chiang mai university journal of social sciences...

ISSN 1905-906XISSN 1905-906X

CMU. Journal of Social Sciences and Humanities website: http://cmuj.chiangmai.ac.th

Ch

ian

g M

ai U

niv

er

sit

y J

ou

rn

al o

f So

cia

l Sc

ien

ce

s a

nd

Hu

ma

nit

ies

Chiang Mai University JournalResearch Administration Center239 Huay Kaew Road, T. Suthep, A. Muang, Chiang Mai 50200 THAILAND.Tel: 66-53-942676-7 Facsimile: 66-53-943600E-mail : [email protected] [email protected]

CONTENTS

CHIANG MAI UNIVERSITY JOURNAL OF SOCIAL SCIENCES AND HUMANITIES

CHIANG MAI UNIVERSITY JOURNAL OF SOCIAL SCIENCES AND HUMANITIES Volume1 June 2007 Number 1



Vol.1, June 2007, Num

ber 1

Publ

ishe

d by

Chi

ang

Mai

Uni

vers

ity, C

hian

g M

ai,

Thai

land

CONTENTS Page

TRANSITION ECONOMY How Far Are the Transition Countries from the Economic Standards in the European Union? 1 Supawat Rungsuriyawiboon and Alexei Lissitsa

Leadership in Dairy-Cattle Production in Southern Thailand : Dimensions in Socio-Economic Status, Adoption of Recommended Practices, and Communications 21 Kriengsak Pattamarakha, Boonlert Srinoy and Somkiat Saithanoo

Projection of GDP of Manufacturing Industries in Bangladesh: An Approach to Growth Model 35 M. N. A. Bhuiyan, Kazi Saleh Ahmed and Roushan Jahan

The Existence and Rationale of Predatory Pricing Strategy in the Network Industry: An Experimental Approach 49 Yingyot Chiaravutthi

Understanding Housing as an Economic Stimulus: A Regional Computable General Equilibrium (CGE) Model for New York State Housing 69 Rainer vom Hofe

LEARNING DEVELOPMENT

A Study of First-Year College Students’ Metacognitive Awareness of Reading Strategies 93 Xu Xianming

Analysis of Dimensionality Underlying EFL Proficiency of Thai Students 109 Sumalee Sawatdirakpong

Development of an Instructional Set for a Laboratory in Biology Course for First-Year Thai Science Student Teachers 119 Nantarat Puengpang, Vantipa Roadrangka and Bronwen Cowie

Page


A Case Study of a Thai In-service Primary Teacher’s Understanding of the Nature of Science 137 Tepkanya Promkatkeaw, Michael Forret and Judy Moreland

A Review on the Teaching and Learning Resources for the Deaf Community in Malaysia 165 Siew Hock Ow, Salimah Mokhtar and Roziati Zainuddin

COMMUNITY HEALTH

Relationships of Service Quality, Patient Satisfaction and Patient Behavioral Intention of Pharmacy Service of Highly-Active Antiretroviral Therapy in Community Hospitals in Chiang Rai Province, Thailand 177 Wacharin Chaitha, Chupen Viboonsunti, Tanattha Kittisopee,

Unchalee Permsuwan, Ratanaporn Awipan and Penkarn Karnjanarat

Development of Quality Indicators on Important Aspects of Nosocomial Infection Prevention and Control 197 Wilawan Senaratana, Wanchai Moongtui, Wichit Srisuphan,

Nunta Leksawasdi, Poonsap Soparat, Susanha Yimyam,

Jittaporn Chittreecheur, Wilawan Picheansathian

and Wipada Kunaviktikul

CHIANG MAI UNIVERSITY JOURNAL

OF SOCIAL SCIENCES AND HUMANITIES

C

TM

T

A

Editorial Policy hiang Mai University Journal of Social Sciences and Humanities is a journal for publication of original Social Sciences and Humanities research work. Articles must be original reports of research not previously or simultaneously published in any other technical journal. Manuscripts are critically reviewed before they can be published. Chiang Mai University Journal of Social Sciences and Humanities is published one time a year in June. It was first published in June 2007. Articles are contributed by Chiang Mai University faculty members as well as by those from other ai and also foreign institutions.

Aim and Scope his international journal is a publication bringing together state- of-the-art research results in Social Sciences and Humanities and is dedicated to those who strive to create the better world.

Manuscripts anuscript should be submitted to the editor-in-chief. All papers must contain an abstract and key word, and must be written in English. e mode of presentation must be in the form of social sciences papers, i.e. Introduction, Materials and Methods, Results and Discussion, whereas Conclusion is an optional part. A paper should be limited to 30 typed pages in length, and should contain no more than 10 figures. e typed original and three copies of the paper and artwork, should be submitted to the editor-in-chief. Color photography may be considered with a charge to the author. Typing should be double spaced with ample margin on A4 size paper. Submission may also be made by fax or e-mail attachment in MS-Word format.

Reprints en copies of each published paper will be supplied free of charge to the corresponding author. The order of a minimum of 100 reprints can be forwarded to the Editor-in-Chief. Invoice of the extra reprints will be sent to the author(s).

Inquiries ll inquiries should be directed to Managing Editor of Chiang Mai University Journal, Office of the President, Chiang Mai University, Chiang Mai 50200, ailand.

OWNERChiang Mai University, Chiang Mai 50200, ailand

EXECUTIVE EDITORProfessor Dr. Pongsak Angkasith PresidentAssociate Professor Vajara Rujiwetpongstorn, M.D. Vice President for Administrative AffairsAssociate Professor Dr. Daoroong Kangwanpong Vice President for ResearchAssistant Professor Dr. Pong-In Rakariyatham Vice President for Academic Affairs

EDITOR-IN-CHIEFProfessor Dr. Anan Ganjanapan E-mail: [email protected] E-mail: [email protected]

ASSOCIATE EDITORSAssociate Professor Dr. Prapa Sookgasem Associate Professor Dr. Sanay Yanasarn

ASSISTANT EDITORAssistant Professor Dr. Chusak Wittayapak

MANAGING EDITORAssociate Professor Dr. Jakkapan Sirithunyalug E-mail: [email protected] E-mail: [email protected]

ASSISTANT MANAGING EDITORSammanoon Noumanong E-mail: [email protected] Rithaporn E-mail: [email protected]

EDITORIAL OFFICEOffice of the President, Chiang Mai University, Chiang Mai 50200, ailand E-mail: [email protected] Website: http://cmuj.chiangmai.ac.th

EDITORIAL BOARDProfessor Dr. Michael Howard Simon Fraser University, CanadaProfessor Dr. T.E. Wagner University of Cincinnati, USAAssociate Professor Dr. P.I. Erera University of Washington, USAAssociate Professor Dr. Philip Hirsch Sydney University, AustraliaAssociate Professor Dr. Annop Pongwat Chiang Mai University, ailandAssociate Professor Dr. Sompong Witayasakpan Chiang Mai University, ailandAssociate Professor Dr. Songsak Sriboonchitta Chiang Mai University, ailandAssociate Professor Dr. Chaiyan Rajchagool Chiang Mai University, ailand

CM

U. Journal (2007) Vol. 1(1)

1

C H I A N G M A I U N I V E R S I T Y J O U R N A L

Supawat Rungsuriyawiboon1* and Alexei Lissitsa2*

How Far Are the Transition Countries from the Economic Standards in the European Union?

1Faculty of Economics, Chiang Mai University, Chiang Mai 50200, ailand.2Institute of Agricultural Development in Central and Eastern Europe, 06120 Halle (Saale), Germany.*Corresponding authors. E-mail:[email protected]; [email protected]

ABSTRACT

Economic reform in the Central and Eastern European countries in the 1980s helped transform the structure and volume of

agricultural production, consumption and trade, and resulted in significant agricultural productivity improvements. However, there are large differences among the transition countries in the magnitude and direction of these changes. e main objective of this study is to measure and compare the levels and trends in agricultural productivity in transition countries with those of the European Union (EU) countries, making use of the most-recent data available from the Food and Agriculture Organization (FAO). is study employs a parametric distance function approach to measure Malmquist productivity index as well as the magnitude and direction of technological change. e Malmquist productivity index is decomposed into technologi-cal change, technical efficiency change and scale efficiency change. ese measures will provide insightful information for researchers in designing policies to achieve a high growth rate in transition countries.

Key words: Agriculture, Efficiency, Parametric, Productivity, Transition Countries

INTRODUCTION

The Central and Eastern European Countries (CEECs) and Newly-Independent States (NIS) of the former Soviet Union began major

market-oriented reform of their planned economies in the late 1980s and early 1990s. Economic reform has transformed the structure and volume of these countries’ agricultural production, consumption and trade and introduced important agricultural productivity changes. However, there are large differences among the transition countries in terms of the mag-nitude and direction of these changes. While the agricultural output has fallen drastically in Russia, Ukraine and some Central European Coun-

CM


2

2007Cmu. Journal of Soc. Sci. and Hum.

tries, gross agricultural output increased considerably in Hungary, Poland and other leading transition countries (FAO Agrostat, 2004). ere exist extremely-wide differences in agricultural productivity among transition countries. For example, agricultural output per worker in Moldova is approximately one-fiftieth of that in the Czech Republic. Relatively few transition countries have achieved levels of output per worker as high as an average in the EU countries. ese differences have widened during the last twelve years. is lag in the rate of productivity in agriculture represents a serious constraint on economic growth in many transition countries. ese differences need to be investigated and compared by how far agricultural productivity of the transition countries are from the economic standards in the EU countries. In the past decade, the number of papers investigating cross-coun-try differences in agricultural productivity levels and growth rates has expanded significantly. is is most likely driven by different factors like the availability of new panel data sets such as that produced by the FAO of the United Nations and the development of new empirical techniques to analyze this type of data, such as a nonparametric technique known as Data Envelopment Analysis (DEA) and a parametric technique known as Stochastic Frontier Analysis (SFA), described in Coelli, Rao and Battese (1998). ese new empirical techniques have been extensively applied in the literature to measure total factor productivity (TFP) growth because they provide a classification of what sources are attributed to TFP growth. is information is very useful and helps policy makers in designing proper policies to enhance the efficiency and productivity in the industry. However, there are only a few empirical studies which have published the agricultural productivity developments in transition countries. e main purpose of this paper is to measure TFP developments in agriculture of transition countries after breakdown of socialism and to compare their TFP growth with other European countries. In the literature, TFP can be measured by using productivity index. e most widely-used productivity index is Malmquist TFP index presented in Caves et al. (1982) and Färe et al. (1994). is Malmquist TFP index has become common in practice by applying two techniques such as nonparametric and para-metric to calculate the TFP index. In this study, we employ a parametric distance function approach to decompose the Malmquist TFP index into technical change, technical efficiency change and scale efficiency change. e study is empirically implemented by using a panel data set of the European agriculture on 46 countries over the time period of 1992-2002 to measure and compare the productivity growth among the European countries.

CM


3


e remainder of this paper is organized as follow. In Section 2, the theoretical concept of the Malmquist TFP growth decomposition is presented, followed by a discussion of the methodologies to measure the Malmquist TFP index decomposition. Next, the methodology which is applied in this paper is concluded and then an empirical framework to the Malmquist TFP index decomposition is presented. Section 3 discusses the data set and the definitions of all variables used in this study. Empiri-cal results are presented and discussed in Section 4 and then conclusions follow in the final section.

METHODOLOGY

TFP using productivity index is theoretically defined as the ratio of an aggregate output index to an aggregate input index. e most widely-

used productivity index is Malmquist TFP index presented in Caves et al. (1982) and Färe et al. (1994). e Malmquist TFP index measures the TFP change between two data points by calculating the ratio of two as-sociated distance functions. Either nonparametric or parametric techniques can be applied to calculate the component distance functions defined in the Malmquist TFP index.

e Malmquist TFP index decomposition e Malmquist TFP index is defined using distance function. Dis-tance function provides a convenient way to describe a well-behaved multi-input multi-output production technology without the need to specify behavioral assumptions such as cost minimization or profit maximization. Consider a data set consisting of a vector of inputs and outputs for each of the i-th country where i=1,...,I denotes a country index. Let the input and output vectors for the i-th country be denoted xi = (xi1,..., xiN)∈R+

N

and yi=(yi1,...,yiM)∈R+M, respectively. For any input vector x ∈ R+

Nand any output vector y ∈ R+

M, an input vector x ∈ R+N is transformed into net

outputs y ∈ R+M by a production technology. is production technol-

ogy can be defined using an output orientation x P(x) ⊆ R+M where

P(x)represents the subset of all output vector y ∈ R+M obtainable from x

or less than x for any x ∈ R+N.

