be hypothesized that family interaction patternsdevelop four major categories of child goal attainment behavior:

(1) Static: low in motor and symbolic responses.

(2) Motoric: high activity, low in symbolic skills.

(3) Symbolic: high verbal skills, low motor skills.

(4) Indirect-Obstructive: Both motor and symbolicskills force others to act.

If we, assume that these initial goal attainment behaviorshave (in the jargon) great response strength and thus form abasic life theme, then later responses tend to be complexpatterns and variations of this initial behavioral melody.The tree leans as the twig is bent; the child is father to theman.

It may be that for thousands of years, the most adaptiveand successful human beings were high motoric--low verbal,and that such children would have been much favored over theverbalizing studious, sometimes challenging or contentiouschild. Certainly the motoric pattern would be highly adaptivein a hunting and farm culture. There are dozens of patternsin "The Family as a Learning System" Scale which would resultin such a pattern. Almpst any combination which would include

-44-

A-1, A-2 and A-3; B-1 and B-2; C-3, C-4, C-5; and D-2, D-3and D-4.

However, for the child in our culture, this is an eraof symbolic mastery, a meritocracy where early symbolic skillslead to classroom victories, quickly followed by the attainmentof technical skills as an engineer, lawyer, medical doctor,educator, or businessman. Cultures which emphasize symbolicskills for the very young may produce high performance adultsas an artifact of early symbolic stimulation, parental responsivityand feedback. The higher adult performance levels of themiddle class, the professionals and the upwardly mobile whoemerge out of the urban and rural ghettos, may reflect asimilar (though less probable) accidental pattern of inter-action in the home rather than some genetic template whichpredestined them for success.

In this analysis, the family as a learning system isthe crucial generator of the behaviors possessed by the veryyoung; genetic determinism is a metaphysics which utilizes thejargon and technique of research to propound predestination.Predestination theories ultimately lead to a kismet philosophy;it has been written (in the chromosomes) and, therefore, thepsychologist can only observe. Genetic determinism presumesthat systematic psychoeducational approaches will be eternallypowerless. In the present analysis, the alternative optionappears far more reasonable -- that is, it is assumed thatsystematic and organized efforts to improve the learningefficiency of our children will make substantial differencesin learner performance capabilities. We cannot afford ametaphysics of genetics to explain educability, any more thanwe can pin our faith on the infallibility of the teachingmachine, TV, or teacher creativity. We must face the necessityto (1) design pre-school systems to meet the specific lifeobjectives of children today; (2) develop and test protypesystems and evaluate how well they meet their objectives;and, (3) proliferate and continually evaluate, revise andupgrade the systems to improve the learning efficiency of thechildren.

-45-

A SYSTEMS PERSPECTIVE

Coordinated and planned efforts involving money,personnel, methods, equipment, materials, transportation, etc.,to achieve a given objective are classifiable as systems.A bank, a school, or an aircraft factory may be a system inthe sense that each involves an input and a complex method ofprocessing that involves many factors to attain a definableoutput. Systems are purposeful; their processing methods andquality control techniques convert a relatively raw inputinto a specified output. Some systems, such as banks, offerfew problems. The product (increased in dollars) is seeable,touchable, feelable, discrete, and above all, countable. Theinput of the system can be readily identified by agreed-upon accounting procedures and the efficiency ascertained bythe difference between numbers of dollars of input and numbersof dollars of output. The difference in dollars, whetherstated in earnings, percentage, stock dividends or profits, isan index of gain generated by the system. The efficiency orinefficiency of the system is easily judged by the relativerate of gain in output compared to other banks in similarcircumstances.

The input in a pre-school, on the other hand, is a childwhose capabilities are largely unknown and, the output, forthe most part, is of no concern to educators. Further educationinvolves non-observable capabilities which are uniquely combinedin each individual in unknown ways. Measures of learnerperformance level may be in serious error because of factorssuch as: language, including misunderstanding of instructions,usage, expression, colloquialism and dialect differences;resistance of the examinee; bias of the examiner; pooradminisration; failure to standardize the test environments;distractions; anxieties evoked in the examinee which suppressknown information or skills; preknowledge of test questions;errors in assessing the adequacy of the response, etc. Beyondthis, the numerical sequence of test scores fails to satisfythe unit distance requirement for cardinality and, in addition,the test items blithely span dimensions. This results in scoreswhere greater numbers do not necessarily mean greater knowledge

48

-46-

or skill along any clearly defined dimension. This in turnmakes all later statistical manipulation of dubious value.

This analysis holds that, in spite of the difficultiesof goal-product definition and measurement, and the complexityof our technology and social system, education is powerlessto solve its problems without highly sophisticated systems.Humans, unlike objects, are idiosyncratic and dynamic, varyfrom moment to moment, evolve with different learning histories,perceive, excite, conceptualize, and differ in their objectivesand values; thus, educational systems must be much moresophisticated than systems concerned with earnings, gettingmen to the moon, or processing food. Within this framework,learning systems for children should be concerned with developingthe following output products:

(1) A strong sense of personal independence andadequacy combined with the personal initiativeto establish goals and the stamina and flexi-bility to attain them.

(2)a. A strong language base that provides the termsnecessary for description, denotation, andclassification by category. Ability to useterms functionally as tools to organize and planin order to attain personal goals.

b. Skills in communicating to others at a verballevel and the attention and comprehensionnecessary to obtain information from others.

c. A sufficient level of reading skill to permitpre-school children to increase their vocabularyindependently, along with printing skillssufficient to record the words they know.

(3)a. Ability to work easily and readily with othersto cooperatively perform a task which couldnot be performed alone.

b. Readiness to work with adults as required,with minimum dependency on adults for supervisionand control.

49

c. Readiness to accept responsibility and assistothers where required and to accept assistancein the interest of achieving a personal orshared objective. All aspects of learnerperformance should reflect an empatheticawareness of the importance of the other'sneeds.

(4) The curbing of impulsive and anti-social behaviorswhich prevent learning by the individual learner,or interfere with the efforts of others to perform.Thus, a learning system should be so designed toshift response patterns, where required, fromanti-social or asocial behaviors to behaviorswhich provide basic satisfactions throughlearning itself, self-expression, and social-interactions.

The output measures for a pre-school learning systemshould correspond to the system objectives. The adequacy ofthe measures will be crucial for determining the effectivenessof the system and central to the problem of making comparisonsof the relative effectiveness of different systems. Thepresent measurement of performance in educational settings is,at best, inadequate. The development of learning systems willrequire new, imaginative and pertinent measuring methods.For pre-school systems, several kinds of output measures appearto be indicated. Performance should be sampled across thetotal range of life skills including:

(l)a. Language skills in which meaningfulnessof language covers verbal comprehensionskills, precision, denotation, and functionalvalue of terms in English and in learner'sfamily linked language.

b. Ability to classify and analyze higher orderterms into lower order relationships; understandingof mapping relationships related to the body,home, classroom, etc.; understanding of thefunctions of objects and related terms. Thedepth, extent and precision of language termsshould also be known.

50

-48,-

c. Adequate reading skills (as assessed by diagnosticreading skill tests which provide valid indicesof areas of uncertainty of sound and letterpattern relationships at varying levels ofstimulus and response complexity).

(2) Social interaction skills with peers and adultsin standardized performance task situationsinvolving goal attainment.

(3) Goal attainment skills involving measures ofstamina, flexibility, perseverance and ingenuity,

(4) Adequate degree of control over impulsive avoidanceand aggressive behaviors under conditionsinvolving combiried efforts with others.

(5) Ability to utilize and act to attain goals atvarying points in future time.

(6) Ability to maintain attention to tasks overextended time periods in the absence ofimmediate personal gratification or reward.

Pre-school learning systems, in particular, must bedesigned to overcome handicaps generated in the home and community,at every socio-economic level. They should provide the basicskills necessary to meet the standards of elementary schoolwhile preparing children to contribute, participate and adaptto the infinite variety of conditions in the home andcommunity. Above all, our pre-school systems must be viewedas providing the keystone learning for adult Americans.

The freedom to make choices in a democracy implies thatcitizens are sufficiently skilled to make those choices whichfunction for goal attainment. Education must move beyondteaching punctate elements to satisfy multiple choice testrequirements, and provide the language, planning, emotionalcontrol and flexibility necessary to maintain control over one'spersonal destiny. This writer views a well articulated patternof learning systems, beginning with pre-school, as offering thefollowing capabilities through education:

51

L49-

(1) Developing a commitment to personal freedomthat extends not only to the individualhimself but to all members of a free society,functioning within a constitutional democracy.

(2) Developing problem-solving skills that requireinter-relating, complex concepts, establishingsequences of activities, and the flexibilitynecessary to adapt to situational variations.

(3) Developing sufficient emotional control tocontinue to perform effectively under difficulty,whether as a member of a team or as an in-dividual.

Learning systems, properly designed, can become the basicagency for developing the complex skills required to feeltechnically and socially adequate to function as a free in-dividual in our technical society. (See Figure 2, The Wedge ofChoice and Freedom In a Hypothetical Thirteen Learning ElementSituation).

A Preliminary Learning System Overview:

There are eight crucial considerations in designing alearning system: (See Figure 3 - Lattice for LearningSystem Prototype).

(1) Theory (Cell A-1): The theory should be fullyarticulated with interlocking assumptionswhich provide direction, focus, and a guidingrationale for those responsible for conceptualiz-ing a system design. Above all, it shouldprovide a basis for selection of methods ofassumed greater effectiveness. The theory mustbe directional (rather than circular) so that itcan be assessed as adequate or inadequate.depending on its predictive power. Further,it should be capable of modification and im-provement as a consequence of data obtained

52

-50-

A SYSTEMS APPROACH TO PRE-SCHOOL EDUCATION

The Wedge of Choice and Freedom In AHypothetical Thirteen Learning Element Situation

Number ofLearningElements

13

12

11

10

9

8

7

6

5

'4

3

2

1

r.C>0

As the learner increases the number of successful choices, hebecomes increasingly free to obtain personal goals. The number withineach successive segment of the wedge (choice fields) are illustrative

only. They indicate that at each successive choice field, the numberof choices increases by a factor of 2x + 1 where x is the number ofchoices which could be made in the just previous choice field.

If the learning elements are considered to be letters of the alphabet,this wedge indicates that when one letter is known, the learner can readone word, two letters permit him to read three words, three letters permit

him to read seven words and so on. When all thirteen learning elementshave been mastered, then the learner can make all possible choices toattain any goal involving these thirteen elements. At this point he has

total freedom within this (thirteen learning element) field of choice.

53FIGURE II

-51-

from realistic tests of the performance of outputlearners.

(2) The Philosophic Aspect of a Learning System (CellA-2) :

A theory specifies certain formal assumptionsabout the learning process; however, a philo-sophy provides the direction, boundariesand limits for the theory's application. Thetheory functions within a value system whichspecifies how it will be used. A systemwhich functions effectively to control behaviorcould be used to enslave or to free. Thephilosophy underlying the use of the systemdetermines the way it is put into effect.

(3) The Assumptive Framework (Cell B-2): Theconceptual convergence of the theory and theeducational philosophy form the basis for thetotal set of assumptions used to develop agiven educational system. "The assumptionsunderlying the theoretical system combine withthe philosophic assumptions to form theassumptive set on which the learning system isbased. These establish the degrees of con-ceptual freedom and the limits which will controlthe design of the new learning system.

(4) The System Objectives (Cell C-3): Once thesystem assumptions have been established,educational goals must be sharply and unambiguouslydelineated. An educational system, for example,can limit, itself to learning of a particularbody of content or it can seek to affectbroader ranges of behavior such as aggression,withdrawal, and probleth solving skills. Systemobjectives are absolutely limited by assumptionsin the theory, and philosophy, whether or notthese have been formally exposed. They may be

54

ASY

STE

MS A

PPRO

AC

HT

O

Lattice for L

earning System Prototype

4111111#1011 TO IMOD= TIM

COPYRIGHTED MATERIAL HAS BEEN GRANTED

Witrok

!Am

y&

TO ERIC AND ORGANIZATIONS OPERATING

UNDER AGREEMENTS WITH THE U.S. OFFICE OF

StS

EDDCATION. FURTHER REPRODUCTION OUTSIDE

THE BIC SYSTEM REQUIRES PERMISSION OF

%PS)

111111.DT OREL"

PRE

-SCH

OO

LE

DU

CA

TIO

N

Operational

System

Test Phase

; E0C

System Model

Specifications

System

Objectives

Philosophy

Direction

and Scope

Criteria

for Model

12

Model

Implementation

Learner

Input Stage

45

Successive Stages In System Development

A lattice indicating the key stages involved in developinga learning 'system.

6

Quality

Successive

Control

Evaluation of

Methods

Learner Output

7

FIGU

RE

III

8'

ID 1970 M

YR

0 N W

O O

LMA

N

as broad as a commitment to the whole person,or as narrow as a change in the frequency ofa learner who pulls the lever on a pin-ballmachine.

(5) Establishing the Specifications for the Proto-type ystem (Cell D-4): All planning and spec-ifications for the system are essentially theestablishment of a network of interlockingactivities which utilize the assumptions tosatisy system objectives. The design of theclassroom, the learning materials, equipment,learner behavior, role of the teacher and otheradults, administrative methods, discipline, thesequencing of inputs, the classroom decor, etc.,must be specified in detail and interlock intoa single organic system. The basic constraintduring the planning phase is that each aspect ofeach plan must be based on criteria (Cell A-4)which are consistent with the assumptions, philosophyand ultimately the basic theory under test. Thelevel of detail of the specifications should besufficient to put a team to work to transformthem into an operational learning system, muchas a set of blueprints is used by an architectto specify construction requirements.

(6) Operational Readiness (Cell E-5): Once theprototype of an innovative system exists, anew range of activities is required for itto function as designated in a community orschool context. These activities include suchareas as: initial selection of children;community and parent relationships; teacherselection; relationships with the sponsoringorganization and the local school system; andcareful orientation of political leaders andpersonalities to avoid misunderstandings.Community understanding of the system and itsobjectives is fundamental to a meaningful

-54-

operational test. Once system specificationshave been operationally transformed, the vitalproblem is to insure that staff aresufficiently well trained and are knowledgableabout the system. The operational staff'sresponsibilities should be committed to testingthe system as planned rather than the im-provisation of casual alternatives. In addition,teacher training methods, in-service training,teacher critiques, etc., must be planned toinsure that the learning system operateswithin its own theoretical and philosophicalboundaries.

(7) System Diagnostic Measures (Cell G-7): Theevaluation of system performance requires acomparison between what happens in the courseof actual system operation and what wasplanned in the system specifications. Suchevaluations cover learning rates and whetheror not the equipment and material functionas planned in terms of teacher and learnerperformance, parental reacticn, etc. Thediagnostic evaluations are designed to insure,to the extent possible, that each of the elementsof the operation system function as initiallyplanned. The diagnostic method should exposenon-compliance or misinterpretation so thatsteps may be taken to correct the use of aparticular component. On the other hand, wherea component is being used properly but fails tofunction effectively, it should be modified,eliminated and/or replaced. This aspect of theevaluation should be indifferent to those partsof the system which are functioning effectively.It is a method of evaluation by exception.

(8) System Output Measures (Cell H-8): The purposeof a learning system is to insure specific kindsof life skills after some pre-specified time

57

-55-

interval. System output measures should bedesigned to provide an accurate measure of thesystem's effectiveness in attaining its ob-jectives, as related to the total learnerpopulation. Such measurement might coverlanguage skill gains; change in social effective-ness in interactions with peers and adults;level of emotional control as related toaggressive, impulsive, withdrawal and anti-socialbehaviors; ability of the learner to achievegoals on his own initiative and willingness tosustain efforts to achieve goals; and the useof creative ingenuity in problem solving.

A very important measure of system effectiveness lies indetermining whether changes are restricted to the learningsetting or generalize to the home and community. The most vitalmeasure in this analysis involves the learner's adaptability andperformance after graduation from the pre-school system, thatis, in first grade and later.

The preceding preliminary view of the components of alearning system will be presented in somewhat greater detailin the following sections. A pre-school learning system, nowin its operational phase, will be used to illustrate systemrelationships. This system has recently completed its firstoperational year and its actual output capabilities are notyet known.

2. The Micro-Social Learning Center, a New Jersey StateDepartment of Education Demonstration Project, whichoperates in cooperation with the Vineland Board ofEducation, Vineland, New Jersey.

58

TOWARDS A PSYCHO-SOCIAL THEORYOF

CLASSROOM LEARNING

Classroom as a Life Simulator

In a complex, mobile society which is ethnically andlinguistically heterogenous, the pre-school is the only in-stitution which can provide the cohesion in language and behaviorthat is imperative if later school experiences are to havevalue. Pre-schools are the most important institutionalizedlearning systems because they provide the matrix skills whichmust transfer to elementary, secondary, college, vocationalperformance, and to the family and community.

A democratic, individualistic, technical society posesspecial problems for education because the adult product mustpossess: (1) a broadly based language delivery and compre-hension capability; (2) social skills required to work in groupsas a team leader, member and/or subordinate; (3) tolerance forstress and frustration to resist impulsive, avoidance, andaggressive behaviors in work, family and community contexts;and, (4) ability to plan, organize and maintain goal directedactivities over extended time periods. The classroom complex,when viewed as a system for generating responses, must be soorganized as to insure life survival skills for learners;learner responses must be transferrable from the classroom andfunction to provide goal attainment capabilities.

The simulator is a method for controlling complexsituational variables, insuring the safety of the respondent,and developing skills to the level necessary to insure certainskills levels for the learner. The following excerpt byGagne is alucid indication of the value of the simulator inteaching complex responses patterns:

...First of all, as has already been implied, asimulator attempts to represent a real situation

in which operations are carried out. (By'operations' is meant a set of events in which aman or men interact with machines or with theirenvironment to bring about a particular result.)...Secondly, in representing a real operationalsituation, a simulator provides its users withcertain controls over that situation. It mightbe argued, in fact, that this characteristicconstitutes the major difference between asimulator and the operational situation itself.The latter is usually, by definition, uncontrolledand subject to unpredictable variations; whereasthe simulator provides for control (and often, plannedvariation) of these same aspects of the realsituation . . .Thirdly, the simulator is de-liberately designed to omit certain parts of thereal operations' situation. . . .In all simulators,a greater or smaller portion of the operationalsituation is purposely omitted. . . (Gagne, 1965;pp. 225-26)

In each learning stage, a learning system should in-creasingly simulate the responses required for the next lifestage. This is stated more formally below.

Assumption #1

The Classroom as a Life Simulator:

The classroom is most functional when used as asimulator for the development of responses thatare transferrable to the home, community andnext life stage.

Human Learning and Socialization: Though animal learning isbasically controllable through objects (reinforcers) whichreduce primary and secondary drives, this is not the case for

human learners beyond the first few months of human life, as thewomb giving birth and support to human responses is not primarilyphysical but social. Human learning takes place, mainly inthis analysis, as a consequence of events (intra-familialand social advantages) rather than through the ingestion offood and water or the avoidance of externally inflicted pain.

Assumption #2

The Human Learning Assumptions:

(a) Learning in humans occurs when a positive affectshift takes place following a response.

(b) Learning is assumed to have taken place when apattern of responses is systematically modifiedso as to increase the probability of attainingpre-established goals.

(c) Human learning is facilitated when learnersperceive themselves as achieving a requiredstandard under conditions where there is aperceived probability of failure.

(d) Human learning is fundamentally linked to thesocial (rather than the material) consequencesof a response.

(e) Material objects have learning value to producepositive affect only to the degree that they areperceived as indicators that some perceivedsocial standard is being met.

(f) Learning takes place most efficiently underconditions where the perceived consequences, ofa response involve the attainment of goalsperceived as important within a society in whichthe learner is a member.

61

-59-

(g) Learning systems increase in efficiency whenlearners perceive that the attainment of goalsis tightly linked to upward shifts in the roleand/or status of the learner in the society.

Motivation as Learned Behavior: Motivation is defined as thedegree to which goal-directed responses are sustained beforean individual shifts from the effort to reach a goal to somealternative behavior. The degree to which goal directedresponses are maintained is held to be a function of priorgoal achievement. Those who have learned to sustain responseswhich achieve goals learn to respond over extended periods oftime; they become motivated for future goals. This outcomeresults from a learning system which: (1) provides goals ofinterest to the learner; (2) systematically increases the timeand effort required to achieve goals; (3) insures a richreward history for goal achievement over extended periods oftime, and, (4) provides high probabilities of goal achievementwhen linked to active goal directed response patterns. Theseconditions generate motivated learners, in this analysis,regardless of the prior level of motivation.

Motivation can be systematically developed as an integralaspect of classroom progress by designing the system to pro-vide a learning history where the anticipated gratificationin goal achievement is perceived as outweighing the energynecessary to attain it. If Motivation (M) is conceived to bea vector consisting of willingness to expend energy (E) and aperceived goal (G), then the perceived value of G must alwaysexceed the perceived cost of E. The learning system must beso designed that G)E.

Assumption #3

The Motivation and Self-Image Assumptions:

(a) Motivation is learnable; the first and most crucialproduct of a learning system is the developmentof learner motivation to achieve systemgoals.

-60 -.

(b) Humans in a social setting have a greater 1-.;:klirningpotential than humans learning in isolation:learning which improves social status producesan affective increment which increasesthe general level of learning efficiency.

(c) Learning systems which involve perceived upwardstatus shifts for all learners, on the basisof progress within the system, result inimproved feelings of self-worth, greater self -confidence, and increased motivation toperform.

(d) Involvement in educational materials will befacilitated to the degree that such learningis linked to role and status gains in a socialcontext perceived as important to thelearner's image of himself.

(e) Improved self-image occurs when the in-dividual (1) shifts roles which indicate anincrease in his status and importance as agroup member and (2) is responded to by othersas being instrumental for attainment oftheir goals.

(f)_, Evaluations of learner performance function toimprove self-image and increase motivationunder conditions where such evaluations areevidence to the learner of mastery rather thana threat to expose learner inadequacy.

Symbolic Skills: Our present social system provides rewards andoffers upward mobility to those who have an extensive andprecise language base. The minimum level necessary is theability to deliver information by mouth or in print, and tocomprehend verbal and printed information accurately. Minimalsecond level skills involve organizing information, planning,

3

-61-

scheduling, and describing relationships, and the ability toclassify, develop procedures, forms, lay out work requirements,etc. At higher symbolic levels, there are requirements togeneralize to the present from the past, using such things asdocuments or a wide variety of existing classification systemsand morphologies, and to utilize this ability in skills whichinvolve repairing of equipment, malfunction analysis, officeadministration, etc. Beyond this, there are requirements fordeveloping and evolving new equipment, methods, procedures, andtechniques across business, science and the supportingtechnologies. Such effort involves understanding of complexprinciples and concepts, either compressed into formulae ororganized in an ever-changing, interlocking specializedjargon.

(1) In short, the ability to manipulate languagewith minimal distortion and to classify andconceptualize is the highest priority liferequirement for adjustment to the imperativesof a technological meritocracy. The initialdiscrimination of textures, forms, colors,smells, distance, weights, etc., must quicklylead to identification of objects, events andpersons and then to classification andgeneralization. The language code must beestablished as a precondition for the decodingof print which we call reading.

(2) Thus, the education of the young, whetherfrom disadvantaged or typical homes, mustfocus on the development of an extensivevocabulary base so they may learn the valueof classifactory and conceptual terms asa means of compressing information andexposing critical relationships.

However, the energy involved in developing an extendedlanguage base (including identificational, classificationaland conceptual terms) will not be expended unless language

64

functions to produce goals more efficiently than non-languageresponses. The development of symbolic skills in a classroomcontext therefore requires that language be learned as afunctional tool for achieving important perceived goals. Toshift learners to the use of language, goals must be attainablewith minimal delay and be perceived as relevant by the learner.Language linked energy will be expended only if word use isperceived as producing goals more efficiently than the existingresponse mode.

Assumption *4

The Response Pattern Shift Assumptions:

(a) Language learning will occur only to thedegree that language produces learner goalswith lower energy expenditure than non-language behaviors.

(b) Those classes of response which function mostefficiently for the learner to achieve goalswill displace less efficient responses inthe response repertoire.

(c) Where the individual perceives himself to be amember of a social system, those behaviorsand activities which result in perceived statusgain will tend to displace competing responses.

(d) A classroom which operates on the basis of learnergroup norms will be more efficient than onewhere learner activity is an outcome of require-ments based on authority.

(e) Where learners interact directly with each otherin dyadic relationship, learning contingenciesare more efficient.

6

-63-

Learner attitudes on self-worth improve wherelearners perceive themselves as continuing toachieve learning goals under conditions wherelearning partners change frequently.

The preceding assumptions are specific to the Micro-Social Learning System. This pre-school system involves thedevelopment of the spectrum of skills necessary for adequateperformance in an elementary school system.

The theoretical framework for Progressive Choice ispresented below. While the theory is more comprehensive inapplication than its use for the resolution of pre-schoolproblems, it is presented here at the broader theoretical level.

The Progressive Choice Positions:

The theoretical framework underlying the design andorganization of the Micro-Social Learning System is termed TheProgressive Choice Theory. This theoretical position, firstexplored in 1955, has been applied to the development of avariety of learning systems in education and in varioustechnical learning areas. (Woolman, 1955, 1960, 1962, 1964).

Constantly Expanding Fields of Choice:

Learning is fundamentally linked to the control ofchoices to which the learner is exposed. Where the environmentis very rich and stimulating, the number of choices is verylarge. Before learning has occurred, the probability of acorrect choice will be very low in a highly stimulating learningenvironment. Learning systems must control the probability ofa correct choice based on: previous learning history; kind ofresponse required; number of choices offered; the kind offeedback given, etc. A Progressive Choice System providesnew elements only on the basis of proven skills in making thecorrect choices for all previous elements and functional com-binations. Learning takes place most efficiently when organized

CG

=64--

into successively expanding fields of choice where mastery of agiven choice field must occur before learning exposure to thenext area becomes possible.

Within each successive choice field the learner chooseswithout constraint. As he demonstrates initiative and skillsin attaining personal objectives independently within a givenchoice field, he moves to the next choice field and so on untilhe can attain any personal objective within the given domain.Ideally, on mastery of the final level, he could make any chainof choice (decisions) to attain whatever goals he might wishwithin the given informational and skill domain.

The Progressive Choice System involves: (1) analysis ofall learning elements which are involved in making successfulchoices for total mastery; (2) organization of these learningelements into a network or Learning Lattice (Woolman, 1962).These lattices are so arranged that successive elements can belearned individually and then combined with previous elements ina network of ascending conceptual or organizational complexity.

The Progressive Choice Assumptions:

(1) For learners to obtain gratification fromresponses they must perceive their learningtask as offering a real probability of error.

(2) In organizing a learning system, the expectedlearner error level must be controlled toinsure that(above .75)frustration

success probabilities remain highthroughout learning to reduceand avoidance reactions.

(3) Learning systems should be so designed thatthe perceived probabilities for failure aresubstantially greater than the actual pro-bability of error.

67

,65-2

(4) The greater the number of responses madewhich result in a positive affect shift by thelearner, the greater the rate at which learningtakes place.

The Sequential Hierarchy:

Learning is facilitated where the elements to be learnedare ordered in such a fashion as to combine successive learningelements (LEs) in a mutually inclusive fashion. First LE1is learned to the criterion, then LE2 then LE12 are combined,the LE3 is learned individually and in combination and so on.At all times the learner is required to make choices to estab-lish mastery and all materials are organized to emphasize function.

LE12

LE2

LE1 2 3

LE3

Diagram 1 shows that all materials to be learned areorganized as mutually inclusive and subsumptive (LE12 and LE2,LE1, 2, 3 subsumes LE1, 2 plus all base line cells).

The Sequential Assumption:

Learning complex materials to different levels of

68

L66-

abstraction and/or complexity is facilitated whenthe materials are pre-organized into a structure whichprovides for the learning of individual elements whichare successively combined in a mutually inclusivesubsumptive fashion.

In addition, there are certain rules of thumb in ProgressiveChoice Learning Systems which include such relationships as:

(1) Controls must be exercised over learner errorprobability at all points during learning.

(2) Tension levels which reach the anxiety thresholdincrease error rate and tend to producemechanized and stereotyped responses. Efficientlearning systems cannot exist under con-ditions of authority where punishment, favor-itism, scolding, scapegoatism, etc., can occur.

(3) Irrelevant, digressive, aggressive, withdrawaland anti-social responses reduce learning effi-ciency and should be minimized within thelearning context.

(4) Learning sequences should be so organized thatlearners continue to obtain gratificationfrom adequate responses over extended timeperiods.

(5) Ideally, learning is complete when the learnercan make all possible choices successfully toachieve personal goals. Freedom, similarly,is defined as the capability to make allchoices necessary to obtain a personal goal.Freedom, in an educational sense, exists onlZwhere sufficient learning has occurred for theachievement of personal goals.

69

-67-

An assumptive network or theory should be capable ofgenerating an infinite variety of learning systems. The con-trolling factor in developing a particular system is thephilosophy of education which forms the ground in which thetheory 3.s set as the figure. The theory and philosophy, takentogethe , converge to establish the framework within which theeducational learning system is imbedded. The beliefs and con-victions of the system makers (beyond the theoreticalassumptions) are vital elements in shaping the system throughthe use of the theory. If the theory is powerful, it can beused to produce conformists or individualists, sterile orexpressive students, alienated or empathic persons. Theoryis indifferent to whether students learn by rote or solveproblems by individual initiative and imagination. Aneffective learning theory is independent of the kind of learn-ing which takes place. The theory is the vehicle; what isdone with it is the highway selected by those who possess thetheory. The educational philosophy forms the avenue and goalsviewed as important to the learning system builders. Thephilosophy which controls the use of a theory must be madeexplicit to the system staff used for designing the prototype.Without the underlying philosophy, the staff would lack a mapwhich defines the objectives and functions underlying theireffort.

70

THE PHILOSOPHIC POSITION

If quality of manpower is the fundamental nationalresource, then improvement of the quality of our manpower re-sources must have the highest national priority. It may be farmore important from the viewpoint of national policy andnational survival to develop effective pre-schools rather thanA-Bombs, SST's or space voyagers. If we accept the reasonablethesis that adult behaviors are strongly influenced by ex-periences during the first six years of life, then it may wellbe that our social survival depends on the development ofoptimal methods of early education. Pollution, drugs, crime,alcoholism, welfare, and riots, grip the attention of newsmenbecause they provide immediate shock, personalities, and briefmoments of high drama suited to modern media coverage. However,these adult behaviors indicate that a substantial proportionof American adults feel themselves to be inadequate to meetadult problems. Various forms of escape such as drugs, alcohol,viewing of media, etc., are utilized as a means of minimizinganxieties and apprehensions. Crime is essentially characterizedby: (1) failure to maintain emotional control under conditionsof stress; and, (2) a requirement for property in the absenceof the life skills and/or credentials necessary to purchasethe objects desired. In essence, crime may be defined as thefailure to limit responses within the boundaries defined (bylaw) as aeceseary to protect the rights of others. It isessentially compulsive in the sense that the indimidual feelscompelled to perform an act beyond the limits of the allowableresponse spectrum. Possible punishment does not deter becausethe respondent is operating in immediate time to satisfy animmediate overriding and compelling need. Our increases inescapists and in crime stem from the inadequate preparation tomeet the demands of an individualistic technocracy. Thechurch, the community, as well as many homes, no longer pro-vide the means of establishing codes of behavior, character,and moral commitment -- rubrics that indicate well organized,socialized responses and well-defined response boundaries. Inaddition, where skills have not been developed in planning

71

future goals and in moving through sub-goals, tension is re-duced in immediate time through physical agents (such as drugsand alcohol) acting out, violence, sensual and/or sexual re-lease, and/or retreat into the vast panorama of escapistspectator activities on which large industries have been built.

Other societal problems are less in the spotlight. Forexample, in industry and even in universities, turnover ratesare high; employees are uncommitted and show increasinglyhigher rates of absenteeism and job mobility. Job trainingprograms founder whether supported by the government or in-dustry. The fact is that the difficulties involved in trainingadults are now becoming apparent. The stability of adultperceptions, habits, language base, goal acquisition skills,task involvement, and responses to satisy immediate needs arebased on ancient, deeply organized habits, all of which havebeen rationalized into attitudes and beliefs. When an adultpersonalityof behaviorthen, often

is formed, modification of even limited piecesis enormously time-consuming, expensive, and evenshortlived.

Behavioral modifiability is limited by the history ofprior experience. Experiences which result in anxietiesrestrict responses and limit the response spectrum and be-cause anxieties are cumulative, the number and kind of availableresponses tend to become increasingly limited with increasingage. Over time, more and more personalities and eventsbecome associated with apprehension, pain, and frustration.Avoidance of these areas to reduce anxiety increasingly limitsbehaviors. Our national manpower programs, to become evenmodestly effective, must go far beyond job skill and literacytraining. Much more sopisticated learning systems will berequired to produce the perceptual and behavioral changesthat lead to genuine upward mobility and substantial contri-butions to the tax base. The development of basic lifeskills at an early age is a simple and straightforward solutionto the national manpower problem.