With a specific time period, t, the production technology St transforms inputs xt ∈ R+

N into net outputs yt ∈ R+M for each time period t=1,...,T.

e production technology consisting of all feasible input-output vectors on the production possibility set at time t is defined as . (1)� � NM

ttttt RxfyyxS �� )(:,

CM


4


e distance function can be defined by rescaling the length of an input or output vector with the production frontier as a reference. e output distance function is defined as , (2)

where (xt , yt)≤1 if and only if (xt , yt) ∈ St. Furthermore, (xt , yt)=1 if and only if (xt , yt) is located on the outer boundary of the feasible production set which occurs only if production is technically efficient. e output-orientated Malmquist TFP index as defined by F?re et al (1994) measures the TFP change between two data points by calculating the ratio of the distances of each data point relative to a common technology. e output-orientated Malmquist TFP change index between periods t and t+1 is the geometric mean of adjacent-period output-orientated Malmquist TFP index which is given by

. (3)

e above Malmquist TFP change index can be decomposed in a way that highlights what sources attributing to the TFP growth which can be written as

(4)

where (xt+1, Yt+1, xt , ty) refers to technical efficiency change which measures the change in the output-orientated measure of Farrell technical efficiency between periods t and t + 1 and (xt+1, Yt+1, xt , ty) refers to technical change which is the geometric mean of the shift in technology in time t and t + 1at input levels xt and xt+1. e component distance functions in the above Malmquist TFP index decomposition can be measured using either nonparametric or parametric techniques. One main criticism of the Malmquist TFP index is that it is constructed under a constant returns to scale assumption of production technology. Hence, this Malmquist TFP index does not provide an accurate measure of productivity change because it ignores a measure of scale economies contribution. Ray and Desli (1997), Ray (1998) and Grifell and Lovell (1999) overcome this problem by developing a method using a nonparametric technique to decompose the Malmquist TFP index in which the contribution of scale economies is taken into account. e

� � � �� tttttot SyxyxD �� ,:min,

otD

otD

� � � ��

� ��

2/1

1

1111111 ,

,

,

,,,, �

�

��

��

�

��

ttot

ttot

ttot

ttot

ttttcrso yxD

yxD

yxD

yxDyxyxm

� � � ��

� ��

� ��

� � � �,,,,,,,

,

,

,

,

,

,,,,

1111

2/1

1111

1111111

ttttcrsotttt

crso

ttot

ttot

ttot

ttot

ttot

ttot

ttttcrso

yxyxTCyxyxTE

yxD

yxD

yxD

yxD

yxD

yxDyxyxm

��

��

��

��

��

��

��

crsoTE�

crsoTE�

CM


5


contribution of scale economies attributing to the Malmquist TFP growth can be measured using the ratios of distance function values correspond-ing to constant and variable returns to scale technologies. However, this framework can not be applied to a parametric technique because the constant returns to scale distance function measured by the parametric approach does not necessarily envelop the distance function with variable returns to scale leading to an inaccurate measure of the scale efficiency contribution. Subsequently, Balk (2001) extends the results obtained by Ray (1998) and derives the framework using a parametric technique to decompose the Malmquist TFP index into technical change, technical efficiency change, scale efficiency change and input- or output-mix effect. Although Balk’s approach is appealing, it does require the prior calculation of scale efficiency measures in which the scale effects are measured using the most productive scale size as a reference. As Orea (2002) pointed out, the scale efficiency measures are not bounded for either globally increasing, decreasing or constant returns to scale or for ray-homogenous technolo-gies. More simply, in the case of single output, a U-shaped average cost curve is required for the most productive scale size to exist. erefore, some practical problems may occur when adopting Balk’s approach. As this result, Orea (2002) presents an alternative approach using a parametric technique to decompose the Malmquist TFP index in which the contribu-tion of scale economies is taken into account without requiring the prior calculation of scale efficiency measures. Orea applies Diewert’s (1976) Quadratic Identity Lemma to derive a generalized Malmquist TFP index decomposition which overcomes the practical problems of measuring the scale efficiency contribution shown in Balk. A parametric decomposition of a Generalized Malmquist TFP index presented in Orea (2002) is sum-marized in the following sections.

A Generalized malmquist TFP index decomposition Following Orea (2002), the logarithmic form of changes in output distance function between periods and can be written as

(5)

where represents the distance elasticities for the m-th output in period t and represents the distance elastici-ties for the k-th input in period t.

� � � � � �

� � ,lnln

21

ln21

ln21

,ln,ln

11

11

1

11111

��

��

��

��

��

��

��

��

��

��

��

��

�

��

�

�

t

D

t

D

x

xee

yy

yxDyxD

ot

ot

kt

ktK

kmtmt

mt

mtM

mmtmttt

ottt

ot ��

motmt yD lnln ��

ktotkt xDe lnln ��

CM


6


An output-oriented Malmquist TFP change index between periods t and t + 1 is defined as the difference between aggregating the growth in outputs and inputs between periods t and t + 1. By following Denny, Fuss and Waverman (1981), aggregating the growth in inputs is defined using distance elasticity shares rather than distance elasticities in order to satisfy all desirable properties of the TFP index1. e logarithmic form of a generalized output-oriented Malmquist TFP change index between periods and can be written as

, (6)

where represents the distance elasticity share for the k-th input in period t.Rearranging Equation (6), the logarithmic form of the generalized output-oriented Malmquist TFP change index can be decomposed as

(7)

where represents the technical efficiency change which measures the change in the technical efficiency prediction of the i-th firm at the period t and period t + 1, represents the techni-cal change which measures the mean of the technical change evaluated at the period t and period t + 1 data points and rep-resents the scale efficiency change which measures the change in scale efficiency at the period t and period t + 1 data. Equation (7) is expressed in terms of proportional rates of growth instead of a product of indices as in Equation (4). e is viewed as the parametric counterpart of the Malmquist TFP index.

Empirical framework to a parametric decompositionof a generalized malmquist TFP index

e components of the generalized Malmquist TFP change index in Equa-tion (7) can be measured by estimating a translog output distance function. For the case of M output and K inputs, a translog output distance function for a panel of i = 1,...,I and t = 1,...,T can be defined as follow

� � � � � ��

��

��

��

��

��

�

��

kt

ktK

kktkt

mt

mtM

mmtmttttt

vrso x

xss

y

yyxyxm 1

11

1

1111 ln

21

ln21

,,,ln ��

��

�K

kktktkt ees

1

1Four desirable properties are ideentity, monotonicity, separability and proportionality.

� � � �

� � � � � �,,,,ln,,,ln,,,ln

ln1121

lnln21

lnln,,,ln

111111

1

1 11

11

1111

ttttvrsotttt

vrsotttt

vrso

kt

ktK

kkt

K

kktkt

K

kkt

ot

oto

tottttt

vrso

yxyxSCEyxyxTCyxyxTE

xx

sese

tD

tD

DDyxyxm

��

�

� ��

��

��

��

��

��

��

�

��

��

��

��

��

��

��

��

��

��

��

� ��

� �ttttvrso yxyxTE ,,, 11 ��

� �ttttvrso yxyxTC ,,, 11 ��

lnmovrs

� �ttttvrso yxyxSCE ,,, 11 ��

CM


7


(8)

where β s are unknown parameters to be estimated. Young’s theorem requires that the symmetry restriction is imposed so that βxkxl = βxlxk and βymyn = βynym . e additional restrictions required for homogeneity of degree +1 in outputs are (k=1,...,K) and . ese restrictions can be imposed by estimating a model where the M-1 output quantities are normalized by the M-th output quantity. Equation (8) yields the estimating form of the output distance function, in which the distance term, D o, can be viewed as an error term as follow

(9)

where and . By replacing the distance term, , with a composed error term, vit – uit, Equation (9) can be esti-mated as a standard stochastic frontier production function where s are the random errors, assumed to be i.i.d. and have –distribution, independent of the uit, the technical inefficiency effects. e u s are assumed to be i.i.d. normal random variable, . Given the distributional assumptions of the random variables defined in the above, the unknown parameters in Equation (9) are estimated us-ing the method of maximum likelihood (ML). Battese and Corra (1977) suggested that the two variance parameters can be replaced by the two new parameters and . e γ-parameterization has advantages in seeking to obtain ML estimates because the parameter space for γ can be searched for a suitable starting value for the iterative maximiza-tion routine. e unknown parameters are obtained by using the computer program, FRONTIER (Coelli, 1996a). e components of the Malmquist TFP change index presented in Equation (7) can be computed after estimating the output distance func-tion in Equation (9). e technical efficiency change can be calculated by

(10)

� �

,lnln

2

1lnlnlnln

2

1

lnlnln2

1ln,ln

11

2

1 11 1

11 110

��

��

��

��

� ��

��

��

��

��

M

mmittykit

K

ktx

ttt

K

kmitkit

M

myxlitkit

K

k

K

lxx

kit

K

kx

M

m

M

nnitmityy

M

mmitytt

oit

tytx

ttyxxx

xyyyyxD

mk

mklk

knmm

��

��

��

,11

��

M

mym

� 0 (m=1,...M) ,1

��

M

nyy nm

� 01

��

M

myx mk

�0

1��

�

M

mtym

�

,lnlnln

2

1lnlnlnln

2

1

lnlnln21

lnln

1

1

*

1

2

1

1

1

*

1 1

1

1

1

1

1

**1

1

*0

oit

M

mmittykit

K

ktx

ttz

K

k

M

mmitkityxlitkit

K

k

K

lxx

kit

K

kx

M

m

M

nmitmityy

M

mmityMit

Dtytx

ttyxxx

xyyyy

mk

mklk

kmmm

��

��

��

�

��

�

�

��

�

�

�

�

�

�

�

��

��

��

� �Mitmitmit yyy �*itit

oit uvD �� ln

oitDln�

),0( 2vN �

� �2,0~ uNu �

222uv �� 22 �� u�

� � � ��

��

��

��

��

��

ititit

ititit

it

itvrso uvuE ,

uvuE

TE

TETE

exp

explnlnln 1111

CM


8


where TEit represents the technical efficiency prediction of the i-th firm in the t-th time period. e other components of the Malmquist TFP change index can be written in terms of the parameter estimates of the output distance function in Equation (5). e technical change yields

.(11) e scale efficiency change in terms of the parameter estimates of the output distance function yields

, (12)

where and .

DATA DISCUSSIONS

A data set used in this study is adjusted for quality which measures agricultural outputs and inputs. Data on 46 countries over the time

period of 1992 through 2002 are used in the empirical analysis. Countries are divided into three categories, using the following definitions. e first category called “EU 15” countries consists of countries which founded the EU and countries which joined the EU before 1996. We also include Norway and Switzerland into this group. e second category called “EU 10” countries consists of countries which joined the EU in 2004. e last category called “transition” countries consists of all transition countries after the breakdown of the former Soviet Union as well as Turkey. A list of the countries in each group is summarised in Table 1.

Table 1: Classification of Selected Countries.Group Country

EU15* Austria, Belgium-Luxembourg, Denmark, Finland, France, Germany, Greece, Ireland, Italy, the Netherlands, Norway, Portugal, Spain, Sweden, Switzerland, UK

EU10** Cyprus, Czech Republic, Estonia, Hungary, Latvia, Lithuania, Malta, Poland, Slovakia, Slovenia

Transition Albania, Armenia, Azerbaijan, Belarus, Bosnia-Herzegovina, Bulgaria, Croatia, Georgia, Kazakhstan, Kyrgyzstan, Macedonia, Moldova Republic, Romania, Russian Federation, Serbia-Montenegro, Tajikistan, Turkey, Turkmenistan, Ukraine, Uzbekistan

*Countries joined the EU before January, 1995 including Norway and Switzerland **Countries joined the EU in May, 2004

� ��

��

�

��

��

1

1

**1

11 lnlnlnln212

21

lnM

mmitmitty

K

kkitkittxttt

vrso yyxxtTC

mk��

��

��

��

�

��

��

��

��

��

��

� ��

��

kit

kitK

kkit

K

kkitkit

K

kkit

vrso x

xseseSEC 1

1 11

11 ln11

21

ln

ittxkit

M

myxkit

K

kxxkit

otkit tyxxDe

kmkkkk��

�

��

*1

11

lnlnlnln ��

�K

kkitkitkit ees

1

CM


9


e primary source of data is obtained from the website of the FAO of United Nations and, in particular, the agricultural statistics provided by the AGROSTAT system, supported by the Statistics Division of the FAO. e FAO dataset used in this study has been used in many studies about agricultural productivity. e data used to measure the TFP decomposition contain the measurements of agricultural output and input quantities. In this study, the production technology is presented by two output variables (i.e., crop and livestock) and five input variables (i.e., land, tractor, labour, fertilizer and livestock). e definitions of these variables are summarised as follow: e output series for the two output variables are derived by aggre-gating detailed output quantity data on 127 agricultural commodities (115 crop commodities of average 1999-2001 and 12 livestock commodities), which are produced in the studied countries. Construction of output data series uses the following steps. First, average aggregate for the base period 1999 to 2001 are calculated. ese aggregates are constructed using output quantity data and international average prices (expressed in US dollars) derived using the Geary-Khamis method2. e next step is to extend the average base period output series 1999 to 2001 to cover the whole study period 1992-2002. is is achieved using the FAO production index number series for crops and livestock separately3. Given the constraints on the number of input variables that could be used in the analysis, only five input variables are considered to be used in the study. Definitions of these input variables are defined as follow. Land input variable represents the arable land, land under permanent crops as well as the area under permanent pasture in hectares. Tractor input variable represents the total number of wheel and crawler tractors, but excluding garden tractors, used in agriculture. Labour input variable refers to economically-active population in agriculture. Following other studies (Hayami and Ruttan 1970, Fulginiti and Perrin 1997) on inter-country comparison of agricultural productivity, fertilizer input variable represents the sum, in nutrient-equivalent terms, the commercial use of nitrogen, potassium, and phosphate expressed in thousands of metric tons. Livestock input variable used in the study is the sheep-equivalent of the five categories of animals used in constructing this variable. e categories considered are: buffaloes, cattle, pigs, sheep and goats. Numbers of these animals are converted into sheep-equivalents using conversion factors: 8.0 for buffalos and cattle; and 1.00 for sheep, goats and pigs.

2Detailed information on how international average prices are constructed see Rao (1993). 3See the FAO STAT (FAO, 2004) for details regarding the construction of production indwx numbers.

CM


10


Descriptive statistics of the variables used in the study are presented in Table 2. ere are large variations observed in output and input variables across countries.

Table 2: Data Overview for 46 Selected Countries, 1992 to 2002.Variables Units Average Min Max S.D.

Input Land ×103 hectares 18,005 9,000 221,747 43,481

Machinery tractors 275,217 450 1,660,000 395,242

Fertiliser metric tons 603,607 700 5,510,000 946,751

Labour ×103 persons 1,236 2 14,697 2,461

Livestock ×103 heads 50,223 330 593,697 75,742

Output Crop value ×103 dollars 3,436,528 15,710 21,851,139 5,011,663

Livestock value

×103 dollars 3,244,477 30,000 26,888,503 4,543,331

RESULTS

Prior to estimation, all variables are scaled to have unit means. is transformation does not alter the performance measures obtained,

but does allow one to interpret the estimated first-order coefficients of the translog output distance function as elasticities of distance with respect to inputs and outputs evaluated at the sample means. Livestock output is used as the normalizing output (see Equation 9). e translog output distance function is estimated using the approach described in Section 2.3. Hypothesis tests regarding the structure of the production technology such as the presences of technical inefficiency and technical change are conducted using the likelihood ratio tests. All null hypotheses are rejected which imply the existences of technical inefficiency and technical change in the model. We began by estimating the translog output distance function in Equation (9) using the method of ML. e set of ML estimates is presented in Table 3. All the first-order coefficients have the expected signs, implying that the output distance functions are increasing in outputs and decreasing in inputs at the sample mean. e estimate of the output elasticities is 0.2986 and 0.7014 for crops and livestock, respectively. e estimates of the input elasticities are –0.1175, –0.1945, –0.2154, –0.0259 and –0.4675 for land, tractors, fertilizer, labor, and livestock, respectively. e sum of the input elasticities provides information on scale economies. e sum of these input elasticities is –1.0208, indicating that the technology exhibits moderately increasing returns to scale at the sample mean. e first-order coefficients of the time trend variable in Table 3 provide estimates of the

CM


11


average annual rate in technical change. e output distance function estimates suggest that the technology is improving at a rate of 2.57% per annum. Following the estimation, tests of the regularity conditions are checked at each data point in the sample of 506 observations. We find that the convexity condition and the monotonicity constraints in outputs are satisfied at all observations in the output distance function. e monotonicity constraints in inputs are violated at 12, 0, 2, 14, and 0 percent of all observations in the case of land, tractors, fertilizer, labor and livestock inputs, respectively4. en, the parameter estimates presented in Table 3 are used to calculate the components of the Malmquist TFP growth decomposition. Table 4 presents unweighted average values of technical efficiency scores and the components of the Malmquist TFP growth for the 46 countries over the period 1992 to 2002. We begin by discussing the results of technical efficiency scores, followed by the results of the Malmquist TFP growth. Estimated technical efficiency scores for each firm in the sample are presented in the third column of Table 4. Technical efficiency scores range from 0.582 by Belarus to maximum 0.933 by Bulgaria with an unweighted average of 0.818. e average technical efficiency score implies that the countries in this study were, on average, producing 81.8 percent of the outputs that could be potentially produced using the observed input quantities. Some transition counties such as Bulgaria and Moldova Republic exhibited quite impressive technical efficiency scores over the sample period. According to the Government’s figures, Bulgaria’s agriculture currently generates about 12 percent of gross domestic prod-uct and provides a livelihood for about 368,000 people and a subsidiary source of income for almost one million people. e Bulgarian agricultural sector has been highly subsided by Government unlike other transition countries. e subsidies were not covered by this analysis and this could lead to the overestimation of its technical efficiency scores. Moldova Republic showed a significant increase of technical efficiency scores after the year 1997. An increase in technical efficiency scores could be explained by a decrease in its uses of fertilizers and plant-protection agents. e new ownership conditions and the fragmentation of plots do not permit farmers to undertake the necessary expenditures. Ninety-nine percent of the crops were produced without the use of fertilizers or plant-protection agents. erefore, yields were highly depended on the natural conditions. Average country technical efficiency scores by the group of the countries

4ese regularity conditions can be imposed using the Bayesian approach to imposing curvatures on dis-tance function presented in O’Donnell and Coelli (2005) or performing deterministic parametric linear programming frontier.

CM


12


indicate that average country technical efficiency scores is 0.853 by the EU 15 countries, 0.844 by the EU 10 countries and 0.777 by the transi-tion countries group. Average country technical efficiency scores of the transition countries were lower than those of the EU 10 and 15 countries, respectively, in every single period. e components of the Malmquist TFP growth decomposition are calculated from the parameter estimates presented in Table 3. e Malmquist TFP growth can be decomposed into technical efficiency change, technical change and scale efficiency change effects. e Malmquist TFP growth ranges from –0.49 percent by Ireland to 6.43 percent by Tajiki-stan with an unweighted average of 2.28 percent. Two countries such as Ireland and Turkey which exhibited TFP regress over the sample periods. TFP regress for Ireland was driven by deterioration in both technical and scale efficiencies whereas TFP regress for Turkey was due to technologi-cal regress and deterioration in scale efficiency. e technical efficiency change ranges from –2.67 percent by Turkmenistan to 3.97 percent by Tajikistan with an unweighted average of –0.14 percent. Twenty-four countries showed deterioration in technical efficiency change. Of these countries, nine countries are within the EU 15 countries; six countries are within the EU 10 countries and nine countries are within the transition countries. A significant deterioration in technical efficiency change in the republics of former Yugoslavia could be simply explained by civil war and a political instability during the study periods. A deterioration in techni-cal efficiency change in some countries within the EU 15 countries such as UK, in our opinion, needs a special statement. e negative technical efficiency change in UK is correlated with BSE5 and FMD6 crises in the country. Simultaneous occurrence of these diseases in England by the end of the 1990s caused confusion among consumers about the issues and interrelationships, and the combined costs to the UK economy have been shared by agriculture, consumers, tourism, and trade. For example, the 1996 outbreak was followed by an immediate 40 percent decline in sales of beef products and 26 percent decrease in households consumption of beef and veal. e estimated costs of BSE to the UK were at over 5 billion (USDA, 2001). A significant acceleration in technical efficiency change in Post-Soviet republics like Slovenia, Tajikistan, Albania and Moldova Replubic could be explained by a drastic reduction of the variable inputs use like fertilizers, machinery and livestock numbers. e technical change ranges from –0.21 percent by Turkey to 6.95 percent by Kazakhstan with an

5BSE-Bovine spongiform encephalopathy6FMD-Foot and mouth disease

CM


13


unweighted average of 2.57 percent. All countries except Turkey indicated technological progress. Many countries within the EU 10 and transition countries showed significant technological progress over the time periods. e scale efficiency change ranges from –1.97 percent by Kazakhstan to 0.84 percent by Belarus with an unweighted average of –0.15 percent. Twenty-five countries showed deterioration in scale efficiency change. Of these countries, six countries are within the EU 15 countries; five coun-tries are within the EU 10 countries and fourteen countries are within the transition countries. Table 5 presents weighted growth rate of the TFP growth decom-position and its components by the group of the countries over the period 1992 to 2002. Examining the growth rate by the group of countries will allow us to explain agricultural productivity trends in the European coun-tries and to answer the question we raised earlier that how far agricultural productivity of the transition countries are from the economic standards in the EU countries. TFP growth by all countries increases by 16.80 percent over the sample period with a weighted average of about 1.527 percent per annum. Overall, technical change and scale efficiency change increase by 16.46 and 0.59 percent over the sample period for a weighted aver-age of about 1.496 and 0.054 percent per annum, respectively, whereas technical efficiency change decreases by 0.3 percent over the sample period with a weighted average of about –0.027 percent per annum. e EU 15 countries indicated the TFP growth increases by 14.21 percent over the sample period with a weighted average of about 1.292 percent per annum. Technical change and scale efficiency change increase by 14.90 and 0.66 percent over the period 1992 to 2002 for a weighted average of about 1.355 and 0.060 percent per annum, respectively, whereas scale efficiency change decreases by 1.26 percent over the sample period with a weighted average of about –0.114 percent per annum. ere were deceleration in technical efficiency change during the periods 1992 to 1996 and deceleration in scale efficiency change during the periods 1992 to 1994. e TFP growth by the EU 15 countries was low during the periods 1992 to 1994 and 2000 to 2002. e EU 10 countries indicated the TFP growth increases by 15.31 percent over the sample period with a weighted average of about 1.392 percent per annum. Technical efficiency change and technical change increase by 1.28 and 13.87 percent over the sample period for a weighted average of about 0.117 and 1.261 percent per annum, respectively, whereas scale efficiency change decreases by 0.02 percent over the sample period with a weighted average of about –0.002 percent per annum. ere were deceleration in technical efficiency change during the periods 1992 to 1994 and deceleration in scale efficiency change during the periods 1994 to 1998 and 2000 to 2002. e TFP growth by the EU 10 countries was

CM


14


low during the periods 1998 to 2002. e transition countries indicated the TFP growth increases by 20.70 percent over the sample period with a weighted average of about 1.882 percent per annum. Technical efficiency change, technical change and scale efficiency change increase by 0.45, 19.53 and 0.53 percent over the sample period for a weighted average of about 0.041, 1.775 and 0.048 percent per annum, respectively. ere were deceleration in technical efficiency change during the periods 1992 to 1998 and deceleration in scale efficiency change during the periods 1998 to 2000. e TFP growth by the transition countries was low during the periods 1994 to 1996. TFP growth for each group of countries was mainly driven by technology progress. e results indicate deterioration in technical efficiency by the EU 15 countries but acceleration in technical efficiency by the EU 10 and transition countries. is result implies that the EU 10 and transition countries increased the outputs by improving technical efficiency more than the EU 15 countries group. Technological progress by the transition countries was higher than the EU 15 countries and EU 10 countries, respectively. e results show deterioration in scale efficiency by the EU 10 countries but acceleration in scale efficiency by the EU 15 and transition countries.

Table 3: Estimated Parameters of the Output Distance Modela.Parameter Estimates t-Statistic Parameter Estimates t-Statistic

β0 0.3694 -.0.1225 βx2x5 8.4473 -1.9714βy1 0.2986 0.0760 βx3x4 12.7308 1.7289

βy1y1 0.8281 -0.0860 βx3x5 9.4575 -1.2925βx1 -0.1175 -0.1442 βx4x5 -5.2767 -2.0204βx2 -0.1945 -0.1999 βx1y1 -9.2042 -3.5792βx3 -0.2154 -0.1379 βx2y1 -8.3084 -3.7666βx4 -0.0259 -0.1351 βx3y1 -1.0267 -3.7880βx5 -0.4675 -0.1762 βx4y1 -10.9166 -3.1119

βx1x1 0.0936 0.5039 βx5y1 2.1319 5.6849βx2x2 0.0010 -0.0257 βt 0.0288 -8.2120βx3x3 -0.1328 0.0023 βtt -5.0181 1.0809βx4x4 0.3414 -0.1789 βx1t 6.8838 -3.7616βx5x5 0.4778 -0.0121 βx2t 4.1586 -0.2991βx1x2 -0.0455 0.0424 βx3t -1.3196 0.8754βx1x3 0.0943 0.1328 βx4t 3.0327 2.5739βx1x4 -0.1264 0.0995 βx5t -3.7959 1.228βx1x5 -0.0897 -0.0599 βy1t -1.4725 -0.9429

CM


15


βx2x3 0.1276 0.0754 σ2 7.1548 5.6135βx2x4 -0.0554 0.7186 γ -1.8007 5.7077

aSubscripts on βx coefficients refer to inputs: 1 = land; 2 = tractors; 3 = fertilizer; 4 = labor; 5 = livestock input and subscripts on βy coefficients refer to outputs: 1 = crops; 2 = livestock output

Table 4: Unweighted Average Values of Technical Efficiency Scores and TFP Growth by Each Country.