Future problems involving our manpower pool, crime,escapist behaviors, etc., would be reduced if learning systems

72

for pre-school children would result in: (1) deep satisfactionsand skills in the area of language; (2) enjoyment in workingalone or with others to achieve short and long-range goals; (3)skills in controlling impulsive and aggressive behavior; and,(4) ability to tolerate present frustration to achievegratification in the future.

Educational learning systems, which can provide suchdeeply habituated activity patterns would result in elementaryand high school students who would be: (1) competent to performtheir academic tasks; (2) able to obtain some level ofsatisfaction (rather than anxieties) in the classroom context;and, (3) able to enter the job market, before or after collegeswith a much broader range of skills, fewer anxieties, and abetter self-image. When such persons enter their adult years,their gratifications should be linked to social effectiveness,achievement of goals, and meeting personal standards that arewithin the boundaries of a personal and defined ethical system.Education, to the degree that it provides skills in achievingpersonal goals, is the essential prerequisite of freedom.

Personal freedom is a political given in an individualistic,democratic society. Trapped by limitations of language,caroming in response to emotional needs like billiard balls,unable to live or work harmoniously in a sustained fashion,and, lacking useful skills on the job market, large numbers ofAmericans are paralyzed, impotent and frightened. They reactto farces impinging on them but lack the simple fundamentalproperty of freedom, freedom to decide on one's own life goalsand the personal resources to work towards their achievement.

Allport (1955) makes this essential linkage betweeneducation and freedom very pointedly in the following passage:

Similarly, relative freedom, we know, dependsupon the individual's possession of multiplepossibilities for behavior. To state the pointparadoxically, a person who harbors many determiningtendencies in his neuropsychic system is freer

-71-

than a person who harbors few. Thus a person havingonly one skill, knowing only one solution, hasonly one degree of freedom. On the other hand, aperson widely experienced and knowing many coursesof conduct has many more degrees of freedom.It is in this sense that the broadly educated manis freer than the man narrowly trained . .

(pp. 84-85)

Thus education's fundamental value in a democraticsociety is the development of those skills and capabilitiesnecessary for some degree of upward mobility. Education mustprepare its graduates to be willing to invest their energiesinto the future with a reasonable probability of success.Effective educational systems are social instruments whichinsure that adults can enjoy the fruits of their society.To the degree that men are powerless to choose, they lackfreedom. Efficient learning systems should provide thoseskills ultimately necessary to make appropriate choices andjudgments, and the stamina and flexibility required to attainmajor life goals. Failure to provide these skills violatethe essence of educational responsibility.

In addition to developing personal independence andgoal attainment skills, this philosophy holds that educationalsystems have the responsibility to provide those capabilitiesrequired for participation in the society as a citizen,community member and family member. Further, an effectiveeducational system should be capable of developing languageskills, social interaction skills, self-control and motivation,to the degree that such skills are required for effectiveadult performance. Educational systems at the pre-school levelshould possess the resources to overcome any handicaps imposedby the home and community. By the time of entry into theelementary school system, the pre-school pupil's skillrepertoire, behavioral control and ability to adapt should besufficient to insure acceptable performance. On completionof the entire educational system, graduates should be equippedwith the means to meet their general adult responsibilitiesand also maintain some degree of upward mobility based on theirown decisions.

74

The Learning System Objectives

The theory combined with the philosophic frameworkcomprise the assumptive framework and set the stage for es-tablishing the objectives of the system. In establishingobjectives, in the interest of economy of time, money and per-sonnel, non-essential and redundant skills should be avoided.Objectives of learning systems are strongly affected by othersocial institutions; however, where these institutions failto provide certain kinds of understanding, such as basic socialawareness, the educational system is obligated, as an adultreadiness system, to do so.

A pre-school system should move across all socioeconomiclevels, have no barriers based on familial or community handi-caps in language or behavior, and be indifferent to variationsin color, sex and age. Further, it should be capable ofaccommodating to sharp differences in language skills andemotional states.

The general objectives of a pre-school learning systemare, in this view:

(1) To insure the spectrum of behaviors and skillsnecessary for effective performance in anelementary school setting.

(2) To provide a basic set of behaviors andskills which will facilitate adaptationand effective performance of individualsin the home and community.

The General Terminal Objective:

The terminal objective of the learning system should bea broad statement which embraces its purpose and the range

and level of the skills required, such as:

To provide all the skills necessary for pre-schoolchildren with and without cultural and/or languagehandicaps to perform effectively at the first gradelevel, despite the substantial variation in schooladministrative procedures, curriculum, and differencesin style, manner and expectancies in the elementaryschool setting.

The Pre-school Terminal Objectives:

These terminal objectives reflect those key aspects ofbehavior which the system model should provide to meet its owngoals. In this writer's view, these objectives should bespecified at considerably higher levels than any given situationmight require, as the children will be distributed across abroad band of schools and teachers. The system's learningrequirements should cover those response requirements whichall learners can be expected to face across all first grades.In addition, the system should be designed to develop habitsof action which extend beyond the first grade so they willnot be of limited value. The following system objectiveswere employed in developing the Vineland Center:

Language Objectives:

(a) Develop a language base sufficient tomeet the range of performance requiredin first grade, the home and thecommunity, that is, a capability toindependently extend the languageoutside of the classroom learningcontext (2,000 functionally usefulwords);

(b) Develop reading skills to avoid dependenceon mouth to ear communication as the basisfor language growth (300 words);

2-74-

(c) Develop take-home materials to producesocial interactions in home and communitycontexts specifically related tolanguage expansion;

(d) Organized classroom activities which givevalue to the functional meanings of termsused in the home and community.

(e) Train parents in the learning system so theyoan facilitate the improvisation of relatedactivities in the home to increase therate of language growth.

Social Interaction Objectives: Develop skills in social inter-action, such that learners will:

(a) Actively assist others in their mutual interest;

(b) Increase the rate of verbal exchange touse their new terms;

(c) Develop skills in improvising activitieswhich can be done best on a cooperativebasis;

(d) Use social interaction as a means ofshifting roles to improve their self-image;

(e) Use social interaction as a means ofavoiding the development of socialbarriers related to differences insex, language, color, age, etc.;

(f) Use social interaction as a means of

77

transferring information from higherskill to lower skill learners;

(g) Use social interaction as a means of re-ducing anti-social behaviors.

Motivational and Task Involvement Objectives:

(a) To develop in students the capability tomaintain response patterns to a given taskcovering a period of one hour;

(b) Develop a high rate of self-initiatedresponses to printed material;

(c) Develop personal involvement in Centerlearning tasks, so that learning itselfranks higher in producing personalgratification than do motoric, aggressive,and avoidance behaviors;

(d) Develop satisfactions in performing self-initiated expressive activities whichextend beyond what has been formally pre-sented in the learning system;

(e) Provide a broad range of learningactivities which are self-initiated, expresslearner preference and do not require im-position by the teacher;

(f) Provide situations which develop bothimmediate and delayed rewards to permitlearners to project into future time, developfrustration tolerance, and learn planningskills for attaining rewards in the future.

78

CRITERIA AND SPECIFICATIONS

The criteria for generating specifications for theprototype serve as a guide to staff to plan the operationalprototype. Criteria are written in fairly concrete terms andconform to the theory, philosophy and objectives; they functionas an action tool to guide staff and should not unduly limitimagination and initiative.

The following illustrate the kind of criteria employedby the writer ir. setting up system specifications:

(1) The system will emphasize the developmentof responses and deemphasize stimulus enrichment.The environment is conceived as a responsegenerating system.

(2) There will be a minimum level of non-relevant stimulation which could intrude ordistract learners from the given learningtask.

(3) Responses will be learned across a broadspectrum. Language, for example, will includespeech comprehension and reading; however,competing responses that reduce learningrates, such as those involving aggression, with-drawal, and those behaviors which prevent othersfrom learning, are to be reduced in rate oreliminated;

(4) Responses to others should be mutuallysupportive and cooperative, and indicatebenefits from shared efforts while providingfreedom to respond as an individual;

79

(5) Responses will provide learners with largenumbers of opportunities for immediate per-ceived success;

(6) Major goals and sub-goals will extend into futuretime and provide delayed goal achievement in arhythm of increasing time distance between goals.

The Classroom:

(1) The learners will interact with each other to theextent possible.

t2) The response rate for each learner shouldultimately reach fifty recorded responses perhour.

(3) There will be no punishment or imposed pressurefrom adults.

(4) The classroom will have two levels of structure,both in terms of physical arrangement andtask peh.formances:

(a) r fairly high degree of structure forsocial interaction and language learning;and,

(b.) No apparent structure for expressiveand creative activities

(5) The teacher will be responsible for insuringthat methods are followed. The teacher will bethe planner, organizer and administrator of allclassroom activities. She will control the flowof activity, direct the aides and insure thatthe learning system paradigm is followed. The

80

-78-

teacher will decide whether the children havereached the skill level required to move on toadditional learning goals. She will not belimited to the materials contained in thesystem and will be encouraged to improvisesupplementary materials which are consistentwith the learning system.

(6) The system will be dominated by learningrather than teaching. The interaction betweenlearners will be the dominant process in-volved in producing content and social interactionskills.

The Learning Materials:

(1) Since learners will be free to move and talkamong themselves, the learning materials mustprovide sufficient emotional satisfactionsto engage in available alternative stimulation.

(2) All learning materials will be designed tointerlock, to be cumulative and mutually inclusive.Each new informational element will be learnedby itself and in combination (where applicable)with all previously learned materials; newmaterials will tend to subsume and be supportedby previous learning. (Note apparent incon-sistency with Bijou's fifth assumption above).

(3) All learning materials will be functional.If the learning element is the term "eye,"the learner must know that it is the "seeing"part of the body, that there are two eyes, andthat they are spaced above the nose, etc. Thefunctional use of terms will provide learnerswith language as a tool for achieving satis-factions and attaining goals. Language willbe perceived as a method of obtaining satis-factions which are both more precise and less

81

- -79-

energy demanding than aggressive, avoidance,obstructive and other alternative behaviors.

(4) All learning materials will be so designed thatlearners can obtain a 100% criterion of mastery.Checks of learner skill level will occurfrequently and take place only after skillshave been learned. The checks will functionas proof of mastery as perceived by the learnerand not as a teacher's evaluation of thelearner's performance level.

The Learners:

(1) The system will be designed to permitchildren to learn and act independently andfreely. Learners will be minimally subjectto adults for their progress, achievementrelated satisfactions, and opportunities tolearn.

(2) The settings at which children work will bedesigned to facilitate performance and socialinteraction. Equipment will be built to makeit easier and more natural to work as partnersin a mutual learning task than to act in-dividually.

(3) The classroom will be laid out so that progressivemastery of the learning materials is translatedinto progress in physical space; the classroom,in essence, is an upwardly mobile social systemwhere progress is perceived as upward socialmobility in a society in which learners viewthemselves as members.

8 2

Specifying the System

Converting system criteria into proper system specifi-cations is a complex, elaborate and time consuming task, evenfor highly effective systems teams. The kinds of productswhich are produced in attempting to meet the criteria areillustrated by Figures 4, 5, and 6. These represent an ex-tremely limited but representative fraction of the type ofmaterials developed to detail the requirements for theprototype based on the criteria.

Figure 4 presents one version of the structured-un-structured classroom learning space in specification form.It shows five learning modules each of which seat six children.The modules are split into three sections so that the childrencan interact in pairs. The module sequence provides the upwardmobility in the classroom. Moving from one module to thenext depends on mastery of a given set of learning materialsto a 100% criterion by children working in pairs. Childrensupposedly perceive the successive sequence of movement inlearning to perform effectively in the system.

We realized that there were two major risks involved inthis plan: (1) perhaps children would not be able to learn touse the learning materials working in pairs largely independentof the teacher; and, (2) perhaps children would fail to per-ceive that the classroom pattern, their skills in masteringmaterials and their movement from module to module were in atriple linkage. As the learning system hinges on independentlearning by partner pairs and perceived upward social mobilitylinked to mastery, the decision to go ahead with the classroomdesign was crucial to the success or failure of the system.

Figure 4 shows two additional modules (demi-modules)designed for four children each, hugging the left and rightwalls. The teacher occupies a central position. The LifeSimulator area, shown at the bottom, provides room for a largevariety of self-initiated, expressive activities whereas themodular section provides a rather systematic learner flow with

83

-81-

1....A SYSTEMS APPROACH T'O PRE-SCHOOL EDUCATIONI

Simulator

Room

Free Time

Room

Art Room

"PERMISSION TO REPRODUCE THIS

COPYRIONTED MATERIAL HAS SEEN CRA

CY..

Demi- module

Module 4

Teacher Position

TO En:;; AND OnicifIZATIONS CFERATI:131:1.113:73 kDECEV:iiiTE Wri TEE U.S. NFIGE OF

FCTIVER F..:PRD:JOCTION OUTSIDE

UAL' ST,3"..in hEQUIRES'PENVIISSION OFOWNER."

Module 2

Demi- module

Learner Cubbies

Module I

Initial. Design for A Structured- Unstructured.

Micro -Social Classroom

84

One Way

Glass

Observation

Roo m

FIGURE IV

Q 1970 MYRON WOOLMAN

-82-

children (though free not to perform work) learning and pro-gressing at their own rate, from module to module. On theupper left are cubbies designed to hold the learning materialsof individual children, since the modules are used by twoclasses of learners each day. (For practical reasons, thesemodules were later laid out in a more linear pattern).

Language learning was basic to the learning systemspecifications, as noted above under "Language Objectives."The language materials were organized into seven language poolsor micro-languages. These were: (1) Partnering, (2) Formsand Colors, (3) Body Parts, (4) Food, (5) Household Objectsand Events, (6) Nature, (7) The Classroom and the Communityand (8) Other Lands. (See Figure 5, p. 241.)

The initial language, partnering, requires the childrento learn the signals or ideographs which permits them to worktogether as dyadic pairs while learning the materials. Thelanguage areas are mutually inclusive; as one moves out, eachsuccessive micro-language includes the language previouslylearned.

The initial area to be learned is "Common Forms" (lines,straight, curved, vertical and horizontal lines). These arecombined into squares, rectangles, triangles, circles, etc.,and are used as the basis for initial activities. All termsare learned functionally and are related to the variety ofactivities which provide opportunities to use newly acquiredlanguage terms in the course of constructions, cut-outs, anddrawings.

For each micro-language there is a story and at leastone workbook. The storybook is designed to: (1) dramaticallypresent the terms in each micro-language; (2) provide a meansfor children who speak in a dialect or in Spanish to hear thegoal words correctly sounded in context; and, (3) provide areliable method to obtain responses from each learner toproperly sound each of the goal words.

The language of color is included to provide languageterms which would be perceived as functional in construction

-83-

and coloring activities and to increase individuality andexpression in these activities. Designation and use of color,in effect, provides a simple micro-language which dramatizesactivities quickly. Color also provides basic terms forselection, description, games, songs and communication. Notethat the language pools move from the self (perception) out-wards to "Other Lands" to provide learning terms such asigloo, pagoda and polar bear. Figure 5 represents the micro-languages as a series of concentric circles. Each concentriccircle is inclusive, continuing and extending the languagebase previously mastered.

The terms in each micro-language focus on functionalvalue rather than frequency counts. The intent was to organizea body of basic terms that offer a wide range of concrete,descriptive terms, classificational terms, and concepts sothat the pre-school setting could be generalized and expandedto the home and community. Each new term must be learned atfive levels. At the first level, the term is a "goal word"in a story. The remaining four levels of the term are to bework-books synchronized to the story line. Two learners worktogether as partners through the four levels. Signals(ideographs) in the workbooks are intended to permit thechildren to work together with almost no teacher intervention.

The organization of the five levels for learning eachnew word (goal-word) is presented in Figure 6, The PedagogicalLattice for Functional Meaning. The five learning levels aretermed: Contextual Association, Discrimination, Identification,Functional Association, and Functional Meaning. At theContextual Association Level, the learner is merely requiredto sound the term correctly based on the teacher's pronunciationand then to point to the illustration in the teacher's story-book (Cell B-3). At the second level (Discrimination), thelearner is in his own workbook and working with a partnerand his responses are limited to selecting and matching formsand illustrations which are associated with the meaning of aterm he had voiced correctly at the lower level. (See p. 243.)