Country Region TE levels

Average Value in PercentageTE

ChangeTechnical Change

Scale Change

TFP Change

Austria EU15 0.859 0.01 2.54 -0.15 2.40Bel-Lux EU15 0.842 0.17 1.31 0.13 1.61Denmark EU15 0.922 0.29 2.12 0.20 2.61Finland EU15 0.884 -0.62 2.65 0.00 2.04France EU15 0.836 0.16 1.43 0.00 1.60Germany EU15 0.893 0.29 0.56 0.13 0.98Greece EU15 0.902 -0.39 3.26 -0.27 2.59Ireland EU15 0.810 -1.32 1.04 -0.22 -0.49Italy EU15 0.855 -0.20 1.25 -0.07 0.98Netherlands EU15 0.903 0.24 0.31 0.35 0.91Norway EU15 0.781 -1.00 1.50 0.09 0.58Portugal EU15 0.751 -0.68 1.49 -0.01 0.81Spain EU15 0.827 -0.89 2.22 0.11 1.44Sweden EU15 0.870 -0.36 2.90 -0.02 2.52Switzerland EU15 0.897 0.06 1.66 0.12 1.83UK EU15 0.811 -0.56 1.43 0.12 0.98Cyprus EU10 0.786 -1.19 2.79 -0.14 1.46Czech Rep EU10 0.855 -0.37 1.57 0.13 1.32Estonia EU10 0.842 1.95 4.31 -0.93 5.33Hungary EU10 0.929 -0.59 2.79 0.01 2.21Latvia EU10 0.801 0.72 4.28 -0.56 4.44Lithuania EU10 0.802 -0.83 3.42 -0.13 2.45Malta EU10 0.884 -0.50 3.75 0.03 3.28Poland EU10 0.834 0.46 0.11 0.04 0.60Slovakia EU10 0.899 -0.38 2.34 0.12 2.08Slovenia EU10 0.808 1.09 3.75 -0.61 4.23Albania Trans 0.862 0.86 0.55 -0.75 0.65Armenia Trans 0.781 -1.12 3.83 -0.17 2.55Azerbaijan Trans 0.670 -0.93 2.42 -0.10 1.39

Table 3: (Continued).

CM


16


Belarus Trans 0.582 0.83 1.49 0.84 3.17Bosnia Herzg Trans 0.678 -0.58 5.17 -1.11 3.49Bulgaria Trans 0.933 0.12 3.80 -0.20 3.71Croatia Trans 0.842 0.51 3.73 0.11 4.35Georgia Trans 0.786 -1.38 2.82 -0.26 1.18Kazakhstan Trans 0.847 0.10 6.95 -1.97 5.08Kyrgyzstan Trans 0.871 -0.16 4.67 -0.11 4.40Macedonia Trans 0.607 -1.79 4.60 0.12 2.93Moldova Rep Trans 0.906 0.81 3.39 -0.98 3.22Romania Trans 0.866 0.09 1.77 0.03 1.89Russian Fed Trans 0.834 0.01 1.81 0.30 2.12Serbia-Monte Trans 0.820 -0.44 1.95 -0.17 1.34Tajikistan Trans 0.679 3.97 3.43 -0.96 6.43Turkey Trans 0.827 0.06 -0.21 -0.01 -0.17Turkmenistan Trans 0.641 -2.67 6.37 -0.20 3.50Ukraine Trans 0.818 0.25 1.28 0.30 1.83Uzbekistan Trans 0.687 -1.03 2.16 -0.30 1.10Mean EU15 0.853 -0.28 1.68 0.05 1.45Mean EU15 0.844 0.04 2.91 -0.21 2.74Mean Trans 0.777 -0.12 3.10 -0.27 2.71Mean ALL 0.818 -0.14 2.57 -0.14 2.28

Table 5: Weighted Growth Rates of the Malmquist TFP Growth Decomposition by Group of the Countries (in percentage).

Period Region Efficiency Change

Technical Change

Scale Efficiency Change

TFP Change

1992-1994 EU15 -0.793 1.440 -0.001 0.6121994-1996 EU15 -0.181 1.164 0.043 1.0201996-1998 EU15 0.074 0.931 0.033 1.0421998-2000 EU15 0.310 0.690 0.099 1.1092000-2002 EU15 0.180 0.479 0.046 0.7081992-2002 EU15 -0.114 1.355 0.060 1.2921992-1994 EU10 -0.883 1.216 0.014 0.3151994-1996 EU10 0.636 1.171 -0.032 1.7961996-1998 EU10 0.315 0.916 -0.015 1.2241998-2000 EU10 0.151 0.659 0.056 0.8702000-2002 EU10 0.226 0.431 -0.028 0.6321992-2002 EU10 0.117 1.261 -0.002 1.392


CM


17


1992-1994 Trans -0.142 1.496 0.260 1.6191994-1996 Trans -0.384 1.374 0.009 0.9831996-1998 Trans -0.003 1.233 0.008 1.2381998-2000 Trans 0.407 1.058 -0.122 1.3502000-2002 Trans 0.279 0.895 0.021 1.2041992-2002 Trans 0.041 1.775 0.048 1.8821992-1994 All -0.531 1.440 0.106 0.9961994-1996 All -0.176 1.243 0.024 1.0861996-1998 All 0.065 1.037 0.019 1.1241998-2000 All 0.329 0.815 0.018 1.1712000-2002 All 0.220 0.627 0.030 0.8821992-2002 All -0.027 1.496 0.054 1.527

CONCLUSIONS

This study applies a parametric distance function approach presented in Orea (2002) to measure the Malmquist TFP index decomposi-

tion. e Malmquist TFP index can be decomposed into technological change, technical efficiency change and scale efficiency change. is study is empirically implemented by using a panel data set of the European agriculture on 46 countries over the time period of 1992-2002 to measure and compare agricultural productivity in the transition countries with those of the EU countries. e decomposition of Malmquist TFP growth will provide insightful information for researchers in designing policies to achieve a high growth rate in transition countries. e empirical findings indicate that the weighted average TFP growth in the European agriculture over the study period grew at 1.527 percent per annum which was driven by –0.027 percent in technical efficiency change, 1.496 percent in technical change and 0.054 percent in scale efficiency change. Turning to the performance of the different groups of countries, the EU 15 countries operated at higher technical efficiency levels than the EU 10 and transition countries over the study periods. e weighted average TFP growth grew at 1.292 percent per annum for the EU 15 countries, 1.392 percent per annum for the EU 10 countries and 1.882 percent per annum for the transition countries. TFP growth for each group of countries was mainly driven by the technology progress. e results also show that the EU 10 and transition countries increased the outputs by improving technical efficiency more than those located within the EU 15 countries. Transition countries indicated impressive “catch-up” effect comparing with the EU 15 and 10 countries.


CM


18


REFERENCESBalk, B. 2001. Scale efficiency and productivity change Journal of Produc-

tivity Analysis 15, 159-183.Battese, G. E., and G. S. Corra. 1977. Estimation of a production frontier

model: with application to the pastoral zone of Eastern Australia Australian Journal of Agricultural Economics 21, 169-179.

Caves, D. W., L. R. Christensen, and W. E. Diewert. 1982b. e economic theory of index numbers and the measurement of input, output and productivity Econometrica 50, 1393-1414.

Coelli, T. J. 1996b. A guide to DEAP version 2.1: a data envelopment analysis CEPA Working Paper 96/08, Department of Econometrics, University of New England, Armidale.

Coelli, T. J., and D. S. P. Rao. 2003. Total factor productivity growth in agriculture: a Malmquist index analysis of 93 countries, 1980-2000 Agricultural Economics, to appear.

Coelli, T. J., D. S. P. Rao, and G. E. Battese. 1998. An introduction to efficiency and productivity analysis, Kluwer Academic Publishers.

Denny, M., M. Fuss, and L. Waverman. 1981. e measurement and in-terpretation of total factor productivity in regulated industries with an application to Canadian telecommunication in T. G. Cowing and R. E. Stevenson (eds) Productivity Measurement in Regulated Industries. New York: Academic Press, 179-218.

Diewert, W. E. 1976. Exact and superlative index numbers Journal of Econometrics 4, 115-145.

Färe, R., S. Grosskopf, and C. A. K. Lovell. 1985. e measurement of efficiency of production, Kluwer Academic Publishers, Boston.

Färe, R., S. Grosskopf, M. Norris, and Z. Zhang. 1994. Productivity growth, technical progress and efficiency changes in industrialised countries American Economic Review, 84, 66-83.

Färe, R., E. Grifell-Tatjé, S. Grosskopf, and C. A. K. Lovell. 1997. Biased technical change and the Malmquist productivity index Scandinavian Journal of Economics, 99, 119-127.

Fulginiti, L. E., and R. K. Perrin. 1997. LDC agriculture: nonparametric Malmquist productivity indexes Journal of Development Economics, 53, 373-390.

Fuentes, H. J., E. Grifell-Tatjé, and S. Perelman. 2001. A parametric dis-tance function approach for Malmquist productivity index estimation Journal of Productivity Analysis, 15(2), 79-94.

Grifell, E., and C. A. K. Lovell. 1999. A generalized Malmquist productivity index Sociedad de Estadostica e Investigacion Operativa, 7, 81-101.

Hayami, Y., and V. Ruttan. 1970. Agricultural productivity differences among countries American Economic Review 40, 895-911.

CM


19


Hicks, J. R. 1963. eory of wages, 2nd Edition, London, Macmillan.Orea, L. 2002. Parametric decomposition of a generalized Malmquist

productivity index Journal of Productivity Analysis 18, 5-22.O’Donnell, C., and T. J. Coelli. 2005. A Bayesian approach to imposing

curvature on distance functions Journal of Econometrics 126 (2), 493-523.

Ray, S. C., and E. Desli. 1997. Productivity growth, technical progress, and efficiency change in industrialized countries: Comment American Economic Review 87, 1033-1039.

Ray, S. C. 1998. Measuring scale efficiency from a Translog production function Journal of Productivity Analysis 11, 183-194.

Rao, D. S. P. 1993. Intercountry comparisons of agricultural output and productivity, FAO, Rome.

CM


21


Kriengsak Pattamarakha1*, Boonlert Srinoy2

and Somkiat Saithanoo3

Leadership in Dairy-Cattle Production in Southernailand : Dimensions in Socio-Economic Status,

Adoption of Recommended Practices, and Communications

1Department of Agricultural Development, Faculty of Natural Resources, Prince of Songkla University, Songkhla 90112, ailand.2College of Agriculture and Technology, Phatthalung 93000, ailand.3Department of Animal Sciences, Faculty of Natural Resources, Prince of Songkla University, Songkhla 90112, ailand.*Corresponding author. E-mail: [email protected]

ABSTRACT

This study aimed to identify farmers’ systems of dairy production, growth diffusion of dairy cattle raising, leadership in dairy cattle

production, and group social networks used for giving, and finding farm information. It also compared some characteristics between leaders and followers in farm organizations. Muang district, Phatthalung province, was selected as the study area, as there were quite a lot of dairy farmers who lived there. Since there was limited population, all were used without sampling. Ninety-one farmers were interviewed. e findings revealed that dairy cattle farming was first undertaken in 1973 by one farmer, and the number of dairy-cattle farmers gradu-ally increased each year. Five opinion leaders were identified using socio – metric techniques. ey were higher in socio-economic status than their followers. ey received requests for information rather than initiating the dissemination of information. e study implies that opinion leaders acted as consultants for farmers in dairy cattle production. It also implies that such opinion leaders should be involved in information delivery systems. ey should also act as catalysts for development and work together with livestock extension officers.

Key words: Dairy cattle, Opinion leader, Social network, Practice adoption

CM


22


INTRODUCTION

About 60% of ailand’s population are engaged in agriculture. us, agriculture plays a very important role in the economy of the country.

Agricultural extensions have been undertaken in order to increase farm produce throughout the whole country in primary production – crop, animal and fishery. Rural people can have sufficient food for their own consumption, and the rest of the farm produce can be sold for income. e raising of dairy cattle for milk production in ailand com-menced in 1961 with the assistance of the Danish government. At that time, the contributions were in the form of technical support, and the provision of good parent stock. In addition, those involved were trained in various disciplines of dairy cattle raising. Under the 5th National Social and Economic Plan (1982 – 1986), an annual estimate of 33,000 tons of raw milk production or 90 tons per day was forecasted. In 1985, average dairy milk production was 75 tons which was 15 tons lower than the estimated amount (Samnakhon, 1992). From 1985 to 1991, the demand for milk increased gradually. In 1992, the demand for milk products increased sharply, especially for fresh and processed milk. As a result, ailand imported 114,013 tons of milk with a value of 5,570 million Baht (Chantalakhana, 1995). In 1993, raw milk production in ailand was about 900 tons per day. is had paved the way for dairy cattle raising, although most milk production was still carried out by small farmers, each having about 5 – 15 dairy cows, a low income and limited financial investment. ese farmers generally only had a small area of land suitable for pasture, with the result that the cost of production was high and profits were low (Chanatalakhana, 1995). In 1995, ailand had 17,800 farmers with a total of 292,694 dairy cows. e average raw milk production was 959 tons per day, or 350,035 tons per year, while the demand for raw milk was 3,000 tons per day (Office of Agricultural Economics, 1996). However, the 7th National Social and Economic Plan (1992 – 1996) forecasted that the demand for raw milk production would be only 1,300 tons per day, or 475,000 tons per year (Phatthalung Provincial Office of Livestock Development, 1995). Under this situation, the ai government paid much more attention to increase the production of raw milk, in order to reduce imported milk from abroad. Promotion and extension of dairy cattle raising was undertaken with the full support of the government. e 8th National Social and Economic Development Plan (1997 – 2001) stressed an increase in milk production. is led to an increase in the number of farmers raising dairy cattle. However, associated problems

CM


23


still remained from past practices, particularly the unreadiness of farm-ers, a lack of skill and knowledge in dairy raising and its financing, and intermittent aid from the government. At present, ailand has 17,893 dairy farms with 358,440 dairy cattle. Each farm has an average number of 20 cattle. e total milk produc-tion per day is 2,037 tons, which is still lower than the demand of 3,000 tons per day in 1995. It exported and imported milk and other products of milk, with the value of 1,438 and 3,081 million Baht, respectively (Department of Livestock Development, 2003). So the value of the im-ported milk was greater than that of the exported milk. In Southern ailand, most of the land is hilly and mountainous. It is largely covered with forests and mangrove forests along the seashore on both sides of the peninsula. In higher areas, rubber is the main crop and in lower areas, rice is grown as the staple food. Large – scale animal production is relatively new to the area; currently, dairy cattle production is undertaken mainly in three provinces : Chumporn, Phatthalung and Songkhla. Previously, farmers in Phatthalung province raised native beef cattle for local consumption, supplied nearby provinces and exported some to Malaysia. In the early 1960s, demand for fresh milk began to rise and some dairy cattle farming was started in 1965. Until 1982, the development of dairy cattle projects was overseen by the ai government (Depart-ment of Livestock Development, 1987). In 1985, a milk processing plant was established which paved the way for the significant development of dairy cattle farming. Farmers who raise dairy cattle live relatively close to one another, and have joined together to form a dairy cooperative. Also, after a milk processing factory was established, it helped the cooperative members get to know each other better. As time went by, they developed self and mutual help approaches. During this process, group interaction took place and leadership emerged. Although dairy cattle projects were launched, the outcomes of these projects were rarely known, especially regarding social group networks for farm information or the farmers’ use of practices recommended by government officers, which was an important aspect in increasing the nation’s milk production. To this assumption, questions which this study are addressed: 1. Are most of the innovators and early adopters socially accepted as opinion leaders by dairy farmers? 2. Are opinion leaders higher in socio – economic status than the followers? 3. Do opinion leaders prefer to be receivers than initiators in personal contact?

CM


24


is study aimed to identify farmers’ systems of dairy cattle produc-tion, the growth and diffusion of dairy cattle raising, opinion leaders and the social group network used for seeking farm information. It also com-pared some characteristics between leaders and non-leaders (followers).

LITERATURE REVIEW

In this study, two existing theories were employed. Firstly, the adoption theory which proposes that behavioral changes of an individual can be

observed by their adoption of recommendations; thus, the adoption of innovations is a function, and varies according to individual characteristics such as education, income, farm size, etc. ose who score highly in such characteristics will adopt recommendations faster than others. us, the adoption of recommended practices can be seen as a function of the socio - economic status of the individual (Rogers, 1983). Secondly, the social acceptance theory which maintains that social acceptance is related to the similarity of members in a group. e more similar the group members are, the greater the overall social acceptance. When a group is formed, communication among group members or the social network shows the direction of their relationship – and the more the group members interact, the greater the speed of diffusion of information. Also, group pressure which has an influence on conformity and deviant behavior of group members is thus lessened. More member-to-member interaction is observed when the group members are similar in terms of their frame of reference. ey must also be socially close and respect each other (Bass, 1983). Rogers and Shoemaker (1971) indicated that communication among members within a group was a function of homophilly. e more similar they were in beliefs, values, education, socio-economic status and so on, the more they would interact with other members. ere was a likelihood that followers would interact with opinion leaders when they were similar, even if those opinion leaders were slightly higher socio-economically, or had a higher education. Dubey and Dwivedi (1972) found that opinion leaders in agricul-ture posed generally higher characteristics than followers in terms of their standard of living, education, farm size, value orientation, aspirations, personal contact with the extension officer, social participations and the adoption of improved farm practices. Bembridge and Burger (1976) also found that when members in a group were similar, group interactions would be more frequent. When disparity among group members was more common, the communication

CM


25


would lessen or even stop. Khadka (1979) found that innovators and early adopters had a higher socio-economic status than followers. Also that followers tended to initiate personal contact with innovators and early adopters, while those innovators and early adopters developed personal contact with extension officers. Pattamarakha (1986) employed sociometric techniques to identify leaders in rice farming in Songkhla province, Southern ailand. It was evident that leaders appointed by the government (the village headmen) were regarded as opinion leaders. Pattamarakha et al. (2003) found that the opinion leader in one group of rubber processers in Songkhla province, was accepted across the group borders by members of the other group.Conceptual model Based on the theoretical framework and empirical research reviewed, the conceptual model was developed. As shown in Figure 1, the model postulates that the socio-economic status of the individuals would result in and be associated with (1) social acceptance by reference groups, and (2) the speed in the adoption of recommended farm practices in dairy cattle production. As a result of social acceptance, opinion leaders emerged. As a result of the adoption of recommended farm practices, innovators and early adopters emerged. Innovators and early adopters may or may not be regarded as opinion leaders (indicated by a dotted line), depending on their reputation. With reference to communication, some opinion leaders may interact with other farmers by initiating, or receiving the personal contacts.

Figure 1: Conceptual model of the study.

Adoption of recommended farm practices

Innovators and early adopters

Communication

Opinion leaders

Socio-economic status ofthe individuals

Social acceptance byreference group

CM


26


e following general propositions are made: 1. e adoption of innovations is dependent on, and will vary according to personal characteristics. 2. ose who adopt recommended practices faster or earlier will have a higher level of social, economic, communicative and psychological characteristics than those who adopt slowly such practices or do so at a later stage. us, from the propositions, the following hypotheses are postu-lated:Hypothesis 1: Opinion leaders will prefer to be receivers rather than initia-tors in personal contact.Hypothesis 2: Opinion leaders will have higher socio-economic status than followers.

RESEARCH METHODOLOGYSelection of the study area

In Southern ailand, a greater number of dairy cattle farms can be observed in Chumporn and Phatthalung provinces. As dairy farmers

in Phatthalung province live closer to one another, plus the fact that Phatthalung has established a dairy processing factory, this province was selected.

Background of the study area Phatthalung province consists of 9 districts, covering an area of 3,424.5 km2. About 80% of the total population are engaged in agricul-ture. Rice growing, rubber plantations, orchards and fishing are the major occupations. About 40% of the cultivated area has access to irrigation.

Population and sampling procedure ere were 96 dairy farmers in the selected area: three had just started, and 93 had raised dairy cattle for more than one year (the three new ones were excluded from the study, giving a total of 93 farmers in the sample.)

Data collection and analysis Personal interviews were assigned as a method of data collection. e questionnaires were then designed and tested to be a precise and accurate tool. e interviews lasted from January, 2003 until March, 2003. Ninety-one farmers were interviewed as two dairy farmers were unavail-able for interviews as quite often they stay with their relatives outside the district.

CM


27


When the field survey was finished, all the questionnaires were checked and found to be completed correctly. Coding and a code book were prepared. e analysis of data was undertaken at the Computer Cen-ter at Prince of Songkla University. e Statistical Package for the Social Sciences (SPSS) was employed as the tool for data analysis. e t-test was used as a means for testing the hypotheses.

Term definition Innovators and early adopters were those who adopted the dairy raising at the beginning when the innovation was introduced. Opinion leaders were those who were socially accepted by a majority group of dairy farmers in disciplines concerned. Leadership identification was undertaken by the socio-metric techniques.

RESULTS AND DISCUSSIONCharacteristics of the respondents

Most of the respondents were male and married. ey had obtained formal education to Grade 4 and were literate. e average age

was 41 years. Some 27.5% were migrants from nearby provinces. Most had a gas stove, refrigerator, color television and motorcycle as material possessions. Raising dairy cattle was regarded as their major occupation (considered as time spent rather than income received). ey also had a minor occupation in rice farming, rubber cultivation, fishery, retailing, or as hired labor. ey received income from their major occupations at an average of 130,193 Baht per family per year. ey had land tenure status as owner, although a few rented their land. About one-third received irrigation water but only 4.4% could access water all year round. Most used family labour for dairy cattle raising and about 12.1% hired temporary external labour. ey had a high level of awareness of the persons engaged in dairy cattle extension, from talking with those concerned. About one-half had paid a visit to the local livestock office while about one-half of them had received a visit from the livestock officer at their farms in return. Personal contact between farmers and livestock officers mainly involved consultation about diseases and their prevention. About one-half received training in production, and approximately one-fourth had gone on study tours.

Production systems As shown in Table 1, most farmers (59.3%) raised Holstein Friesians. is breed gave more milk and was found to be more suited to the local topography. Raising methods were split of three types: feeding inside a shed

CM


28


(47.2%), feeding outside a shed (37.4%), and free-range on the pasture (15.4%). e best raising method was feeding inside the shed, which was best for the prevention of epidemics, pests and problems associated with heavy rainfall. is required a higher initial investment, for construction of the barn, which some farmers tried to avoid. e number of cattle raised varied, but was usually 6-15 head per family, with an average of 11.7. e number of milking cows also varied, but generally ranged from 1-9 head per family with an average of 5.2. Concentrates were provided for each cow according to the amount of milk produced. e pasture plot averaged 11 rai (1 ha = c6.25 rai), which is quite small. Para grass and Plicatulum grass were grown for use as roughage feed. Rice straw, hay, and corn stems were additionally provided in summer to compensate for a shortage of grass. Most followed the recommended practices for internal and external parasite control. About four-fifths applied teat infection medication. Breed-ing was done mostly by artificial insemination (AI). About one – half of farmers (47.3%) used their hands for milking as they had few cows, and this also saved them having to buy milking machines. About one-half of the farmers used machines for milking as they had more cows and the milking machines saved them time and labour. e daily milk production also varied according to feed provisions, ranging from 7-18 litres per head with an average of 11.7 litres. Most (59.3%) received the current selling price of 12-12.5 Baht/litre (1$US = c44 baht) with an average of 12.2 Baht per litre. ey were satisfied with the price and felt that it was fair. A low conception rate of artificial insemination, teat infection and a shortage of grass were mentioned as major problems facing milk production. Most needed additional training in maintaining their pasture plots.

Table 1: Production systems.

Attribute Number (n = 91)

Percentage

Dairy cattle breeds Holstein Friesian (HF) Sahiwal Crossbred HF x Native Crossbred HF x Other breeds

54141914

59.315.420.915.4

Raising methods Feeding inside the shed Feeding outside the shed Free to roam on the pasture

433414

47.237.415.4

CM


29


Herd size (head) 1 – 5 6 – 10 11 – 15 16 – 20 More than 20 Average = 11.7

123030145

13.133.033.015.45.5

Number of milking cows (head) 1 – 2 3 – 4 5 – 6 More than 6 Average = 5.2

15292819

16.531.930.820.8

Milk production (l/cow/day) 8 – 11 12 – 14 More than 14 Average = 11.7

423811

46.241.712.1

Milking methods Hand Machine

4348

47.352.7

Latest selling price received (per litre) 11.01-11.50 Baht 11.51-12.00 Baht 12.01-12.50 Baht More than 12.50 Baht Average = 12.20 Baht

4155418

4.416.559.319.8

Opinion of selling price Not fair Fair

685

6.693.4

Diffusion of dairy cattle adoption Cattle raising in Phatthalung province started in 1973 with one farmer. Seven years later, another one began. Over the following years, the number of farmers who raised dairy cattle grew slowly but steadily. e most rapid increase in the number took place from 1993-1994, largely as a result of the 1991 ailand Revival Plan for Agriculture which allowed farmers in poor areas to obtain low-rate loans. By 2001, the total number was 94 (Figure 2). e main reason why there has been such relatively slow growth in the number of dairy farmers, is probably that this enterprise requires a high initial investment. ose who want to raise dairy cattle must save for many years, or be able to access loan money. According to time to begin raising dairy cattle, the innovators and early adopters were farmers 28, 7, 16, 40, 18, 59 and 10, respectively.