8f;

-84-

A SYSTEMS APPROACH T-0 PRE-' SCHOOL EDUCATION

EXPANDING LANGUAGE CAPABILITY

OTHER LANDS

COMMUNITY,

PooNATURE:piants.

vA0USEHOLD

FOOD

,tioDY PARTS

\-10̀

FORMS; $4,.40

.P.

CO0

COD9,.PARTNERING

FIGURE V

SUCCESSIVE SPEECH POOLS

87

-.85-

At the next higher level (identification) the learnerjoins the two lower levels (Discrimination and ContextualAssociation). He makes the appropriate word sound which hehas learned from Contextual Association upon seeing the appro-priate illustration or form he has learned from the Discrimi-nation Level. When he can give the correct term on seeingthe form and select the correct form on hearing the term, hehas identified the object. In addition, using ideographicsignals he verbally identifies all terms previously learnedon seeing a variety of forms in his work-book.

All terms are to be learned functionally, e.g., the"eye" is used for "seeing", the elbow "bends." TheFunctional Association Level requires that all learners under-stand why a term they know at the Identification Level isimportant to them. Functional values are to be shown in specialillustrations and presented by the teacher or aide to set upthe three-way linkage (illustration-ideLtification-andfunction).

At the Functional Meaning Level, two partners are toestablish that each knows the correct name for the objectsillustrated and can tell what the object did or its use value.

At the Functional Meaning Level (top right hand cell),the learner is required to designate the word which would bemost appropriate for accomplishing something within an illu-stration designed for that purpose. He must analyze thesituation, determine what happens and designate the appro-priate term and the function being performed in the actioncontext. These tasks increase in complexity to develop itemsof increasing difficulty. Thus, for any one goal word at theContextual Association Level there is at least one functionalrelated term to be learned at the fourth and fifth levels ofmeaning.

Life Simulator Space activities are planned to be cross-coded to the terms learned so that the functions can beactualized in a much more realistic and concrete setting. The

88

cot

ASY

STE

MS A

PPRO

AC

H T

O PR

E- SC

HO

OL

ED

UC

AT

ION

PE

DA

GO

GIC

AL LA

TT

ICE

FO

R

VE

RB

AL

FU

NC

TIO

NA

L ME

AN

ING

FIGU

RE

VI

IITE

RM

ISS

IOR

TO

RE

PR

O=

TH

ISD

MY

RIC

HT

ER

MA

TE

RIA

L HA

S B

EE

R IM

AM

fiknertV

TO

ER

IC A

ND

OR

GA

NIZ

AT

ION

S O

PE

RA

TIR

OU

ND

ER

AG

RE

EM

EN

TS

WIT

H T

HE

U.S

. OF

FIC

E R

IF

ED

UC

AT

ION

. FU

RT

HE

R R

EP

RO

DU

CT

ION

OE

MT

HE

ER

IC S

YS

TE

M R

EQ

UIR

ES

nuns= O

fT

HE

CO

PY

RIG

HT

OW

NE

R."

Word

I

Picture of

Word

CO

NT

EX

TU

AL

AS

SO

CIA

TIO

N:

Understanding

Word In M

ean-

ingful Context

and Associoting

Word w

ithP

icture

Story C

ontext

23

0 1969 MY

RO

N W

OO

LMA

N

DIS

CR

IMIN

A-

TIO

N:

Choosing

Picture of

Word from

thoseof P

reviouslyLearned W

ords

IDE

NT

IFIC

A-

TIO

N: V

erbal-izing of W

ordw

hen Word In

Tw

o or Three

Dim

ensionsIs S

een

Picture of

Word and

Previously

Learned Words

Choice

45

FU

NC

TIO

NA

LA

SS

OC

IAT

ION

:

Verbalizing

the Word for

the Function

of theO

bject

Verbalization

of Word

Function of

Word

Perform

ance

of Function

of Word by

Learner

FU

NC

TIO

NA

LM

EA

NIN

G:

Perform

ance ofF

unction ofW

ord In Rela-

tion to Functions

of Previously

Learned Words

Existing

Choice F

ield

9

-87-

Life Simulator Room activities are designed to be simple, quicklycompleted, and to result in a tangible and concrete construction.It was anticipated that these linked Simulator Space activitieswould give the children an opportunity to work individually orin groups, see the new terms and their functions in a realisticgoal achieving context, and provide the basis for adding newterms in a play activity setting. For example, after learninglinear forms and colors, an activity would involve theirassembly into a figure such as a robot, house, car, etc. Thesenew terms, cluster words, function as descriptors for a totalintegrating effort which describes a total activity cross-linked to terms learned in the work-books.

At a theoretical level, it is essential that the termsin each micro-language be presented in orderly and mutuallyinclusive sequences. It is also important that the sequencemove from concrete Isensorily available) terms, to classifi-cational terms and ultimately to conceptual (multiple class)terms. The stories (for Contextual Association) and the learnerworkbooks (for the other four levels), are planned and con-trolled through the use of lattices. A Lattice for Body Partshas also been designed as a method of organizing body termsinto a sequence such that each successive level would be sub-sumptive and mutually inclusive. In the course of actualimplementation, lattices may be fractionated into finer detailsor modified as required; however, their use was important inproviding a pre-planned structure laying out the sequence inwhich all terms and functions would be introduced independentof the particular team member ?erforming the task.

-88-

THE OPERATIONAL STAGE

The problems involved in translating the specificationsinto an actualized operational system may be briefly described,though the actual process is both complex and time consuming.The basic problem is to assemble a staff which will: (1) learnand conform to criteria and specifications; (2) function asa team capable of synchronizing their individual efforts intoa cohesive whole; (3) work with imagination, ingenuity andunderstanding under the various r.istraints imposed; and,(4) have the flexibility to shift. from task to task to conformwith the specification requirements as they change for variouscomponents.

The Vineland classroom in operational use presents avaried scene. Some children may be at the Contextual Associa-tion Level, others may be working independently at their modules,and still others are in The Life Simulator Space.

The partnering illteractions are linked to workbooks andcontribute to the system in several ways to:

(1) Increase the rate of learner response sharply;

(2) Free the teachers and aides Lc focus on problemsof learning, classroom flow, evaluation andbehavior control;

(3) Provide social interactional skills basedon mutual success in joint tasks;

(4) Provide a basis for movement from module tomodule;

(5) Set up opportunities to verbalize andutilize language terms;

(6) Provide the basis for shifting social roles basedon level of mestery;

(7) Become the basis for rotating all learners witheach other to eliminate cliques and/or isolates;

91

(8) Generate the language base which can then beused functionally in activities in The LifeSimulator Space and the home and community.

In one sense, the entire learning system hinged on thepartnering concept, since it was based on the belief thatlearning efficiency, socialization value, and behavioral controloutcomes would be substantially reduced without the independentlearner interaction involved in partnering. It soon becameapparent that the children (even the youngest) had no diffi-culty in learning this new "game." Within a month, the majorityof the children were responding to the partnering ideographicsymbols with awareness of their meaning, though misinterpre-tations did occur frequently. By the end of the third month,the 'lasses routinely used the method as planned with littlediscernible difficulty.

Teacher Training: No effort was made during the Micro-Social training of teachers to persuade the staff assigned bythe school system that this new method would be superior toexisting methods. Teachers were initially skeptical and distant.However, despite doubts about the partnering concept, the hugenumber of terms, the lack of punishment or discipline, thepeculiar look of the modules, the goldfish feeling conveyedby the observation rooms and many more reservations, theteachers learned the methods and techniques, studied and under-stood (though often they wouldn't accept) the rationale forthe system. In brief, though the teachers lacked faith in theapproach, each worked and learned how to perform the teachingjob in the classroom. In three weeks of training (includingone week in the classroom) they developed a good verbal commandof everything we thought they should know; however, they wereunsure of themselves in the classroom context.

From our experience staff training should cover thefollowing:

92

(1) Optimal methods of teacher activity in theclassroom;

(2) Techniques for using special materials andequipment; social interaction and behaviorcontrol techniques, etc.

(3) A specific text or guide for covering allaspects of the method involved in the classroomcontext.

(4) Opportunities to observe and critique teachersin action in the classroom and relate these tothe teacher training materials.

(5) Free discussion sessions where the value orlack of value of the new system can be vigorouslyand candidly discussed.

(6) Opportunities to work in the classroom in theroles of parent-aide, aide, and teacher tounderstand different role requirements.

(7) Structured discussions covering vital aspectsof the new method.

(8) Case studies designed to involve "bestchoices" under conditions where no patsystem solution is possible.

The in-service training of staff should be intensive atthe outset. Steps must be taken to minimize the forming ofinitial habits which violate learning system procedures. Thisintensive training and in-service period should not exceedfour months, after which the number of meetings should besharply reduced. Where teachers are unable to accommodateto the learning system approach, they should be free to return totheir previous positions. Conversely, if teachers continue to

93

use incorrect methods after initial training, intensive in-service training should be maintained.

Teacher Selection: To the extent possible, teachersshould not be selected on the basis of special criteria. Apre-school system, with an effective training program and agood in-service training system, should be able to train anduse a high proportion of the teachers sent by a public school.The alternative is to compete for some special category ofpersonnel, often in very short supply. A pre-school systemassociated with a public school produces minimal problemsif it accepts those teachers (certified or not) who areassigned to it. Under such an arrangment, however, it isimportant that teachers be able to leave the pre-school withoutprejudice on their simple and unsupported request.

Teacher-aides are vital to the success of a pre-schoolsystem and are readily recruited from the community. Ifpossible, it is often best to locate teacher-aides whosechildren are students in the Learning System. If a substantialnumber of children do not speak English, bilingual teacher-aides can be invaluable, not only in the classroom, but alsoto interpret at meetings, to translate take-home circulars, speakon the radio, and to explain the purposes and methods to thenon-English-speaking community directly in the homes. Aidetraining should overlap completely with teacher trainingwherever possible to develop understanding and the range ofskills required. Aides should participate in at least one in-service training session per week and, if funding permits,in all sessions. In the classroom, of course, the teacher hasthe responsibility to insure the fidelity of the system; well-trained aides who understand the system simplify the teacher'sproblems considerably.

In addition to the wide variety of specific, pre-schoolsystems related materials that must be developed, there mustbe administrative support accompanied by various forms and pro-cedures. Where an innovation system is linked to other organi-zations, administration is complicated by relationships with

94

the school system and funding agency. Staff morale can beaffected by skepticism and uncertainty, particularly if teachersare recruited from the on-going school system in the samecommunity. Problems of added workload, failure of schooladministrators to recognize special circumstances, communi-cation gaps, and slights -- real or imagined from non-systemprofessional staff, parents and community leaders, etc. --produce added complications which require considerable energyand occasional finesse in maintaining good teacher and aidemorale. The Superintendent of the host system, Board ofEducation members, higher level school functionaries andteachers in the host system should be oriented to the purposes,objectives, methods, materials and equipment whenever possibleto minimize misunderstanding and conjecture.

Teacher involvement in both the prototype learningsystem and the normal chores of the classroom is vitally re-lated to the fidelity of the system itself. Teachers areimportant in helping to solve the following kinds of problemswhich occur in making an innovative system function effi-ciently:

(1) Differences between the operational systemand the system as initially designed whichoccurs through failures in training, mis-interpretations, availability of materials,supervisory and/or staff errors, per-severation of old teaching habits, etc.These simple discrepancies must be locatedand corrected before they become crystallizedinto the operation.

(2) Situations where the operational systemis conforming to the specifications butthere are problems in terms of learnerperformance, e.g. children become confusedor disinterested.

-93-

(3) The materials and methods function asplanned, but there are typographical errors,misplaced pages, missing symbols, confusingillustrations, unforseen secondary connotationsin language terms, etc., which involve locatingerrors within the minutiae of the learningsystem.

(4) The system is functioning as planned, butimaginative and resourceful staff familiar withthe method, objectives and approach makesuggestions which are consistent with themethod and enrich it, based on their classroomexperience and their intimate moment-to-momentawareness.

The Operational Pre-Tests: Objective arms-lengthevaluation of key base-line data was considered a necessityby the funding agency and the Vineland program. The fundingagency was to make the arrangements using psychologists oftheir choice. The initial pre-tests of the children were(1) Wechsler Intelligence Scale for Children (WISC); (2) ThePeabody Language Test; and (3) The Goodenough Draw a Man. Dueto problems in coordination, testing began after the center'sopening. It was considered more feasible to have the childrenenrolled during pretesting even though the pre-test scoreswould be confounded by whatever learning took place duringthat period. This discussion was also viewed as a conser-vative testing posture, as presumably these base-line scoreswould be increased to some unknown extent during this initialperiod. It was assumed that score differences on the follow-up tests would tend to be reduced in comparison to those thatwould have been obtained on the pre-tests before entry.Nevertheless, the means were quite low (median WISC was 71.8for N = 117). Also after testing only a fraction of the Controlgroup, their testing was discontinued abruptly and withoutnotification as the importance of this procedure was not fully

96

appreciated outside of the project itself. This slip-up inobtaining the total requested Control data emphasizes theabsolute necessity (discussed at some length in a latersection) for erecting an external, specialized and highlysophisticated evaluation superstructure which operates to com-pare all pre-school learning systems using measures which lackat least some of the defects of those in current use.

The first interim testsof project operation have beensample of the children and arescore differences appear to bepated.

covering the first six monthsgiven to a restricted randomnow being analyzed. Mediansomewhat greater than antici-

The Middle-Class Children: The first genuine evidenceof the staff's positive reaction to the program was not in theform of ratings, discussions, or recommendations. Rather,about three months after operation began, teachers, aides andtheir respective husbands forcefully and formally requestedthat their own children not be excluded on the basis of theirhigher income levels. Since it was felt that these childrenwould increase the skill range and be catalytic in a varietyof ways, their inclusion was authorized.

97

EVALUATING AND MEASURING LEARNING SYSTEMS

The only necessary condition for an effective systemis that it result in the kind of output behaviors required.If there were a choice between theoretically based methods,one of which delivers a narrow band of skills whereas the otherdelivers a broad spectrum to the same level, the obvious choicewould be to favor the system which generated the greatestrange of output learner skills. Further if an eclectic, non-theoretical system delivered a stronger output than one basedon theory, the eclectic system should prevail. Systems forlearners must provide measurable results.

The claims of eclectics, developmentalists and theoreticiansmust be tested against the hard realities of learner capabili-ties. The efficiency of systems is not resolvable in journals,seminars, colloquia or regulations. Evaluation of learningsystem effectiveness must be independent of partisan claims.The consequences of a learning system, properly evaluated,should offer no refuge in theory, authority, or rhetoric.

Learning systems are useful to the degree that they:(1) predictably increase the rate of learning, retention, andappropriate knowledges and skills; (2) deal with the problemof total human beings (rather than micro-behaviors attainedin limited time frames); (3) increase skills in making choices,judgments, and decisions related to life itself; (4) reduceimpulsive and aggressive behaviors where these restrict learn-ing opportunities; (5) provide skills in interacting effectivelywith peers and adults; (6) substantively increase the propor-tion of children who satisfy elementary school standards, and,(7) reduce the number of children who are unable to respondwithin the limits of institutional regulations and publiclaws.

Measuring and Evaluating Learning Systems: There arethree basic considerations involved in the measurement and

98

evaluation of a Learning System:

(1) The System Fidelity Evaluations;

These evaluations concern the degree to whichthe system prototype operates in accordancewith pre-existing specifications. Systemfidelity evaluation is directly concernedwith locating areas of system discrepancy.These evaluations are directed to insuringthat the materials, equipment, methods and pro-cedures in use are as consistent as possiblewith initial plans and specifications.

(2) The Individual Criterion Measures;

These are measures of learner performance whichdirectly reflect the system objectives. Ifthe system was designed to increase languageskills and social-interaction skills then thebattery of criterion measures should providesuccessive sets of measures from the outsetof the project, over time, that yield a basisfor evaluating the degree to which the specificobjectives of the system were attained.

(3) The Total System Output Measures;

These are measures of the total capacityof the system. These are not limited tothe specific objectives of a particularlearning system but are designed to pro-vide a basis for comparison between learningsystems which may have differing objectives.

The Fidelity Evaluation Approach:

These measures concern the degree of consonance with

99

specifications and are primarily of value to the systemdesigner. The measures form the basis for change to minimizedeviations from operation of the system as initially planned.