CM


30


Figure 2. Diffusion curve showing adoption of raising dairy cattle.

Leadership in social acceptance by reference group (opinion leaders)

Farmers were asked to note the names of the best farmers in situations concerning (1) knowledge in dairy cattle production, and (2) consultancy. As shown in Table 2, farmers 40 and 10 received the highest number of first mentions. Farmer no. 7 was also mentioned as second priority in knowledge of dairy cattle raising, human relations and as a consultant. From the two situations, those who were mentioned would be regarded as opinion leaders. Farmers 7, 9, 10, 21 and 40 were identified as opinion leaders through the socio-metric techniques. ere was a close relationship between knowledge of cattle raising (situation 1) and consultants (situation 2). at is, those who were regarded as having a good knowledge of cattle raising, would be those who had a very important role as the consultants. In addition, of the five opinion leaders, three (no. 7, 10 and 40) were also regarded as innovators and early adopters.

Table 2: Perceived leaders in situations concerning dairy cattle production.Situation

Rank Knowledge of cattle raising Consultant

1 40(15) 10,40(9)*2 7(12) 7(5)3 10(8) 9,21(3)

*Figures in brackets indicate the number of voters

Social network in seeking dairy cattle information Dairy farmers were asked to mention the names of persons who had come to visit and consult them about dairy cattle production. As shown in Figure 3, all opinion leaders (no. 7, 9, 10, 21 and 40) received personal

0

20

40

60

80

100

73 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 00 01

Cum

ulat

ive

num

ber o

f dai

ry fa

rms

Year

CM


31


contact as consultants. Not one opinion leaders initiated personal contact for seeking dairy cattle information. is supported Hypothesis 1-that opinion leaders would prefer to be receivers than initiators in personal contact. One reciprocal contact took place between farmers 17 and 22. is occurred when they were very similar in social and cultural settings such as, in values or attitudes. Other dairy farmers also received personal contact as consultants but to a lesser extent. e implication for this was that opinion leaders might be used as a channel for an information delivery system in dairy cattle production. is will help support the development of dairy cattle raising of the public concerned as the number of extension officers is too low in relation to the number of target persons who raised dairy cattle.

– Opinion leadersFigure 3. Social network used in seeking dairy cattle information.

Opinion leaders and socio – economic status Socio – economic status is the composite index of (1) farm size (2) farm income (3) status of land tenure (4) education (5) social participation and (6) urban contact. is index is of equal weight. us, the total score

CM


32


of this index is six. As shown in Table 3, opinion leaders had a higher socio - economic status than followers, which was statistically significant (t = 4.648, p = 0.000). is supported Hypothesis 2 that opinion leaders would have a higher socio - economic status than the followers. Also, opinion leaders who were higher in socio – economic status would prefer to be receivers rather than initiators in personal contact with followers. is study was similar to that of Rogers and Shoemaker (1971) in which followers were found to be to interact with opinion leaders who were of higher social status, or were more technically advanced.

Table 3. Comparisons of socio - economic status between leaders and followers.

Average score t P

Leaders ( n = 5 ) Followers ( n = 86 )3.40 2.63 4.648 0.000

CONCLUSIONS AND RECOMMENDATIONS

Most farmers raised Holstein Friesian dairy cattle, and feeding was undertaken in a shed. Rice straw, hay or corn stalks were additionally

provided in the summer to compensate for a shortage of grass. Breeding was mostly undertaken by artificial insemination and milking was done by hand or machine. With regard to leadership characteristics, five dairy farmers were regarded as opinion leaders by socio - metric techniques. ree opinion leaders were also regarded as innovators. For seeking dairy cattle information, all opinion leaders received per-sonal contact as consultants in giving dairy cattle information at home. ey preferred to be receivers rather than initiators in giving farm information. Of the hypotheses tested, it was found that (1) opinion leaders preferred to be receivers rather than initiators for providing farm information and (2) opinion leaders had higher socio-economic status than followers.

e following steps are recommended: 1. For delivery of farm information to target persons, the extension officers should be active and initiative in providing the information to those opinion leaders rather than to farmers individually. 2. Farmers 7,10 and 40 who played a key role as opinion leaders should be used as catalytic agents to improve the low adoption rates of some recommended practices, especially in pasture plot and health prac-tices. In addition, as they had a very important role as consultants for farm information, these leaders should be used as a key channel for the

CM


33


dissemination of information to the dairy farmer group, which will reduce the routine work of the livestock extension officers. 3. As a number of farmers who raised dairy cattle is limited, expan-sion of a number is obvious in order to expand the dairy industry and meet the market demand. 4. Locality groups should be established to develop a relationship among members within a group. Linkage of locality groups should be encouraged to provide farm information on dairy cattle. 5. Farmers had a very small size of pasture, averaging 11 rai, making it impossible to provide roughage all year round, especially in the hot, dry summer months. All farmers should take steps to prepare for this shortage, using supplements as necessary. 6. Most farmers needed intensive training in pasture management, and agencies concerned should organize training courses, as this was as-sociated with the shortage of roughage. 7. A low conception rate of AI was observed, thus, farmers should repeat with care by observing and mating the cattle heat period, and im-mediately contact the livestock officer for support without delay.

ACKNOWLEDGEMENTS

The authors wish to thank Mr David Patterson and Miss Maria Bow-ering, English teachers at Prince of Songkla University, who assisted

with editing this paper.

REFERENCESBass, B. M. 1983. Some observations about a general theory of leader-

ship and interpersonal behavior, pp. 64-79. In R.L.William and R.F.Richard (eds.) Leadership and Social Change. 3d ed.. California:University Associates.

Bembridge, T.J., and P.J. Burger. 1976. e importance of opinion leaders in extension. South African J. of Ag. Ext. 5 : 13-17.

Chantalakhana, C. 1995. e development of dairy cattle production and milk products in ailand : strategies for research and development in the future. Office of Foundations for Research Support, Bangkok.

Department of Livestock Development. 1987. Development of dairy cattle production in ailand. Bangkok.

Department of Livestock Development. 2003. Data-base information. Bangkok.

Dubey, V.K., and N.N. Dwivedi. 1972. Identification of opinion leaders and their characteristics. Indian J. of Ext. Ed. 8(2) : 76-83.

CM


34


Khadka, S.S. 1979. A study of factors associated with diffusion and adoption of high yielding rice varieties in Nawalparansi district of Napal. Master’s thesis, Department of Agriculture, University of Queensland. (Unpublished)

Pattamarakha, K. 1986. Differential characteristics of adopters and non-adopters of high-yield varieties of rice. Songklanakarin J. of Sci. & Tech. 8(2) : 145-153. (in ai)

Pattamarakha, K., and B. Somboonsuke. 2003. Group interaction, participa-tion and leadership : comparison between successful and unsuceessful rubber sheet improvement and sales groups in southern ailand. CMU J. 2(1) : 49-60.

Office of Agricultural Economics. 1996. Agricultural statistics of ailand Crop Year 1995-96. Bangkok.

Phatthalung Provincial Office of Livestock Development. 1995. Informa-tion about dairy cattle production in the 9th Regional Livestock Zone. Phatthalung.

Rogers, E.M. 1983. Diffusion of Innovations. 3d ed. New York : e Free Press.

Rogers, E.M., and F.F. Shoemaker. 1971. Communication of Innovations. 2d ed. New York : e Free Press.

Samnakhon, A. 1992. Communication exposure and adoption of innova-tion affecting the accomplishment of farmers’ dairy production in Changwat Petchaburi between members of dairy cooperative and non-members. Master’s thesis, Kasetsart University. (Unpublished)

CM


35


M. N. A. Bhuiyan1*, Kazi Saleh Ahmed2 and Roushan Jahan1

Projection of GDP of Manufacturing Industries in Bangladesh: An Approach to Growth Model

1Department of Mechanical Engineering, Tokyo Metropolitan University, 1-1, Minami-Osawa, Hachioji-shi, Tokyo 192-0397, Japan. 2Department of Statistics, Jahangirnagar University, Savar, Dhaka-1342, Bangladesh. *Corresponding author. E-mail: [email protected]

ABSTRACT

In this paper, forecasting of the GDP of Manufacturing Industries in Bangladesh is made for the future thirteen years, based on the

last twenty three years data, by using growth model. To select the best growth model, some statistical tools, e.g., R2, R 2, AIC, BIC, AME, RMSE and MAPE are used. Quadratic type of growth model is found to be the appropriate model for forecasting on the basis of these criteria. Nowadays, time series data have been frequently and intensively used in empirical research. In our present study, time series data are used to make forecast. Forecast line shows the slowly-increasing quadratic trend for the GDP of Manufacturing Industries (Total, Large and Medium Scale and Small Scale) for the next thirteen years. Key words: Forecast, Growth model, R2, R 2, AIC, BIC, AME, RMSE, MAPE, Quadratic model, Empirical.

INTRODUCTION

Industrial sector accounts for a significant growth to the national economy. It is a very important potential sector of Bangladesh. is sector plays

a very important role in creating foundation and modernization of other sectors. Manufacturing sector creates opportunity of employment which helps to reduce poverty. Manufacturing combines human ingenuity and technology innovation with natural endowments of a nation. A country is said to be industrialized if 30% of the total GDP comes from the manu-facturing sector. Chowdhury and Bhuiyan (1985), in their study sought to profile the changes in the levels and structure of wages in Bangladesh’s large-scale industries in the post-liberation period. It was found that real wage in large-scale industries essentially stagnated between 1972/73 and 1981/82, according to Bangladesh Bureau of Statistics (BBS) data. How-ever, real wages in public sector industries are found to have grown, albeit

CM


36


at slow rates. Here, projection has been made using extrapolation method and using geometric growth rate y = ert model for projection. Granger et al (1995) use AIC and BIC criteria for model selection rather than testing of hypothesis at the time of deciding model specification. is is because an arbitrary choice of significant level, the null hypothesis is favored to test and from the same data set can end up with different final models to the researchers. Granger (1980) derived the approach of fitting and evaluating technique of curves or models. e basic steps in this approach are (i) the choice of a set of possible curves or models (ii) the estimate of the parameters of the model and (iii) the use of a criterion to evaluate the relative goodness of fit of the models. Here, the forecasting of the U.S. personal consumption is made on the basis of linear, exponential, modi-fied exponential and parabolic (quadratic) model. Paul (1998) works with “Modeling and forecasting the energy consumption in Bangladesh”. In his paper, he selected at first three models on the basis of largest R2 and R2 and then selected a final model for forecasting based on the minimum value of AME, RMSE and MAPE. He made forecast by dividing the total period into three parts, namely, estimation period, validation period and forecast period. Pindyck and Rubinfeld (1976) discussed about simple extrapola-tion models and also made forecast of monthly retail sales of department stores. e time series data are used to make forecast in their study. ey applied linear trend model, logarithmic linear trend model (exponential growth), autoregressive trend model and logarithmic autoregressive trend model. ey also suggested to calculate forecast based on quadratic model. Hurvich and Tsai (1989) derived one of the leading selection methods, AIC, for regression and autoregressive time series models which are particularly used when the sample size is small or the number of fitted parameters is a moderate to large fraction of the sample size. e minimum-AIC criterion produces a selected best possible model. is paper reveals the most suitable growth model for forecasting the GDP of Manufacturing Industries, including large and medium scale and small scale industries for future periods beginning from 2002-2003. Secondary sources of data published from Bangladesh Bureau of Statistics (BBS) are used in this study.

MODEL DISCUSSION

The models which are used to describe the behavior of some variables, as they vary with respect to time, are termed as growth models. is

type of model is needed in a specific area and in a specific problem, depending on the type of growth that occurs in the time series data. In general, growth models are mechanistic, rather than empirical ones. A

CM


37


mechanical model usually arises as a result of making assumptions about the type of growth, writing down the differentials or difference equations that represent these assumptions and then solving equations that represent these assumptions and to obtain a growth model. An empirical model, on the other hand, is a model chosen to empirically approximate an unknown mechanistic model. Typically, the empirical model is a polynomial of some suitable order.