Staff as an Information Resource: The teacher andaides have the most intimate and direct contact with the learn-ing system. During the Operational Prototype their classroomexperience can and should be utilized to uncover system dis-crepancies. This requires the following:

(1) The teaching staff must be sensitive andsympathetic to the objectives, methods andtechniques used in the project.

(2) Staff must be able to verbalize the rationalefor inclusion of given methods, materialsand equipment and also possess the verbalskills to pin-point any areas of deficiencyin the actual classroom activity.

(3) Sufficient rapport and confidence in systemstaff to provide objective and candidaccounts of system problems for the record,even under circumstances where staff members'own behaviors are involved.

(4) Maintenance of a log of critical class-room in-cidents which affect system objectives.

(5) Surveillance and detailing of errors, confusingterms and illustrations, etc., in learningmaterials.

(6) Attendance of systems staff meetings, formulatingcritiques and otherwise serving as a resourceto reflect the practical problems of the class-room.

100

-

(7) Making recommendations individually and/orcollectively to improve system functioning con-sistent with the system frame of reference.

(8) Maintaining as standardized a pattern of activityas possible across class-rooms, making certainthat new system modifications are adapted intoall class-rooms in as similar a way as possible.

(9) Completion of required data collection involvingratings of (a) individual children; (b) adequacyof various aspects of the learning system; and(c) personal attitudes.

(10) Observing actual teaching activities and relatingto system methods and procedures as required.

Rate of Coverage Measures: In those learning systemswhere there is an established framework or sequence of learneractivities, some assessment measures must be adopted. Forexample, if there are a substantial number of work-books, thespecification of completion dates for each successive work-book provides an important index of rate of learner motion.Should the rate of work-book coverage measure be used, it isimportant that procedures exist which insure that childrenhave mastery of work-book materials before proceeding forward.Checks of level of mastery should occur frequently (one ortwo per day); rate of motion and mastery of learning materialsmust be tightly linked or the rate measure becomes ameaningless index.

The measures of rate can be broken down and scored interms of pages covered per learner per month or even responsesper learner per hour. The use of the page as a unit ofanalysis presumes that the pages are standard and that theresponse requirements and evaluations within each work-bookare identical for each learner. In the Vineland system pre-viously described, recorded responses are made by each child

101

--99--

on each of the work-book pages. The total number of responsesper work-book is available as are individual attendance records;thus, it can and is possible, using the work-book completiondates, to make means, variances, correlations, etc., based onrate per hour scores against such factors as sex, age, socio-economic level, etc. This writer feels that changes in rateof response to work-book pages (rather than abs)lute rate) willprobably emerge as the most sensitive response rate measure.Analyses should also reflect the initial objectives of the pro-ject; simple plots can determine trends for each learner andthe class as a whole.

Verbal Interaction: The measurement of social-inter-action, if that is a relevant variable in the system, can beperformed readily by setting up forms to determine: (1) thenumber of times and by whom responses are initiated within agiven period of time; and, (2) who is the recipient of theresponse (a) another learner (b) the teacher (c) the teacheraide, or (d) the parent-aide. Carefully collected time samplesof such class-room interaction patterns can generate a broadrange of scores, such as:

(1) Number of responses initiated by learners,teachers, and aides per hour.

(2) Number of responses received by learners,teachers and aides per hour.

(3) The ratio of learner to adult response

(a) initiations

(b) the ratio of learner to adult responsesreceived.

(4) The number and length of "chained" responses.(A response chain is defined as possessing atleast three elements, initiation-receptionand a response back to the initiation (I-R-I).The frequency of chained responses betweenlearners and between learners and adults isreadily obtainable).

102

-100--

This type of measure clarifies who is initiatingresponses and who is rt:.acting to them. In a pre-schoollearning system based on higher learner response rates, ateacher whose response rate is very high (above 45%) isreducing the level of responsivity for all learners in herclass-rocm; the response opportunities for perhaps 25 child-ren are limited to 55% of the available time. If ten obser-vations are taken per week on a rand,Jm basis, social inter-action measures are readily obtained which permit the analysisof ratios of learner initiation to adult initiations; learnerreception to adult receptions; learner initiations to adultreceptions; learner receptions to adult initiations, etc.Th3 number of "chaining" responses among learners andbetween learners and adults also would be of value. Achained response involves comprehension and attention to thelanguage of another, ability to respond in a pertinent andmeaningful way to the information received, incorporationof the new informational element fed back, and the responsein the new enriched context. Growth in the size and numberof clzained responses in a pre-school setting could be a sig-nificant index of growth in social interaction skills.

Verbal-interaction scores can be computed by learnerclass and for all `earners and staff and are useful as a basisfor staff discussions because they provide unambiguous evidence.Though the form should be used anonymously in staff meetings,teachers quickly use the form operaticnally and limit theirresponses where they see themselves as initiating a high pro-portion of responses.

The Critical Social InoWent: Where a learning systemspecifies as an objective the improvement of social skills,a quantitative method for determining the effectiveness ofprocedures related to improvement of social skills should bean organic aspect of system operation.

The measurement method should involve an unambiguousspecification of behaviors defined as anti-social. Thesespecified anti-social behaviors may be modifiable in terms

103

-10123

of experience in operating the prototype system but a clearlisting of such behaviors is necessary to define thisparticular system objective and to minimize differences inteacher reactions to similar behaviors. When such behaviorsoccur, the in,stem procedures (1) should involve whateversystem method is employed to deal with it and (2) require arecord covering the date, child, and perhaps the anti-socialbehavioral element itself. If a card or form is completedfor each learner activity specified within the listed domain,then it becomes possible to establish the initial anti-socialbase-line for each child, each class, and the system as awhole. Curves can be drawn to establish whether learneranti-social behaviors are decreasing, maintaining their ownlevel, or increasing. Anti-social behavioral rates can beobtained and these, in turn, can be related to workbookresponse rates, teacher ratings, absenteeism, etc. Differencescores indicating drop-off in anti-social behaviors could,in turn, be related to IQ shift, Peabody scores or otherindices using correlation and analysis of variance techniques.

Absenteeism: Absenteeism may be viewed as a socialindex; however, absentee scores should be separated intosuccessive and intermittent patterns. Successive scoresindicating the presence of a childhood illness should bediscArded. The greater the social involvement in the learn-ing system, the higher the threshold for a non-attendanceresponse. Absenteeism should be viewed as an index of totalfamily involvement. Absentee scores, totalled on a quarterlybasis, can be used as successive scores over time for indi-vidual children, classes, and the system as a whole. Differ -&nce in scores by individual learners could be analyzedagainst difference scores in number of anti-social incidents.Workbook progress rate, language and performance scores alsocould be analyzed against absentee scores.

10

,102-

The Home and Community as Diagnostic Resources: In-

stitutional systems readily produce conforming behaviors inmany instances. The learning system for pre-school childrenmust be judged, not only on how well the children performwhen their behavioral expectations are known, but also, andperhaps more importantly on how well they use what they havelearned when an response options are open.

The effects of a learning system can only be partiallyassessed in the classroom. The development of measures forobtaining evidence of system effectiveness involves decisionsas to the most valuable kind of diagnostic information. Inthis analysis, major emphasis should be given to determiningwhether the prototype learning system has sufficient force togeneralize beyond its own limits or whether it merely pro-duces conforming behaviors within its own walls. Evaluationsof performance and the impact of the system on the learnershould extend beyond the institutional capsule itself, to thefamily and community setting which are vital dimensions ofanalysis of the value of the system for the learner.

With this purpose in mind, major emphasis should begiven to evaluating the objectives of the learning system asreflected in the child's behavior in the home and community.Dimensions which merit analysis are: Differences in thesystem educated child as compared with siblings given alter-nate methods of training with reference to: (a)vocabulary;(b) level of aggression; (c) impulsivity; (d) avoidance andwithdrawal behaviors; (e) social interaction as evidencedby cooperativeness, empathy, sharing of toys, books, candy,and other goal objects; (f) relationship with parents;(g) general position of peer group as peacemaker, leader,follower, etc.; and, (h) types of activities characteristicof behavior in and out of home, etc.

The data gathering methods for external generalizationmeasures could consist of: (1) structured interviews;

10o

-103 - -

(2) rating scales; (3) ranking methods; (4) critical incidents;(5) behavioral sampling; (6) in-depth observation, etc. Anestimate of the generalization vector beyond the learningsystem is essential if we are to determine whether the proto-type is producing mere conformity or assisting learners tocope with the problem solving, adjustment and social-interactionprocesses of life itself.

There has been considerable controversy over the useof IQ test and other measures of language and performancewhich favor children from specific socioeconomic, culturaland ethnic backgrounds. In this writer's view, the suggestionthat any test of verbal or performance skill is somehow tappinggenetic potential involves the erection of a Mendelian meta-physics on a psychometric altar. Tests which involve makingchoices as to the meanings of terms and following directionswill, if scored correctly, give an advantage in score tochildren who possess the greatest reservoir of meaningfulterms in the language of the test and who comprehend and arewilling to follow directions precisely. Capability to makecertain classes of important meaningful response choices is,in fact, being measured. Genetics are irrelevant to theitem structure, the responses made, the scoring, and thecomputation of the 1Q score itself. The gratuitous assumptionthat the gene pool is being tapped should be relegated to thesame forum which now serves those who debated the number ofangels who could dance on the head of a pin.

Binet and Simon's early statement on the value of theIQ still appears reasonable to this writer. It is a directand immediate measure which produces a score on the "presentmoment" which reflects comprehension, judgment, reasoningand invention.

...In order to recognize the inferior states ofintelligence we believe that three differentmethods should be employed. We have arrivedat this synthetic view only after many yearsof research, but we are now certain thateach of these methods renders some service.These methods are:

100

1. The medical method, which aims toappreciate the anatomical, physiological,and pathological signs of inferiorintelligence.

2. The pedagogical method, which aims tojudge of the intelligqnce accordingto the sum of acquired knowledge.

3. The psychological method, which makesdirect observations and measurements ofthe degree of intelligence.

From what has gone before it is easy to seethe value of each of these methods. Themedical method is indirect because it con-jectures the mental from the physical. Thepedagogical method is more direct; but thepsychological is the most direct of all be-cause it aims to measure the state of theintelligence as it is at the present moment.It does this by experiments which obligethe subject to make an effort which shows hiscapability in the way of comprehension, judgment,reasoning, and invention. (Binet and Simon,1961; p. 883).

Do the standard IQ tests have a function for the measure-ment of pre-schoolers or are they useless because the IQnotion has been a device to suggest the necessary ascendancyof a meritocracy, destined by Mendelian imperatives, toleadership over those who couldn't come up with as manycorrect responses. In this writer's view, these scales havefunctional value for producing useful evidence as to the valueof a learning system. These reasons are as valid now aswhen Binet presented the first tests in 1911 (incidentallywithout suggesting a genetic rationale).

107

--105,

The IQ tests present a standard and limited universeof items which predict with some success how well childrencan perform within certain institutional settings. Binet'sinstitutionalized language pattern, standard reading matter,and standard recitation methods involved a centralized schoolsystem with a national curriculum, standardized teacher be-haviors, recitation methods and student response require-ments. Although our educational system lacks the orthodoxyand centralization that characterized France at the beginningof this century, language and behavioral expectancies, never-theless, are sufficiently homogenous to make the IQ scoreextremely useful. Irrespective of the magical genetic claims,the IQ scores are good predictors of school success. There-fore, systems involved in getting children ready to attendelementary school will find them useful. If the mean IQlevel of pre-school children in a learning system is initiallybelow 75 but has risen to 100 or more after two years, thegain is important. One can predict that a substantiallygreater percentage of these children will survive the firstgrade public school environment because they have thelanguage base, ability to attend to directions given by anadult, and the willingness to respond to an adult withoutresistance.

The Evaluation of the Individual Performer

In evaluating any human pe,..formance, there are a fewbasic conditions which must exist for the measuring instru-ments to have value. Human measures must be related to somedimension which increases in magnitude as a function of thehuman performance dimension being tested. One method ofattaining this dimensionality is to set up a pot-pourri ofitems, obtain inter-correlations, and assign dimensions onthe basis of factor loadings. Another is to construct thetest items along a given dimension such as vocabulary,

108

reading comprehension, arithmetic, grammar or other achievementarea. Such tests may be useful as indicators of individualperformance levels under certain conditions, for example:

(1)

(2)

(3)

Instances where items are not taught inthe classroom overtly or covertlyor are not known by the teacher so thatscores on the test reflect a true sample ofthe relative performance level of thelearners.

Instances where tests are equally fair to allrespondents and maintain dimensional in-tegrity; for example, a test of reading shouldnot require imformation known to some but notall respondents since this confounds readingscores with an information base. Readerswho lack the information will obtain scoreswhich cannot be differentiated from (a) non-readers with the information; and (b) non-readers lacking the information.

Instances where guessing scores and zeroresponse scores have the same value. Anexaminee who does not respond on an initialtext and guesses on a second version of thesame scale should not be assured of a gainon the second measure as his knowledge orskill level was not shifted.

(4) Tests in which the scales do not generatescore differences as a consequence ofage, sex, color, dialect, or any factor otherthan the dimension of measure.

(5) Settings where test directions are notused to shape 'core distributions, sinceskills in accenting directions are notrelevant to the dimension being measured

10D

(6)

(7)

107-

unless the scale is a measure of skills infollowing directions.

Situations where notation is made that failureto score well on a test is due to (a) emotionalfactors (b) illness and/or (c) resistanceto the tester.

Tests in which item correctness is absoluteand not based on the relative frequency ofhigh scorers. The latter criterion begs thequestion of the standard by shifting itback to the learners themselves behind astatistical mask. By the same token, itemcorrectness should not reflect beliefs,customs, language, or accepted truths atparticular socioeconomic levels or be relatedto geographic usage.

Any factor which clouds the value of the score as anindicator of the learner's true performance level should beaccounted for. Though psychological scales are generallybarred from cardinality, every performance score should bereflected by an ordinal number which is indicative of thecapability of the examinee along the dimension being tested.Non-performance should be labelled as such and not exist ashidden error within a statistical analysis.

The preceding criteria for measurement of individuallearner performance would screen out widely used achievementtests. These tests are not suited to the measurement ofindividuals but are useful tools (under some conditions) formaking comparisons of different groups of learners. However,even here, these shot-gun tests have dubious value for evaluat-ing learning systems.

These venerable tests initially had a different functionthan they now serve, namely to provide estimates of thelevel of mastery of content areas within standard curricula.

110

However, these tests have become so popular that much of theclass-work is now directed to teaching informational elementswhich tend to crop up within the tests. In many cases, thesetests do not sample learner skill levels across curriculabut function as measures of teacher and student sagacity andcommitment to locat.d.ng and/or predicting the kind of infor-mation which will be covered.

Under these circumstances, though the initial forces forthese tests have maintained their momentum, the tests havelost their raison d'etre. They no longer measure what theywere designed to measure. More important, they have reducedmuch class-room teaching to informational splinters, testrelated fragments which sharply limit student opportunitiesto integrate and conceptualize relationships within curricularareas.

Or stated more directly, much teaching does not in-volve the total curriculum, but test preparation. Althoughthe educational measurement problem cannot be dealt with atlength here, it should be evident that serious efforts toimprove our educational system through the use of learningsystems must also focus on the development of strongermeasurement tools than those now in popular use. Until then,decisions on the relative merits of alternative learningsystems (including pre-schools) cannot be made with any greatconfidence.

In this analysis, learning system evaluation hingeson five basic factors:

(1) Learner Gain Measures: The difference intotal capabilities of learners asmeasured at the point of entry into thelearning system (base-line scores) withtheir total capabilities on leaving thesystem (out-put scores).

iii

(2) Generalizability Measures: The degree to whichthe learning system succeeded in providingadaptive and socially useful skills when trans-ferred to the home and community.

(3) Value for the Next Life Stage: The degree towhich graduates from one school level (e.g. pre-school) are able to adapt and perform effectivelyin the next life stage. From the entry level,follow-up studies are necessary to determine thevalue of pre-school learning systems as related toimpact on elementary, adolescent and adult perform-ance in school settings and in the community.

(4) Cost of pre-schooling per learner in relationshipto gains in_performance: A system which pro-duces substantial gains in four areas is fourtimes more efficient per dollar expended than asystem which produces like gains in only one area.

(5) Gains attributable to the learning system in termsof savings in tax dollars: These could berelated to drop-off in special education pro-grams, vandalism, drop-out, welfare, medicalcosts, incarceration, parole costs, etc., asa function of more adequate pre-elementary schoolreadiness. Using extended longitudinal studies,pay back for pre-school systems could be relatedto tax yield differences across pre-school learnersas compared with non-pre-school trained childrenand differences across children given differenttypes of pre-elementary school readiness training.

(6) Analyses across pre-school systems to comparetheir relative level of efficiency in assumingelementary school readiness: Such analysesshould be made within the first month in firstgrade.

112

-7110-

The major danger in requiring that learning systems belocked into theoretical frameworks is that easily manipulatedmicro-theories would have an immense advantage which wouldsharply limit the range, depth and scope of class-room learn-ing systems. Theoretical tunnels could reduce pre-schoolsystems to magnified laboratory settings focused on verynarrow (but measurable) responses. The essence of the pre-school classroom must be the organic growth of the totalhuman being.

It is imperative that theories be developed whichcan deal with the problem of an ever-evolving learner.Eclecticism is preferable to the erection of theoreticalbarriers which limit the rate, depth, and extent of learnerdevelopment. Learning systems must be gauged by learneroutput performance: however, where usable within a pre-school, and across the range of skill development requiredthe system should be tested and articulated.

Measuring Learning Process: The measurement of in-dividual performance within a learning system is viewed, inthis analysis, as one means of determining the efficiency ofthe system in process. One can estimate system efficiencythrough measurements of changes in learner performance overtime as related to costs in personnel, materials, equipmentand facilities. The measures of the learning system as aprocess of obtaining changes in performance over time in-volve such dimensions as:

(1) Rate, depth and extent of learning languageskills and content information.

(2) Control over impulse and aggression, generallife adjustment in the school, home andcommunity setting.

(3) Social value of the learning system interms of adjustment, adaptability, awareness,

113

and inter,:action with others of differingages, sexes, colors, socioeconomic status,language, etc.

(4) Ability to work independently (or as a teammember or team leader) to determine ob-jectives and to work effectively and solveintermediate problems to achieve objectives.

(5) Ability to solve problems to attain goals innon-standard, idiosyncratic, and creative ways.

(6) Ability to move into new situations and adapt,perform, and meet the standards required.

(7) Ethical and moral understanding involvingthe rights of others and the obligation ofthe self to others.

The Measurement and Comparison

Total Learning Systems: The measurement of learningsystems by IQ measures, available achievement tests, ratingscales, the use of authority, theoretical conformity, orconformance with accepted dogmas offers little basis for de-cision among alternative approaches. Achievement tests havebeen devised (and used by the writer) which purport tomeasure at the pre-school level. The fact that widespreadachievement testing produces item teaching class-rooms isnow an accepted fact of American education. Learning systemsat the pre-school level should involve the use of measureswhich are categorized as follows:

(1) Impervious to Inflation Criterion: Are unableto be taught in the class-room as a means ofinflating scores.

114

-

(2) The Synthetic Criterion: Function to developskills in integrating and organizing informationfor purpr;ses of problem solving and goal achieve-ment.

(3) The Total Child Criterions Go far beyond thelimitations of content information and relateto total personality Integration and 1,i,feadjustment.

(4) The Future Value Criterion: Function as learn-ing systems in terms of overall impact in le.terlife performance.

Learning sys_ems for the young will be, in this analysis,the most important determinant of the quality of the futurnational manpower base. If this assumption holds to anysignificant extent, the measurement of pre-school learningsystems should be so organized as to provide national policymakers with absolute and incontrovertible evidence of therelative merits of the alternatives which now exist aad whichwill emerge in the future. Decisions so vital to the societyas a whole merit the development of measures which functionto insulate decision makers from exaggerated claims, class-room gimmickry, and well organized marketing strategies. Theessence of the problem is to develop a battery of pre-schoolmeasurements which: (1) function to measure system out-putbased on learner performance; and, (2) are used within anorganizational network which insures the validity of theevidence.

One requirement in the area of systems measurement isto move beyond standard tests to evaluate individuals and pre-tend that combined individual scores permit evaluations ofsystems, not collections of individuals measured on a limitedsubset of discrete items within a given curricular area.

The following analysis assumes that (1) the use of pre-printed standard achievement tests erode and limit teacYinqand learning in the class-room and (2) offers substantial

115

advantages and great temptation to systems builders to insurethat test item-related materials are completely covered withinthe learning systems. This could well mean that a weak learn-ing system which heavily focused on test items would appearto have higher output scores than a substantially strongersystem which covered the total curriculum plus various liferelated skills.

The suggested method of testing and evaluation of pre-school children which follows is general and is assumed tocover content, social-interaction, general performance, pro-blem solving, generalization to home and community, follow-uptests, etc. The presentation here is necessarily abbreviatedbut hopefully the coverage will be sufficient to indicate thegeneral point of view to those concerned with the systemmeasurement problem.

For purposes of illustration we will consider a contentarea within the system containing: (1) 15 major concepts;(2) 35 sub-concepts which are linked in various ways to the15 major concepts; and (3) 200 relationships involvingdefinitions, inter-relationships, and functions.

For purposes of illustration: (1) these three patternsform the total universe of information in the given contentarea; and f2) the inter-relationships to be covered amongthe three patterns is limited and consists of 1000 informationalelements. These 3.000 informational elements which are bydefinit+cm non-redundant are then used to generate 1000 items.Of course, this inforairltional base could be used to producetwo, three or more, items per element, but no such item re-dundancy will be assumed here for purposes of simplicity.

The next step would be to: (1) store the 1000 itemsinto a computer with a print-out capability; and (2) programthe computer to randomly sample across items and print outtest:4i consisting of, say, 100 items each. Thus each testwould contain a sample of ten percent of the total itempopulation. But as the items would be drawn at random, each

116

-114-

learner would have his own unique ten percent sample of thetotal universe of content information covered in the curriculum.If the computer were to print out 1000 tests, there wouldbe 100,000 items. Each of the 1000 learning elements in thesystem would occur 100 times, by chance, across all testsgenerated.

Without considering problems of test administration --which are not relevant at this point -- we may move quickly tothe data collection process. For purposes of simplicity, wewill test 1000 learners across six pre-school learning systems(A,B,C,D,E and F). On entering into a given learning systemeach learner would be given a base-line test covering thecontent area. Each learner would be tested on his uniquesample of 100 items, and would respond to his unique combinationof items. However, all learners taken together would coverthe total universe of items in the curriculum and each itemwould occur equally often for each system being tested.

Though no single individual would be required to respondto more than a small fraction of the curriculum, the totalcurriculum would be covered 100 times over by the learners ineach of the set systems. The effectiveness of each learningsystem would involve the degree to which the universe ofinformation (or behaviors) had been learned across the sampleof learners covering the population of information in thecurriculum. Rather than testing individuals on repeated sub-samples of the curriculum, samples of individuals would beused to determine degree of mastery of the total populationof information in the curriculum. Statistical methods couldbe used to evaluate item difficulty, reliability, internalvalidity, etc.

However, because the emphasis in this approach wouldbe on the ability of the system to produce a total learningproduct, there would be room for many types of analysis whichare not now part of the psychometric armamentarium.

In this context, means, variances, difficulty indices,score-ranges, distribution-shapes, median mean relationships,

117

etc., become diagnostic tools for improvement of the learningsystem. For example, large variances indicate that a systemdistributes scores at the extremes and delivers inadequatesupport for "weak" learners; high item difficulty indiceswould pinpoint areas of weak coverage in the learning and/orevaluation aspects of the system, etc.

Table 3 on the following page presents a hypotheticaloutcome of the scoring of the 1000 learners in the six systemspreviously mentioned. These hypothetical tests are viewed asoccurring over a period of 24 months. During this period eachlearner was tested six times in each system. Each learner, ofeach system presumably received a different test for a totalof 36,000 different tests. If we assume that the tests wouldbe reasonably reliable (above .90, Spearman Brown), there canbe no argument about validity as items are drawn directlyout of the curriculum itself.

An examination of Table 3 will be used to indicate thewriter's view of how this universal sampling method could beemployed as a decision making tool:

(1) System E started with learners with aminimal content base (Mean 18.3) which almostdoubled during the course of the two yearperiod (gain = 18.2). Of special note wasthe accelerating rate of increase over time.

(2) System D which started at the highest levelin terms of content managed to obtain a sub-stantial increase: (12.7) points with therate of gain maintained throughout.

(3) System F, which had the next to the lowestrank in base line scores appeared to offervery high acceleration at the outset throughInterim Test #3 but reversed slightly (fallingdown from 39.8 to 38.9) on Interim Test #4 andgained slightly for the final test. Its outputdifference of +16.9 was second in rank order

118

-k16-

Table 3

Mean Performance on Six Hypothetical Learning Systems

A, B, C, D, E and F

Total Curricular Universe Test

BaseN Line

TestInterim

#1

Interim#2

Interim#3

SystemInterim Post Output

#4 #5 Differences

System A 1000 37.2 37.8 41.3 42.5 42.6 42.7 + 5.5

System B 1000 28.3 31.4 34.3 36.7 38.3 39.2 + 10.9

System C 1000 41.5 42.3 43.2 42.8 43.6 44.7 + 3.2

System D 1000 44.6 46.3 49.2 53.2 55.4 57.3 + 12.7

System E 1000 18.3 18.6 18.9 23.4 29.8 36.5 + 18.2

System F 1000 23.2 28.5 34.6 39.8 38.9 40.1 + 16.9

Total 6000 32.18 34.15 36.91 39.73 41.43 43.41Means ofSystems

Table 3; A hypothetical array of means from six learning systems basedon six total universe curricular 24 month period.

119

-117-

for the systems, though clearly the system asdesigned lacks momentum for the full two yearsin this content area.

(4) System A, which was slow at the start, managed togenerate a reasonable gain between Interim Test#1 and Interim Test #2 but remained quite staticthereafter.

(5) System B, starting with learners in the mid-range of base-line scores, managed to makequite respectable gains (+10.9) but was de-celerating by the time of post-testing.

(6) System C showed very little gain (+3.2) overthe 24 month period, beginning with a ratherknowledgeable group of children. Thoughsome gain occurred at each test point,there is little evidence that the system hassubstantial value in this content area.

Table 3 offers several decision points:

(1) Check available data for Systems A and B todetermine whether minimal end-product yieldsoccur across all curricular components.

(2) Determine whether staffing, staff training,equipment, materials and facilities handicapactual testing of the method.

(3) If methods are functioning as planned, thensystem personnel and leadership should beinformed that their systems are functioning ina marginal range and will be discontinued unlesssystem outputs move above, say, a 10.0 range overthe twenty four month test period.

If we assume that the data in this curricular areareflect the entire range of test scores across all test areasthen:

120

(1) Systems E and F should be tested with childrenin the mid and high range to determine whetherthey can maintain their acceleration ratewith higher base-line children.

(2) System D should be tested on lower base-linechildren to determine whether the method wouldhave a greater rate of acceleration on theassumption that learning growth rates decelerateas a function of preceding learning growth.

Table 3 could of course be subjected to an analysis ofvariance using difference scores as covariance (based on thebase-line sources). Correlational matrices are readily developedacross successive tests. Without belaboring the point, standardstatistical tests to test the null hypothesis (single tailedand two tailed tests) appear to offer no greater difficultyusing these measures than currently used measures. The problemis to develop researchable learning system models which en-courage sophisticated decision making based on careful re-search.

The tests given to measure a learning system should ofcourse be (1) specific to the stated objectives of the systemand (2) reflect the growth of the child across all dimensions(such as those listed in the 12 points below).

The total system test concept, as viewed by this writer,does not include the type of factual punctate elements nowused in the typical multiple choice test. Essentially testitems are viewed as demanding problem solving skills. Thebrief test descriptions below are suggestive of the kind oftest considered appropriate to evaluate total learning systems.

The completion and testing of measures for total learn-ing systems will involve considerable effort which will bejustified if education can finally.begin to make rational de-cisions as to the best method or system necessary to attainlearner behaviors over an extended period of time.

121

--119-

For purposes of clarifying the differences betweencurrent measures and measures of the total learner, a fewalternative dimensions will be listed which could be developedinto measures to assess total system effectiveness:.

(1) A test of comprehension skills covering language,gesture and expression.

(2) A test' of understanding of the functionalrelationships of objects taken singlyand incombination.

(3) A test of ability to analyze and organizeinformation where all necessary informationand/or equipment is supplied within the itemitself; however, one important conditionis that no information from the home, communityor class-room can form the basis for anadequate response.

(4) A test of creativity in which all items in thetest have a conventional means of achieving asolution to a problem covering a sequence ofsteps. Creative scores occur when (a) newsolutions are produced using an alternatesequence or (b) a solution is obtained in fewersteps.

(5) A test of social-interaction skills which involvestwo or more persons to solve any given problem.Solutions occur best under conditions wherechildren communicate and are aware of eachother's activities and work to cue each other'sresponses to attain their mutual objective.

(6) Test of control over aggression and impulsein which problems are designed so that it isalways possible for one child to obtain a goalor prize for himself. However, if he workswith one or more children the same prize goes

-120-

to each participant. The scores are based onthe choices made (social vs. antisocial)and the kinds of behavior patterns exhibited.

(7) Assessment of task involvement where tasks aredeveloped which require large numbers ofresponses over extended time periods to reachsome pre-established goal. The learner isscored on the basis of the number of responsesmade to each goal-task and the shape of theresponse curve over time. This could be animportant index of motivation.

(8) Measurement of stimulus generalization todetermine the degree of distortion made by thelearner in responding to very similar butdiffering stimuli. This could be a usefulindex of anxiety and uncertainty.

(9) Learning effectiveness'measures to cover a varietyof learning problems. that may develop but havea standard difficulty level. LearningEffectiveness Score would be based on the rateat which material was learned. This scale,though admittedly posing some problems, wouldbe of major importance if a valid and reliablemethod, were to be developed. Positiveshifts in learning system would requirecorrelations with gains in petformance withinthe system to establish validity.

(10) Learning system tests which measure understandingof complex conceptual. relationships (while pro-viding objectivity). These are not beyond rea-sonable limits of attainment.

(11) Past-Future Tests which would be concernedwith whether the learner prefers the safety of

1

-121-

past situations which are totally known orfuture situations which involve uncertainty,risk and opportunity.

(12) The Verbal Chain Test which involves the numberof times the child. can maintain verbal inter-actions with another child or adult on agiven subject where All information is knownand distractions are present at intervals.

The development of suitable performance tests toevaluate total systems would be an interesting, even challeng-ing exercise. It would of course be expensive, time_ consumingand result in some new problems statistically. However, fromthis writer's perspective, none of the available tests (in-cluding his own) are suited to the problems involved in comparingtotal learning systems that have been designed to improvethe development of our young people for tommorrow. It shouldbe emphasized that a new method would not exclude use of allstandard measurement techniques although ingenuity andflexibility would be required in devising others. Without acarefully organized measurement system, such as suggested,learning system decisions can be based on gross but unreliablescore differences, powerful and authoritative voices ofsupport, conventionality and theoretical conformism. Theproblem is not to obtain system endorsement nor even an arti-culate and enthusiastic system constituency. The problem re-duces to the use of objective measurement to determine, asrealistically and objectively as possible, the systems whichare most effective in terms of the development of the child asa whole, intellectually, socially and emotionally.

124

-122,

AN EDUCATIONAL RESEARCH SUPERSTRUCTURE

The problems of educational research are not solubleby the rese&rcher regardless of the adequacy of his theory orhis operational system. Education is a pervasive social in-stitution which affects every level of society. Educationalresearch, therefore, deals with the very marrow of our socialsystem, and the destiny and the opportunity of all our children.The educational researcher cannot anticipate the neat andordered simplicities of the laboratory. He is dealing withother people's children and public money and the success orfailure of his method is of interest to appointed and electedpublic officials of every level. The educational researcheris a man whose work is under public surveillance in a farmore intimate way than the physicist, psychologist, or engineer.

Experimental research is one of the methods men use toselect the best among many possible alternatives. More ex-plicitly, research is a complex decision-making tool which menmust employ when logic, authority, divination, and traditionalbeliefs have proven powerless.

There are many euphemisms used to convey the notionthat experimental research is now being accomplished ineducation-. "Evaluations by experts" -- a standard techniqueused to judge the worth of a project -- ultimately assuresthat traditional views are crystallized and supported by currentresearch."Educational innovation" is often interpreted as avariation of an accepted method or belief which has managed togather a large number of enthusiastic adherents.