Analysis of Growth Model In this study, eight types of growth model: linear, quadratic, cubic, logarithmic, exponential, power, inverse and logistic are fitted by using the software SPSS 10.0. From these models, only three models: quadratic, cubic and exponential are primarily selected on the basis of high R2, R 2. Finally, the best model is selected from these three under the criteria that the model has largest Coefficient of Determination (R2) and Adjusted Coefficient of Determination (R 2) and smallest Akaike Information Cri-teria (AIC), Bayesian Information Criteria (BIC), Absolute Mean Error (AME), Root Mean Square Error (RMSE) and Mean Absolute Percent Error (MAPE).

Estimates of the Selected Models e estimates of the selected models are given below:Quadratic ModelFor GDP of Manufacturing Industries, Bangladesh Yt =94720.446076+1545.685827 t+411.024619 t2

For GDP of Manufacturing Industries (Large and Medium Scale), Ban-gladesh Yt =67207.228120+1065.613552 t+295.793111 t2

For GDP of Manufacturing Industries (Small Scale), Bangladesh Yt =27513.217956+480.072276 t+115.231508 t2

Cubic ModelFor GDP of Manufacturing Industries, Bangladesh Yt =98485.821570-158.387954 t+584.81118 t2-4.827404 t3

For GDP of Manufacturing Industries (Large and Medium Scale), Ban-gladesh Yt =70703.718238-516.772361t+457.169578t2-4.48268 t3

For GDP of Manufacturing Industries (Small Scale), Bangladesh Yt =27782.103331+358.384407 t+127.641602 t2-.344725 t3

ˆ

ˆ

ˆ

ˆ

ˆ

ˆ

CM


38


Exponential ModelFor GDP of Manufacturing Industries, Bangladesh Yt =84447.119001 e.060955 t

For GDP of Manufacturing Industries (Large and Medium Scale), Ban-gladesh Yt =59794.616409 e.061292 t

For GDP of Manufacturing Industries (Small Scale), Bangladesh Yt =24648.103963 e.060131 t

Model Selection e criteria by which a model is selected as the best one for these three categories are shown in the following tables:

Table 1: Criteria for GDP.Criteria Quadratic Model Cubic Model Exponential Model

R2 .99640* .99660 .99235R2 .99604* .99607 .99199

AIC 459.9250989* 460.7911902 470.669992BIC 463.3315815* 465.333167 472.9409804AME 3899.361739* 3917.373913 5014.627391

RMSE 4970.644215* 4985.135133 6448.141199MAPE 2.112328314* 2.219061614 2.833078668

(e value of the criterion for a model with starlet shows that the model is better than other two models with respect to that criterion).

From the above table, it is observed that Quadratic model is the best among all of the criteria.

Table 2: Criteria for GDP (L&M Scale).Criteria Quadratic Model Cubic Model Exponential Model

R2 .99553 .99586* .99092R2 .99508 .99521* .99048

AIC 449.5742902* 449.959011 502.2555508BIC 452.9807729* 454.5009879 504.5265392AME 3066.882174* 3106.347826 3961.914348

RMSE 3927.752178* 3980.641277 5003.791MAPE 2.336757967* 2.477916941 3.131493068


ˆ

ˆ

ˆ

CM


39


Table2 shows that Cubic model is fractionally better than Quadratic model in case of R2 and R2 but all other criteria suggest Quadratic model as the best model. So Quadratic model is selected to forecast the GDP of Manufacturing Industries (L&M Scale) based on above criteria.

Table 3: Criteria for GDP (Small-Scale).Criteria Quadratic Model Cubic Model Exponential Model

R2 .99628 .99629* .99396R2 .99591* .99571 .99367

AIC 402.8530429* 404.954797 411.1908837BIC 406.2595256* 409.496774 413.4618722AME 1126.017391* 1142.92435 1295.705652

RMSE 1441.614466* 1476.55965 1769.625266MAPE 2.151892391* 2.21419517 2.485355458


It is shown in Table 3 that only R2 suggests Cubic model but all other criteria suggest Quadratic model as the best model. So in this case, Quadratic model is finally selected as the best one. Of all the three cases, it is found that Quadratic model is the one which exhibits better than the other two models. us, forecasting is performed on the basis of this selected model.

RESULTS AND DISCUSSION

The following tables, representing the estimated values and the forecast values, are divided into two parts, namely, estimation period and

forecasting period. After the estimation period, forecasting is made for the next thirteen years based on the selected model in each of the three cases of Manufacturing Industries in Bangladesh.

CM


40


Table 4: Forecasting for GDP of Manufacturing Industries in Bangladesh.Year GDP Quadratic model

Estimation Period1979-80 96311 96677.156521980-81 100523 99455.916211981-82 101783 103056.72511982-83 105272 107479.58331983-84 114664 112724.49071984-85 121030 118791.44731985-86 129882 125680.45321986-87 140210 133391.50831987-88 141118 141924.61261988-89 145029 151279.76621989-90 156163 161456.96911990-91 166133 172456.22111991-92 178391 184277.52241992-93 193771 196920.87301993-94 209554 210386.27271994-95 231517 224673.72171995-96 246351 239783.22001996-97 258795 255714.76751997-98 280908 272468.36421998-99 289882 290044.01021999-00 303679 308441.70542000-01 331308 327661.44982001-02 340176 347703.2435

Forecast Period2002-03 368567.08642003-04 390252.97852004-05 412760.91992005-06 436090.91062006-07 460242.95042007-08 485217.03952008-09 511013.17792009-10 537631.36542010-11 565071.60232011-12 593333.8883

CM


41


2012-13 622418.22362013-14 652324.60812014-15 683053.0419

e following figures show the forecast values of GDP of Manufac-turing Industries in Bangladesh.

Fig 1. Forecast GDP in Mill.Tk.* of Manufacturing Industries in Bangladeshfrom 2002-03 to 2014-15 by using Quaratic type of Growth Model.

In Fig1, GDP is plotted in respect of time period. It shows the future trend of the GDP of Manufacturing Industries in Bangladesh. Every year, the contribution of this sector is increasing but not at a satisfactory level. e projected value shows a quadratic trend but slowly increasing for the mentioned future period.

Table 5: Forecasting for GDP of Manufacturing Industries (L&M Scale) in Bangla-desh.

Year GDP Quadratic modelEstimation period

1979-80 69015 68568.634781980-81 71914 70521.627671981-82 71708 73066.206781982-83 73572 76202.372111983-84 81157 79930.12366

Fig1. Forecast GDP in Mill.Tk.* of Manufacturing Industries in Bangladesh from 2002-03 to 2014-15 by using Quadratic type of Growth Model

y = 411.02x2 + 1545.7x + 94720

R2 = 0.9964

0

100,000

200,000

300,000

400,000

500,000

600,000

700,000

800,000

1979

-80

1982

-83

1985

-86

1988

-89

1991

-92

1994

-95

1997

-98

2000

-01

2003

-04

2006

-07

2009

-10

2012

-13

Year

GD

P in

Mill

. Tk.

GDP(Obs.)

Forecast GDP

Trend

*1 US$=67 Taka

CM


42


1984-85 85507 84249.461431985-86 92100 89160.385431986-87 99882 94662.895651987-88 100576 100756.99211988-89 102657 107442.67481989-90 111126 114719.94361990-91 117817 122588.79881991-92 126462 131049.24011992-93 137846 140101.26761993-94 149220 149744.88141994-95 166297 159980.08141995-96 175732 170806.86761996-97 182704 182225.24011997-98 199668 194235.19881998-99 208033 206836.74361999-00 217086 220029.87482000-01 238252 233814.59212001-02 240554 248190.8957

Forecast period2002-03 263158.78542003-04 278718.26142004-05 294869.32372005-06 311611.97212006-07 328946.20682007-08 346872.02772008-09 365389.43482009-10 384498.42812010-11 404199.00772011-12 424491.17352012-13 445374.92552013-14 466850.26372014-15 488917.1881

e above table for forecast GDP of Manufacturing Industries (L&M Scale) can be represented by the following figure.

CM


43


Fig 2. Forecast GDP in Mill.Tk. of Manufacturing Industries (L&M Scale) in Bangladesh from 2002-03 to 2014-15 by using Quadratic type of Growth model.

Fig 2 depicts the past and future condition of the GDP of Large and Medium Scale Industries. is figure scratches the picture of the GDP of Large and Medium Scale irrespective of time. e past contribution of this sector is very low but year to year the share of this sector is getting better. According to the above figure it can be conclude that the forecast value shows a quadratic trend for the aforementioned periods.

Fig2. Forecast GDP in Mill.Tk. of Manufacturing Industries (L&M Scale) in Bangladesh from 2002-03 to 2014-15 by using Quadratic type of Growth model

y = 295.79x2 + 1065.6x + 67207

R2 = 0.9955

0

100,000

200,000

300,000

400,000

500,000

600,000

1979

-80

1982

-83

1985

-86

1988

-89

1991

-92

1994

-95

1997

-98

2000

-01

2003

-04

2006

-07

2009

-10

2012

-13

Year

GD

P in

Mill

. Tk.

GDP(L&M) (Obs.)

Forecas t GDP

Trend

CM


44


Table 6: Forecasting for GDP of Manufacturing Industries (Small-Scale) in Bangla-desh.

Year GDP Quadratic modelEstimation period

1979-80 27296 28108.521741980-81 28609 28934.288541981-82 30075 29990.518351982-83 31700 31277.211181983-84 33507 32794.367021984-85 35523 34541.985881985-86 37782 36520.067761986-87 40328 38728.612651987-88 40542 41167.620551988-89 42372 43837.091471989-90 45037 46737.025411990-91 48316 49867.422361991-92 51929 53228.282331992-93 55925 56819.605311993-94 60334 60641.391301994-95 65220 64693.640321995-96 70619 68976.352341996-97 76091 73489.527391997-98 81240 78233.165441998-99 81849 83207.266521999-00 86593 88411.830602000-01 93056 93846.857712001-02 99622 99512.34783

Forecast period2002-03 105408.30102003-04 111534.71712004-05 117891.59632005-06 124478.93852006-07 131296.74362007-08 138345.01192008-09 145623.74312009-10 153132.93732010-11 160872.59462011-12 168842.7149

CM


45


Forecast period2012-13 177043.29812013-14 185474.34442014-15 194135.8538

e forecast value of the GDP of Small-Scale Industries can also be shown in the following figure.

Fig 3. Forecast GDP in Mill.Tk. of Manufacturing Industries (Small-Scale) in

Bangladesh from 2002-03 to 2014-15 by using Quadratic type of Growth model.

Fig 3 discloses the overall picture of the GDP of Small-Scale In-dustries for the estimation period and forecast period. From 1979-80 to 1991-92, it can be seen that the GDP was linear and after then it has been changed to curvilinear. So we conclude for the total period that the figure exhibits a quadratic trend. ough Small Scale shows quadratic trend but the increasing portion in every future year is very slow.

CONCLUSION

Our results indicate the future movement of Manufacturing Indus-tries in Bangladesh, starting from 2002-2003. ough performance

of this sector was not satisfactory during the post liberation (1971) in Bangladesh but it is increasing day by day. The share of the manu-facturing sector limped from 9.0 percent in 1973 to a mere 11.4 per cent in 1994-95. In 1994/95, this sector achieved a growth rate of 8.6 per cent compared to 7.2 per cent in 1989/90. Over the period from 1990/91 to

y = 115.23x2 + 480.07x + 27513

R2 = 0.9963

0

50,000

100,000

150,000

200,000

250,000

1979

-80

1982

-83

1985

-86

1988

-89

1991

-92

1994

-95

1997

-98

2000

-01

2003

-04

2006

-07

2009

-10

2012

-13

Year

GD

P in

Mill

.Tk.

GDP(Small-Scale) (Obs.)

Forecast GDP

Trend

CM


46


1995/96, the manufacturing sector grew at an average annual rate of 6.3 per cent, and there have been fluctuations also. e first year of the fourth plan registered the lowest growth rate (2.39 per cent), while the highest growth rate (9.1 per cent) was achieved during the year 1992/93. e growth rates during 1995/96 and 1996/97 were 5.3 per cent and 3.3 per cent, respectively. Among the industrial sub-sectors, ready-made garments (RMG) and pharmaceuticals recorded the highest growth performances (greater than 10 per cent), while growth rates of jute and cotton textiles declined. To take strong courses of action about manufacturing sector it is necessary to analyze the future condition based on the past. It is also important to identify the obstacles which are the causes of less contribution of manufacturing industries in the national economy and strictly handle this sector by formation of some rules. To do this, forecasting plays a sig-nificant role for getting the main feature of this sector. is paper suggests that all the three cases of manufacturing industries show an upward trend but it is slowly increasing. Forecasting helps the policymakers and planners to improve this sector so that Manufacturing Industries in Bangladesh can contribute a remarkable portion in the national GDP and meet the target in the near future. In future, emphasis should be put on larger data system for better statistics.