The student of physics and the student of education,at the undergraduate and graduate level, are each studyingmethods, procedures, and measurement instruments but thedifferences in their experiences are profound. The physicsstudent is learning material which is consistent with theunderlying theoretical system which has a demonstratedefficiency in controlling and predicting events, whereas the

125

-123,

education student has no such rigorous and systematic con-ceptual spine upon which to rely. Consequently, the graduatephysicist can move boldly into designing devices (e a0,0,cyclotrons, radar, spectrometers, etc.) with confidence borne ofexperience that the energy and costs involved will be justified.In education, the converse holds; in the absence of a cohesiveand established theoretical base, educational methods aretentative and uncertain; movement into novelty involves eareerrisks and the educational researcher often retreats intominor modifications of conventional and admittedly inefficientmethods to avoid stepping off into a conceptual abyss un-supported by a network of theory.

For present purposes, an educational- project exists, when,and only when, there is a comparison among alternative treat-ments to determine the method of greatest value in terms ofstudent learning. As research in physics involves the investi-gation of the physical properties within space and time,educational research is limited to the investigation of methodsto improve learning capabilities within well defined learningenvironments.

Educaticnal research involves the interaction of a com-plex chain of interacting variables within a confined spacesuch as the classroom, language lab, lecture hall, seminaror pre-school. The physical space controls the limits ofmobility and visibility and audibility of stimuli; the classroomequipment (desks, chairs, teaching machines, etc.) controlspupil distance, social interaction potential and verbalizationopportunities, etc.; floor covering differences affect theimpact of noise level, as well as visual and tactile stimu-lation; and, walls affect sound and may be blank or be coveredby a barrage of confusing, attention gathering stimuli.Generally, the time consumed by teacher activities largelycontrols the amount of time available for responses by alllearners. Further, the emotional state projected by theteacher is reflected in the emotional reaction by all learners,and perhaps also in satisfaction, apprehension and/cr frustration.

12G

-124-

Sensitivity to differences in language and behavior as afunction of economic level, race or ethnic origin, also canprofoundly affect learning rates. In addition to the abovesample of factors which may affect the learner's ability tocomprehend printed materials and verbal statements, there arefactors such as family preparation, learner health, nutrition,self image, community attitudes, etc., which may add to thevariables which interact in the classroom context.

In short, the classroom or learning environment is acomplex network of interacting variables which produces aproduct or resultant change in student performance abilities.The degree to which these interacting variables are interrelatedto predict outcomes is the measure of the quality of a learningsystem. The classroom then is viewed as a learning system,which has as its single and unvarying objective, the improve-ment of the learner's range of abilities. Restated, EducationalResearch is the systematic effort to improve the classroom asa learning system for the purpose of increasing student ad'ustment,knowledge and performance. This definition of educationalresearch is not consistent with much of the wide range of workwhich is now termed (and supported) under the rubric of edu-cational research.

In the present analysis, educational research differsfrom psychological research in that it is directed towardimproving the rate and quality of human learning in a socialsituation (the classroom), a phenomenon which involves theuse of independent but interacting variables. In psychologicalresearch in learning, however, variables are isolated to thedegree possible (so as not to be confounded) and the operationusually involves one isolated variable on one subject, humanor infra-human, at a time. Psychological research certainlycan contribute important data to educational research; how-ever, it is important that the limits of each research areabe clearly designated, that a sharp and unremitting focusbe wintained to minimize confusion across these behavioralareas which differ in theory, technique, method and objectiveswhere learning processes are involved.

127

-125-

Research Dynamics:

ITI.111,^41.1

A simplified diagrammatic illustration of researchdynamics is presented as Figure 7; The Lattice of ResearchRelationships. In the lower left hand cell (Cell A-l), theterm Problems in Nature is used to indicate the area of investi-gation. In the area of bio-chemistry, Nature would, of course,relate to the requirements and interactions involved inliving organisms, in physics, the boundaries would cover matterfrom the subatomic to the galactic, including pertinent timeand space relationships. In the present context, this wouldbe the total classroom ecology, ranging from a student andteacher on opposite ends of a log, to complex electronic class-rooms and/or life simulation classroom models.

The next cells (A-2, A-3) cover the body of assumptionsused to control and predict Nature. For over two centuries,man predicted natural events in the heavens and on earth, largelythrough the use of imaginative use of Newton's three Laws ofMotion, which proved instrumental in many ways. However,Einstein's modifications of Newton's analyses predicted somewhatdifferent outcomes of similar phenomena. These two overlappingbut partially contradictory networks of assumed relationshipsor theories predicted different outcomes under certain de-finable conditions which we need not elaborate upon here.Viewing these relationships in light of Figure 7, however, onecan see that when predictions (A-3) are made from two con-sistent theories (B-2), then a test situation is constructed(A-4) which ideally is designed as a fulcrum which must tipin the direction of one theory or the other (D-4). As anintegral part of the experimental design, a measurement system(A-5) must be used which measures with sufficient sensitivity,reliability, and relevancy to insure that any true differenceswill be accurately and repeatedly obtained. Once the measure-ments are obtained (D-5), they are subjected to a statisticalanalysis (A-6) to simplify and compress the mass of data andalso to give some estimate of the probability that one theo-retical position has merit over the other (F-6). However, as

128

GF

1

SYST

EM

S APPR

OA

CH

TO

PRE

- SCH

OO

L E

DU

CA

TIO

N

The L

attice of Research R

elationships

FIGU

RE

VII

0

7

- -

"PE

RM

ISS

ION

TO

RE

PR

OIN

CE

TN

R

lOP

YR

ISH

TB

MA

TE

RIA

L MA

I SE

EN

GR

AN

TE

D

s(N ....&

teemT

O E

RIC

AN

D O

RG

AN

IZA

TIO

NS

OP

ER

AT

ING

UN

DE

R A

GR

EE

ME

NT

S W

ITH

TH

E U

.S. O

FF

ICE

OF

ED

UC

AT

ION

. FU

RT

HE

R R

EF

RO

OD

C01313E

TH

E E

RIC

SY

ST

EM

RE

QU

IRE

S P

ER

MI:i.L-it

OF

TH

E C

OP

YR

IGH

T O

WN

ER

."

Experiment

Hypotheses

Theory

Problems

In Nature

Imposition

of Order

Predictions

Fulcrum

Test

12

4

Theoretical

Modification

A

0

Quantified

Results

]Measurement

Sequential Relationships

Partial

Proof or

Disproof

Statistical

Analyses

6

Reanalysis

7

01970 M

YR

ON

WO

OL

MA

N

all theories are in the process of evolution, even the pre-diction of the correct theory would still be in error to someextent, and of course the weaker theory which had lost theFulcrum Test would be in still greater error. Based on thenew data from the experiment, theoreticians from both com-peting positions, would evolve new frameworks for theirassumptions (B-2). Again points are located where theoreticaldifferences are sharp and differing hypotheses can be testedagainst each other; a tiny sliver of Nature is carefully testedand measured to determine which theory holds best for the giventest, and once the data is in, new theoretical images are con-structed. Each pattern of modification increases the levelof theoretical precision and increases the scope, value andprecision of the theory. As the research scientists increasetheir ability to predict and control Nature, so does the society.Ever evolving precision in theory translates to an evergreater control over Nature.

Nature in educational research is, of course, the learningcontext or the classroom as a learning system. The funda-mental research problem is to develop a set of assumptions (A-2)which forms a theoretical network (B-2) with sufficient power todevelop a set of predictions or hypotheses (C-3) which can betested under controlled conditions (A-4) and under circumstanceswhich permit the acceptance or rejection of the hypotheses whichwe term The Experimental Test Situation (D-4).

Theory A: A learning context will be most efficientif based on the following assumptions:

(1) The environment is enriched andprovides an ever changing field ofstimulation using selected learningmaterials.

(2) Learners are always given totalfreedom of choice.

130

(3) A skilled teacher operates as thebasic source of information and isresponsible to inspire and motivatelearners.

(4) Periodic testing generates studentmotivation and involvement.

Learners self-actualize themselves in environmentswhich provide the following basic developmental ingredients.

Theory B might offer an alternative or overlappingset of assumptions, such as:

(1) Learning systems involve controlover choices.

(2) The number of choices offered to thelearner will be small at the outsetand increase gradually based ondemonstrated mastery.

(3) Motivation to perform will occur bestwhen the classroom operates as asocial system which is neither auto-cratic nor anarchic, but where statusgain and social roles are outcomes ofdemonstrated proficiency.

(4) Learning is optimized under conditionsof maximal responsivity by the learnersand minimal intrusion by adults.

(5) Learners will develop highly socializedbehavior under conditions where nolearner can succeed at the expense offailure by another.

131

(6) Language learning in the classroom con-text will occur optimally to the extentthat terms improve the ability of thelearner to function on word-related tasks.

(7) The classroom itself should be designedas an ecological space which providesstructured, orderly and predictablerelationships and also provides un-structured space for free and expressivebehaviors.

(8) The teacher's role should be essentiallythat of an orchestrator of materials,movement and pattern of interaction oflearners and evaluator.

(9) Impulsive, aggressive, and other disorderlyand/or disruptive behaviors are best hand-led by providing opportunities for thesebehaviors to occur under conditions wherethey are not followed by either reward orpunishment, but only by an effectivelyneutral time gap.

(10) The quality of a learning system isvalidated by learner performance inthe next life stage.

These two theoretical systems may then be testedexperimentally by carefully planning the classroom's organi-zation and learning materials to conform as closely to thetheoretical models as possible. Children of equivalentinitial performance capability (by appropriate measures)would be retested to determine rate of gain. Also, measureswould be made of ability to adjust and perform in the nextlife stage and thereafter. The two theoretical models shouldbe studied closely to determine points where predicabledifferences can be developed into hypotheses.

132

It is immediately apparent that Theory B predicts ahigher learning rate of skills, greater social skill develop-ment, higher motivation to perform, and greater control overimpulsive and aggressive behaviors, as well as improved self-expressive capabilities. Hypotheses would then be developed,based on the theories predicting the outcome given thecorrectness of Theory A, and an alternative set of predictionsfor Theory B. Within this framework, one might test thefollowing, as possible hypotheses:

(1) The Language Learning Hypothesis:

Treatment B learners will have greaterlanguage skills as measured by testsof verbal comprehension, speech, read-ing, and writing than Treatment Alearners.

(2) The Social Interaction Hypothesis:

Treatment B learners will inter-relatewith each other and with adults betterthan Treatment A learners as measuredby teacher and parent ratings, andsociometric tests.

(3) The Behavioral Control Hypothesis:

Treatment B learners will havegreater control over impulsive andaggressive behavior in free and un-repressed situations than Treatment Alearners, as measured by behaviorsamples in free play situations.

(4) The Motivational Hypothesis:

Treatment B learners will tend to re-quire less adult control, initiate a

133

= 131.-

greater proportion of responses, andrespond more interactively with eachother, as measured be behavioral samples,than will Treatment A learners.

(5) Post-System Performance:

Learners given Treatment B in a pre-school setting will perform moreadequately in the elementary schoolsetting, including first grade, thanlearners given Treatment A, as measuredby grades, classroom performance, testsof achievement, attendance and dis-ciplinary reports.

A system test, such as the one sketched out above,may be possible once our educational research posture shiftsto evaluating learner performance outcomes under carefullyorganized conditions over extended periods of time. The con-sequences of such a comparison, under appropriate test con-ditions, could have a massive impact, not only on pre-schoollearning systems, but education as a whole.

The Re licabilit Problem:

If a pre-school learning system were to be supportedby research findings over a number of years, it would stillrequire three additional basic characteristics to be ofpractical value as an instrument for improving educationalquality at a national level. These characteristics are: (1)the range of utility; (2) standard; and, (3) exportability.By range of utility, I mean that it should be usable by asmany different kinds of learners as possible; it should spandifferent ages, socioeconomic levels, urban-rural boundaries,ethnic and racial differences, and, of course, it shouldwork for both sexes. By standard, I mean that the system should

.134

-132-

require little or no modification when it moves from the re-search and field setting to general adoption by school systemsin locales with responsibilities for children differing infactors such as socioeconomic level, background, speech, andsystems of values. By exportable, I mean that the experimentalsystem should be sufficiently well articulated so that each ofthe system elements (equipment, learning materials, teachertraining, system administration, evaluation, etc.) should existin a form which permits rapid digestion and utilization by anew school system. Three months after acceptance of the newmethod, effective classroom operation (including teacher train-ing) should indicate whether the system is exportable.

However, a proven research project which is public,standard and exportable will still fail, unless it is palatableto all persons concerned with the system in any way. Thus,new research systems, as part of the export process, requireorientation preparation and public awareness in each newcommunity setting, or their very novelty can become a basisfor community controversy and jeopardize their acceptance.

Obtaining Evidence to Education Decisions:

Educational research differs in quality and kind fromthe physical sciences as well as the behavioral and socialscience activities which can take place in neatly controlledsettings, such as the psychological laboratory, or in de-scriptive field studies characteristic of anthropology andsociology. Education takes place in the life cockpit of itsactivities, directly affects the family and is pervasive forthe community. The educational research realities not onlydiffer in method, but in the process of performing the re-search; the new method must be acceptable to the community andto the families involved, or it will fail to survive the timeframe required to determine its value.

In addition, if we consider the problems faced by ahypothetical educational researcher who has evolved a theoryfrom which a prototype educational system could be built,

135

we may be able to point up some of the key issues for testingthe efficacy of a system which differs radically from existingsystems. Such an undertaking would require:

(1) Money for the materials, equipment,teacher training, etc.

(2) A setting providing conditions where atrue test could be made.

(3) Training and continual upgrading ofteachers and staff to insure conformityto the method.

(4) Freedom from requirements to meet local,state and federal administrative regulationswhich violate system requirements.

(5) Capability to involve parents, professional andcivic organizations as required to insureunderstanding and community support for theproject.

(6) Pre-test and interim test capability toestablish the rate and quality of system per-formance.

(7) Capability to employ experts for advice, sug-gestions and coordinated effort where requiredto solve problems.

(8) Powerful and immediate public support for re-search freedom when and if a project cameunder attack from the press, or any otherpowerful force.

These eight points, though helpful to the individualresearcher, would not provide the basis for determining whichof a number of alternative systems was more efficient in

136

=134-

terms of learner performance, cost, effectiveness, long-rangeimpact at the elementary school, high school and adultlevels. Historically, research in education, because of itscommunity nature, high costs, crucial importance to the in-dividual, and vital manpower implications for the society asa whole, is vulnerable to social pressures. It is no accidentthat Socrates, an educational innovator of considerable merit,was viewed by many as undermining the social system because heattempted to break through certain conventional beliefs witha technology which required the learner to make choicesamong carefully framed alternatives.

The development of effective learning systems for ourchildren will require, in this writer's view, an apparatuswhich is specifically deigned to continuously improve avail-able systems, and a means to facilitate the dissolution ofinadequate systems, while insuring support for those whichimprove learner performance capabilities.

The crystallized attitudes engendered within thesepowerfully emplaced educational fortresses have been neatlypinpointed by Silberman in'a discussion of individually pro-grammed instruction involving the view of James Becker,Director of Research for Better Schools (RBS) in Philadelphia.Silberman states:

. .'IPI's strategy assumes that the ends areknown, that somebody does in fact prescribethe goals, and that a student is supposed toreach those carefully defined goals. Indeed, itis a highly structured kind of strategy.'Becker concedes that 'some concepts of individuali-zation of instruction at some times in some subjectareas might include ends that the student himselfdefines.' For himself, however, the concept isclear: 'I must reject the idea that maximumfreedom is a necessary condition for somethincalled individualization of instruction.'

137

-135-

Few educators, however, would pose the alter-natives in such stark terms: total prescriptionof goals by others, or total freedom for thestudent to pursue his own goals. Indeed, Becker'sdefinition of the alternatives represents a com-plete misunderstanding of the nature of educationand a perversion of its goals. (Silberman; 1970,p. 199; emphasis added).

The essence of research is that it must be tested onthe basis of objective evidence. There is, and should be, noadvocacy or fossilized belief. When a research organizationcan assert imperatives behind a screen of research objectivity,the term research is reduced to a travesty. As long aseducational research organizations can pre-determine a "bestsolution" without conclusive evidence, and then proceed touse their resources for expansion, defense and advocacy,research in education is reduced to a travesty. In a latersection of this paper, it is suggested that funding of re-search programs, evaluation, and report writing be organizedas separate functions. Further, it is held that such evalua-tions be directed to comparing alternative learning methodsto obtain evidence as to (1) which produces the greatestlearning gain and (2) how various approaches can be improved.This is important as the educational research labs may rejectmethods of possible value in pursuit of some idee fixe.

138

-136 -

THE EDUCATIONAL RESEARCH TESTING SYSTEM

The Decision Making Apparatus

The research of an individual in education has extreme-ly limited value, even when its implications are substantial.This is partly due to the number of competing forces, complexand overlapping agencies, and restrictions of funds, time andpersonnel. However, even when substantial resources havebeen made available for Community Action Programs, RegionalEducational Laboratories, and Manpower Programs, the resultshave failed to strike sparks. Adherents of various fashion-able approaches advocate panaceas.

Ungraded schools, total elimination of literary,exotic computerized hardware, and Educational Televisionare currently vying for adherents. The fact is that class-room learning whether graded or ungraded, using books ormachines, requires that certain contingencies occur. Machinesare neither good nor bad qua machines. The issue is one ofdeciding whether one approach or another produces greateryield at less cost in time, personnel and dollars than another.

Education must move beyond advocacy and test methodsin its own crucible--the classroom. Evidence as to learningsystem effectiveness should not force organizations to: (1)

develop research programs;.(2) design the measuring instru-ments for the evaluation; (3) obtain the evaluation data; and(4) write research reports which support or reject theirprograms.

These are public organizations whose functions areprimarily defensive. They are focused on: (1) funding forthe following year; (2) justification of their line ofeffort; (3) avoidance of conflict with key local, state,and federal figures whose support is required to maintainand/or expand the level of effort; (4) minimizing reper-cussions in the media which might damage their image; and,

139

(5) attempting to maximize and justify their organizations'efforts through formal and informal contacts, and effectiveuse of media.

The foregoing statements are not criticisms of educa-tional laboratories, but reflect the fact that their functionsare not primarily concerned with educational research, butwith the maintenance and expansion of their organizationsunder extremely difficult conditions. Research projectsare, by definition, unpredictable and in the context of thelaboratory structure, this can be very threatening. In lieuof research, the labs set up broad contacts with schoolsystems, hold conferences, evaluate and consult with schoolsystems on community problems, help schools fund and installfashionable approaches involving the use of programmed mater-ials, and hardware. The fact is that educational research,as used in this paper, is too controversial for the labsto handle.

The small sampling of defensive maneuvers merelyindicates that the structure of the mission at the educa-tional research labs forces certain patterns of defense.The comments are illustrative and are not to be construedas a criticism of any particular agency which spends fundsin initiating evaluation and reporting its own work. Rather,it is intended to stress the fact that any agency whichacts to initiate, put into action, and evaluate, is takingover three tasks which are similar to those which thisgovernment initially divided among the legislative, execu-tive and judicial branches.

By this time the dynamics of decision making ineducational research should be clear. There should beno system protagonists who evaluate and recommend for oragainst their own work. Further, those who take theresponsibility to fund learning systems saould have noresponsibilities for evaluation, field testing or laterreport writing. Those who initiate and design systemsshould be free to evaluate their systems in the interest ofoptimizing their functional value in terms of learner per-formance, and to write research reports, perfect their

140

-

theoretical position, methods, materials, etc; however,internal research reports should not be the basis for majorpolicy decisions. Rather, decision making should involvecomparisons across real (rather than written systems) andinvolve an organization limited to this particular respon-sibility.

In the area of-pre-school education, it is suggestedthat there be a carefully articulated structure specificallydesigned to designate those learning systems which producethe greatest gains for children at every level. Also, thisstructure would ban all recommendations on hard 'ta andwould report directly to the highest official in the agency.The first element (Pre-School Program Acquisition and Support,termed PPAS) in this decision :ructure would be two functions:

(1) Support, fund, and foster large scale andrealistic research efforts over extendedtime periods;

(2) Insure that research projects once undertakenwould not fall victim to administrative cross-currents and shifts in political viewpointsbut continue as planned.

PPM should, in short, act only to provide the stabilitynecessary to explore the adequacy of a particular theoreticalapproach when it is put into practice in a reasonable testsetting. A research support staff would, on request, ironout and prevent difficulties which could jeopardize theproject and/or consume excessive amounts of professionaltime, such as outside assaults on the system. They would re-present the system in such cases and thus, provide thenecessary time to obtain the data that would be required toaccept or reject the innovative method on the basis of ob-tained data.

PPAS would make available at the request of the researchteams, programmers, statisticians, computer services, andconsultants so that large masses of data could thus be reduced

141

-139-

into manageable statistical statement.

Thus, PPAS is viewed as offering optimal benefits where:

(1) It functions to assist in research planningif and when required.

(2) It helps locate the optimal conditions forthe test of the theoretical position andrelated operational system.

(3) It protects the integrity of the project fromdilution as a consequence of administrativeand/or political intrusions.

(4) It aids project personnel, where requested,to help insure community support by orienta-tion of professionals, community leaders(formal and informal), parents and variouscommunity groups.

(5) It provides resource consultants, where requested,to assist in further articulation of the theo-retical model, data analysis, computer programming,computer availability, and analysis of the statis-tical data obtained.

(6) Using ERIC and other information banks, PPAS pro-vides articles and analyses as requested by pro-ject personnel to assist in theory constructionand report writing, to minimize project myopia.

As is evident, PPAS is viewed here as a device to supportand strengthen project operation. It functions to offset thefundamental vulnerability of educational research projects towhatever chance winds blow at the community, state or even thefederal level. PPAS is viewed as a much needed protagonistand supporter of the educational researcher to assist him insolving problems and obtaining technical support as required.

el 2

----- -

It may be viewed as a home office, ever ready to assist andsupport the men and women in the field with personnel skilledand trained to help keep educational research projects frombeing inundated by forces beyond the limits of project control.PPAS's mission is to attempt to insure, using all availableresources, that a learning system, selected for test, doesin fact get the time and resources required to be tested fairlyand without prejudice.

Evaluating the Learning Systems

PPAS has been viewed as the home base and windbreakerfor researchers in educational systems. It would provide theprestige, power and resources of the federal government andinsure adequate tests of educational systems. However, thevery involvement of PPAS in the funding and support of itsvarious projects is a barrier which effectively compromisesobjective evaluation of its projects. For this aspect of theeducational research effort, a separate and independent apparatusis a necessity or PPAS will be in that ancient federal trap withtwo jaws: (1) rejection of projects it has put money into,or (2) self-congratulation where projects are hailed as success-ful under the suspicion that trap #2 is always superior to #1whatever the data.

The PPAS apparatus would be monitored out of the officeof the Director who would insure that:

(1) PPAS maintains complete compliance with initialregulations;

(2) PPAS is modified to improve capability to selectand support learning systems;

(3) Reports on funded projects are evaluated indepcnd-ent of PPAS and based on objective data using thebest available analytic techniques.

143

-141-

For this purpose, it would be necessary to form a secondorganization such as Learning Systems (ERLS), which wouldalso report to the Director. This organization (ERLS)would be a completely separate monitoring group, locatedin a different city, and would have no contact with PPASother than notification of the funding of the proposal.ERLS would be specifically barred from performing educa-tional research of any kind. It would have no authority,directly or indirectly, (by regulation) to assist or supporteducational research projects in any way and would have norelationship, formal or informal, with PPAS. Rather, itsfunctions would be to:.

(1) Analyze the basic theoretical positionadvanced in the proposal and various pro-ject related papers, and analyze the oper-ational system to determine degree of con-formity with the theoretical structure.

(2) Generate its own data (designing the necessaryevaluation tools where required) to improveits ability to evaluate projects. This wouldbe simplified by the fact that it would havecomplete access to all project data.

Evaluation by one group over another's efforts impliespower to support or to destroy. Also, all evaluations containerror and bias. ERLS will have its own biases and errorsformed out of the mission and the professional histories ofthe men and women who carry it out. The alternative toimprecise evaluation is no evaluation. To insure that ERLSerrors will be correctable and occur within certain pre-speci-fied limits, the following type of guidelines should be rigor-ously written into the regulations, to insure that ERLSoperates to facilitate pre-school learning system developmentand does not use its evaluation authority to destroy projectswhich happen to violate current thinking:

(1) All evaluation instruments, data collection,analysis methods, and reports will be made

144

-142--

public and routinely supplied to the Director,project personnel and any other educationaland research organizations which may requestthem.

(2) All procedures, criteria, scoring methods,weighting techniques, etc., used in a dataanalysis as well as all reports involvingtotal project evaluation, will be made publicand routinely supplied to the Director and tothe experimental project.

(3) ERLS would be responsible for developing andtesting new measures, including total systemmeasurement methods to facilitate realisticmeasurement of pre-school learning systems.It would compare the relative effectiveness-of alternative systems to determine the degreeto which each system was optimizing the child'stotal performance.

A third group, independent of PPAS and ERLS, wouldbe responsible for evaluating and comparing all projectsfunded by PPAS. This agency, termed for convenience asthe Learning Systems. Agency (LSA), would obtain the evalua-tions supplied to the Director by ERLS, through its ownresources, and from the project itself. Its function wouldbe limited to producing evidence for decision. It would:

(1) Evaluate the programs supported by PPAS interms of theory, philosophy, level of articula-tion of methods, limits of objectives, shortand long range implications, etc.

(2) Evaluate the merits of the measuring instrumentsand methods developed by ERLS.

(3) Make statistical analyses across learning systemswith similar objectives to determine the mosteffective system.

145

-143

(4) Write reports to the Director responsible fordetermining which pre-school learning systemsare most effectivc in process and over extendedtime frames.

(5) Make recommendations to the Director for expansionand/or limitation of given learning systems.

(6) Indicate the criteria required for more effectiveevaluations by ERLS.

LSA would have one sharply defined mission; the deter-mination of efficient learning systems based on evidence.The systems covered would be limited to those sponsored byPPAS and evaluated by ERLS. LSA would be barred by regulationfrom evaluating either PPAS or ERLS. Although LSA would haveall available evidence on a large number of systems, thisevidence would be focused on only one objective; to informthe Director, as unambiguously as possible, of which learningsystems were offering the greatest benefits to American pre-school children.

14G

Aliport, G.

Baldwin, A. L.

Bijou, S. W.

Binet, A. &Simon, T.

Bruner, J., andPostman L.

Bruner, J., andGoodman C.

Childs, J. L.

-1444,

BIBLIOGRAPHY

147

Becoming: Basic Considerations fora Psychology of Personality. NewHaven: Yale University Press, 1955,pp. 84-85.

Personality. New York: HoltRinehart & Winston, 1937

Theories of Child Development. NewYork: John Wiley & Sons, Inc., 1967,p. 587'

What psychology has to offer educa-tion - now. Journal of AppliedBehavior Analysis, 1970, No. 3,Vol. 1., pp. 66-68

The development of intelligence inchildren. In Shipley (Ed) Classicsin Psychology New York: Philosoph-ical Library, 1961, p. 883

Emotional selectivity in perceptionand reaction. Journal of Person-ality, 1947, 16, pp. 69-77

Value and need as organizing factorsin perception. Journal of Abnormaland Social Psychology., (1947), 42,pp. 33-44

Education and morals. In Callahan,R. E. (Ed.) An Introduction to Educa-tion in American Society. New York:Alfred A. Knopf, 1960, p. 96

Frank, P. G.,

Fromm, E.

Gagne, R.M. &Fleishman, B.A.

Gagne, R.M.

Gesell, A.

Gewirtz, H.B. andGewirtz, J.L.

Hall, V.R.,Lund, D.,

Jackson, D.

Hall, C. S.& Lindzey, G.

-145-

The Validation of ScientificTheories. Boston: Beacon Press,1956, pp. 13-14

Escape from Freedom. New York:Rinehart, 1941

Psychology and Human Performance.New York: Holt, Rinehart andWinston, 1959, Chapter 7, pp. 90-141

Simulators. In Glaser, R. (Ed.)Training Research and Education.NewYork: John Wiley & Sons, 1965,pp. 225-26.

Child Development. New York:Harper & Brothers, 1949, SectionII, pp. 21-22.

Caretaking settings backgroundevents and behavior differences infour Israeli child-rearing environ-ments: Some preliminary trends.In Fosse B.M. (Ed.) Determinants ofInfant Behavior IV. (Proceedingsof the Fourth Tavistock Study Groupon Mother-Infant Interaction held atthe House of the CIBA Foundation,London, Sept. 1965.) London: Methuen.1969, pp. 246-47.

Effects of teacher attention on studybehavior. Journal of Applied Behav-ioral Analysis, 1968, 1, pp. 1-12

Theories of Personality, (2nd ed.).New York: John Wiley & Sons, 1970.

148

Hilgard, E.R.

Hook, S. (Ed.)

Hovland, C.I.

Hull, C.L.,Hovland, C.I.

Ross, R.T.Hall, M.

Perkins, D.T.and Fitch, F.B.

Huxley, J.

Levitas, G.B.

Madsen, D.H., Jr.,Becker, W.C.,

Thomas, D.R.

149

.7146-

Theories of Personality, (2nd ed.)New York: John Wiley & Sons, 1970.

Theories of Learning, New York:Appleton, 1956

Dimensions of Mind. New York:University Press 1960, pp. 228-29.

Experimental studies in rote-learning theory. III Distributionof practice with varying speeds ofsyllable presentation. Journal ofExperimental Psychology, 1938, 23,pp. 172-90.

Mathematico-deductive Theory ofRote Learning. New Haven: YaleUniversity Press, 1940.

Evolution in action, In Callahan,R. E. (Ed.) Education in AmericanSociety. New York: Alfred E. Knopf,1960, pp. 28-29.

The World of Psychology, Vol. I.New York: George Braziller, 1963,p. 4

Rules, praise, and ignoring:Elements of elementary classroomcontrol. Journal of AppliedBehavior Analysis, 1968, 1, pp. 139-50.

Maslow, A.

Miller, N.E.

Miller, N.E., andDollard, J.

Nietzsche, F.

Parsons, T. andBales, R.F.

Piaget, J.

Plato.

Lessons from the peak experiences.In Farson, R.E. (Ed.) Science andHuman Affairs. Palo Alto, Calif.:Science and Behavior Books, 1965.

Motivation and Personality. NewYork: Harper, 1954, pp. 340-41.

Toward a Psychology of Being(2nd ed.). Princeton: Van Nostrand,1968.

Comments on multiple-process con-ceptions of learning, PsychologicalReview, 1951, 58, pp. 375-81.

Social Learning and Imitation. NewHaven: Yale University Press, 1941

Menschliches, Allzumenschliches.In Nietzsche Werke, 2 vols.(Stuggart, Kroener, 1930), Vol. 1.,pp. 109-264

Family Socialization and Inter-action Process; Glencoe, Illinois:Free Press, 1955.

The Construction of Reality in theChild: New York: Basic Books,1954, pp. 357-38.

Meno. In The Dialogues of Plato,Vol. I, tr. B. Jowett. New York:Random House, 1937, pp. 349-82.

Theaetetus. In The Dialogues ofPlato, Vol. II, tr. B. Jowett.New York: Random House, 1937,pp. 143-220.

Postman, L.

Rogers, C.R.

Rogers, R.

Rousseau, J.

Studies in learning to learn.II Changes in transfer as functionof practice. Journal of VerbalLearning and Verbal Behavior, 1964,3, pp. 437-47.

Client-centered Therapy: ItsCurrent Practice, Implications,and Theory. Boston: HoughtonMifflin, 1951.

Becoming, Coming into existence:The Struggle to Become an Individual.Cleveland: World, 1967.

Emile. In Levitas, G.B., The Worldof Psychology, Vol. 1. New York:George Braziller, 1963, p. 4

Silberman, C.E. Crisis in the Classroom. New York:Random House 1970, p. 199

Skinner, B.F.

Tolman, E.

Woolman, M.

What is the experimental analysisof behavior. Journal of ExperimentalAnalysis of Behavior, 1966, Vol. 9,No. 3, pp. 215-16.

There is more than one kind oflearning. Psychological Review,1949, 56, pp. 144-56.

The Effect of Varying the Numberof Choices in the Identificationof Very Similar Stimuli.Dissertation Abstracts, Vol. 15,No. 7, 1955.

11

Zigler, Edward.

-149--

On-Site Training of Guided MissileOperators, Technical Report 64,Human Resources Research Office,George Washington University,Washington, D. C. 1960

Basic Electricity, a ProgrammedLearning Course, AmericanTelephone and Telegraph Company,New York, 1962

Programming For Conceptual Under-standing, Communications ResearchDepartment, Bell Telephone Labora-tories, Inc. Murray Hill, NewJersey, 1962.

Reading in High Gear, ScienceResearch Associates, 1964

Metatheoretical issues in develop-mental psychology. In Marx,Melvin H., (ed.) Theories ofContemporary Psychology. New York:The Macmillan Company, 1963;pp. 344-345

152