REFERENCESBBS, Statistical Year Book of Bangladesh, 1980 to 2000, Bangladesh

Bureau of Statistics (BBS)BBS, National Accounts Statistics of Bangladesh, (Revised Estimates,

1989-90 to 1998-99). BBS, National Accounts Statistics (Gross Domestic Product, 2000-

2001). BBS, National Accounts Statistics (Preliminary Estimates of GDP, 2001-

2002). Chowdhury Nuimuddin and Karimullah Bhuiyan.1985. Wages in Bangla-

desh industries1972/73-1981/82: Levels and structure. BDS Journal, Vol.-XIII, No. 2.

Granger, C.W.J. 1980. Forecasting in business and econometrics. Academic Press, Inc. New York.

Granger, C.W.J., M.L. King, and H. White.1995. Comments on testing economic theories and the use of model selection criteria. Journal of Econometrics, Vol. 67,173-187.

Hurvich, Clifford M., and Chih-Ling Tsai.1989. Regression and time series model selection in small samples. Biometrika, Vol. 76, No. 2, 297-307.

CM


47


Paul, J.C. 1998. Modeling and forecasting the energy consumption in Bangladesh. Ph.D. esis, Department of Mathematics, University of North Bengal, West Bengal, India.

Pindyck, R.S., and D.L.R. Rubinfeld. 1976. Econometric models and economic forecasts. McGraw-Hill, New York, Second Edition.

CM


49


Yingyot Chiaravutthi*

e Existence and Rationale of Predatory Pricing Strategy in the Network Industry: An Experimental Approach

Mahidol University International College, 999 Buddhamonthon, Salaya, Nakhonpathom 73170, ailand*Corresponding author. E-mail:[email protected]

ABSTRACT

This paper reports data from an experiment confirming the exis-tence of predatory pricing in the presence of network externali-

ties when the technology of competing firms are identical. An experiment was conducted at the University of South Carolina’s Beam Lab, based on a seven-period and one-market design. Twenty-two subjects were recruited and assigned a role as seller. ey were paired together randomly and played the game for seven rounds. Pairs of subjects were re-matched between each round to avoid repeated game effects. Buyers were simulated and had different, pre-programmed preferences over the goods offered by two sellers. In each period, subjects had to make an entry decision and a price decision in a posted offer market institution. e sub-game perfect Nash equilibrium is for sellers to engage in predatory pricing during the first four periods and price higher during the last three periods of a round. e observations strongly support the presence of predatory pricing, especially in later rounds. Another finding is that price offers during the last three periods were lower than predicted by sub-game perfect Nash equilibrium due to the fact that the predatory pricing strategy was not effective in eliminating competitors.

Key words: Predatory pricing; Network externalities; Entry deterrence; Posted offer market

INTRODUCTION

Predatory pricing is defined as price cutting in the short run in order to eliminate competitors or deter potential entrants, in order to realize

monopoly profits in the long run.1 Some economists reject the existence

1is definition is different from the legal definition proposed by Areeda and Turner (1975) which is pricing less than short-run marginal cost (or average variable cost when data on marginal cost is unavailable).

CM


50


of predation on the rationality that predatory pricing would have to be followed by a period of setting high prices for the purpose of recoupment, and the recoupment period would encourage entry again. Nevertheless, the predatory pricing strategy could be more creditable under certain circum-stances. For example, in dynamic industries with network externalities, it has been thought by economists that predation was theoretically possible. e topic of network externalities has been studied for almost two de-cades, but has received more attention recently because many “high tech” products seem to exhibit some kinds of network externalities. A product creates direct network externalities2 when the utility that an agent derives from consuming the product increases with the number of other agents consuming the same product. Katz and Shapiro (1986) and Farrell and Saloner (1986) analyze pricing strategies employed by firms for the purpose of expanding network size. Katz and Shapiro (1986) discuss consumers’ adoption decisions of a new technology when facing firms’ pricing strategies in a two-period model. Farrell and Saloner (1986) mention the possibility of predatory pricing by an incumbent as a means to deter entry in the presence of network externalities. e model assumes that potential users arrive over time. ey show that the incumbent can induce new users to stay with its technology by reducing the price and incurring a short-run loss. When the rival is kept out of the market, the incumbent monopolist can raise its price without inducing reentry because it has a substantial advantage of installed base. While some economists contend that predatory pricing is a myth, others argue the opposite. e reason lies partly on the inaccessibility of the field data. Because the field data are hard to find, experiments are an attractive alternative. Isaac and Smith’s (1985) is the first study to search for predatory pricing in the laboratory. In their duopoly setting, a large firm (seller A) has an advantage in terms of an initial cash endowment and lower marginal cost over the smaller firm (seller B). In each period, each seller selects a price and a maximum supplied quantity correspond-ing to the price. Both sellers A and B make their decisions without any information regarding each other’s cost structures. On the demand side, buyers are computer- simulated. Isaac and Smith define predatory pricing as the fulfillment of the following three conditions:

2e term “network externalities” used in this paper means direct network externalities as opposed to indirect network externalities. A (group of ) product(s) exhibits indirect network externalities when the utility is derived directly from consumption of a durable good (hardware) and indirectly from a set of complementary compatible goods (software).

CM


51


1. PA(QA) < MCA(QA) and PA < min ACB(QB) where Pi, Qi, ACi, and MCi are firm i’s price, quantity supplied, average cost and marginal cost, respectively. 2. Quantities offered by firm A may not leave any excess demand for firm B to satisfy at a higher price. 3. Firm B exits and does not enter even when firm A subsequently raises its price. Predatory pricing, according to this definition, was not observed in Isaac and Smith’s experiment. Specifically, they found “zero” evidence in all of their experiments, even after including a sunk entry cost, complete information about competitors’ costs and additional predatory incen-tives. Harrison (1988) reclassifies the definitions of predatory pricing and introduces a multi-market design. He defines four types of predatory pricing. Harrison separated type I predatory pricing into “strong” and “weak,” depending on the quantity offered by seller A. Isaac and Smith’s definition of predatory pricing corresponds to strong type I predatory pricing while weak type I predatory pricing only requires fulfillments of the first and third conditions of Isaac and Smith’s definition. As a result, seller A’s weak type I predatory pricing leaves a residual demand to be satisfied by seller B. Harrison’s type II predatory pricing allows seller A to price “below” the predatory price range without any quantity constraint while weak type I predatory pricing requires seller A to price “within” the predatory price range. e non-negative profit constraint is applied to strong type I, weak type I and type II predatory pricing. Type III predatory pricing occurs when seller A sets its price and quantity in such a way as to result in a net loss for that period (no non-negative profit constraint). Type IV predatory pricing focuses on when the small seller, seller B, sets its price below the minimum average cost of seller A. Predatory pricing simply could not be found in Isaac and Smith’s design by any of Harrison’s four definitions. Harrison argued that seller A could not successfully engage in predation (or could not prevent seller B from reentry) because there was no “escape opportunity” for seller B. In Isaac and Smith’s design, seller B would have nothing to do if he decided not to purchase an entry permit. As a result, Harrison introduced a major modification: a multi-market nature of the experiment which provides an alternative revenue option for seller B. ere were five markets with iden-tical but independent demand curves. e roles of sellers A and B were assigned to four and seven subjects, respectively. At the beginning of each period, both sellers A and B were required to make decisions regarding which one of the five markets they wanted to enter, and simultaneously

CM


52


to offer a price and quantity in that market. Each seller A was implicitly assigned to one of the five markets. Seller A was informed that if he chose a market other than his own pre-assigned market, his cost schedule would be that of a seller B. is information also served as a “hint” regarding seller B’s cost structure. After all decisions had been made in each period, prices along with subjects’ identification numbers were publicly written on the blackboard for all to see. is allowed reputations to develop. Buyers were simulated. From 132 observations (11 subjects in 12 periods), 9, 3, 1 and 2 pricing decisions consistent with type I, type II, type III and type IV predatory pricing cases were observed, respectively Rutström (1986) attempted to test whether either the experience of sellers A in contested multiple market designs or the clear alternative revenue options given to sellers B could explain the difference in findings between Harrison and Isaac and Smith. In order to do so, she introduced three variations to the Isaac and Smith design: a constant opportunity cost of seller B’s market participation,3 the experience level of seller A in unregulated monopolies and a 37 percent increase of the single period net gain from predation compared to that in Isaac and Smith’s (the difference between profits from monopoly pricing and the dominant firm outcome which is forgone profits from predatory pricing). Again, predatory pricing was not observed. Harrison’s five-market setup was replicated in Goeree and Gomez’s (1998) first series of experiments. In the first series, the authors found only 3 decisions (out of 144 pricing decisions) which were in the range of Harrison’s type II predatory pricing. In their second series of experi-ments, they modified Harrison’s setup in three important aspects. First, sellers B announced their entry decisions and these decisions were made public before all sellers A and B selected prices and quantities in order to reflect the common market fact that a pricing decision can be changed quicker than entry decisions. is would also allow seller A to make a better pricing decision. e second aspect involved a simplified three-step demand schedule.4 ere were only three markets with one seller A in each of markets I and II, and four sellers B who could enter any of the three markets. ird, seller A had complete information about seller B’s cost structure, but not the other way around. Predatory pricing existed

3She created an “alternate market” for seller B. If seller B dicided not to participate in a particular market, it would earn a certain amount of money (which was slightly below the competitive equilibrium’s profit).4e original Isaac and Smith’s demand schedule consisted of 20 steps (20 buyers’ valuations. e three-step demand schedule consisted of 12 buyers but only 3 steps: buyers 1 through 6 had high valuations, buyers 7 through 10 had the second highest valuations, and buyers 11 and 12 lowest valuations.

CM


53


in this series of experiments. e three-market design and three-step de-mand were also replicated and explained in Capra et al. (2000) with the focus on adapting this experiment for the purpose of teaching classroom students. Also see Jung et al. (1994) for a more sophisticated rational theory of predation by allowing “reputation” to occur in their sequential-game design. Predatory pricing may not commonly be observed both in the real world and even in the laboratory. Nevertheless, previous studies have not considered the presence of network externalities. To create the environ-ment of two firms competing to be the dominant technology, I modify and extend the model tested here in four important respects. First, sellers start out from a symmetric situation in terms of costs and buyers’ valuations. Second, the experiment consists of a longer time horizon with seven discrete periods rather than two. ird, goods are not homogeneous which means that buyers’ relative valuation for the two goods is distributed uniformly and symmetrically. And fourth, network externalities accrued through a continuous process of market share acquisition, rather than being a discrete “winner-takes-all” effect.

MATERIALS AND METHODSExperimental Design

A single market lasts for seven periods. Two sellers, A and B have the same cost structure but sell differentiated goods. Both sellers have complete

and symmetric information about each other’s marginal cost. e cost of producing each unit is $1.00 for both sellers A and B, and for all periods. ere is no production constraint for both sellers, so they always meet demand. In the first period, both sellers have to make a pricing decision which is governed by a posted-offer institution. At the beginning of the second period and all subsequent periods up to the seventh period, both have to decide whether they want to enter that period. If they decide to enter, they must pay a fee of $5. ere is no entry fee in the first period, so it can be considered sunk. In fact, there is not even an entry decision to be made, but both sellers are automatically participating in this first period. e entry decision is made public before they are asked to post prices. ere is no constraint on losses in any period. All subjects receive a $5 initial endowment which is used to cover initial losses. Any losses over the initial endowment are forgiven.5

5ere was only one subject who had cumulative losses from all six rounds in excess of $5.

CM


54


Computer-simulated buyers have different preferences over the goods offered by the two sellers. ere are 5 buyers in each period, and each buyer values seller A’s and B’s products as shown in the second row of Table 1. For example, at the first period when no seller makes any sale, buyer 1 is willing to pay $1.00 for seller A’s product but is willing to pay $5.00 for seller B’s product while buyer 5 values product A more than product B. Buyers make purchase decisions by comparing the surplus they would get from buying either of the two goods.6 e surplus is calculated as the difference between the buyer’s willingness to pay and the price posted for the good. If the surplus from purchasing either product is equal, the buyer’s decision will be randomly determined. Buyers are limited to buying only one good each. Table 1: Examples of Buyers’ Willingness to Pay with Different Cumulative Units

Sold.

If the cumulative units sold by seller A or seller B is

en

0

WTP for A WTP for BBuyer 1 1 5Buyer 2 2 4Buyer 3 3 3Buyer 4 4 2Buyer 5 5 1

1

WTP for A WTP for BBuyer 1 1.5 5.5Buyer 2 2.5 4.5Buyer 3 3.5 3.5Buyer 4 4.5 2.5Buyer 5 5.5 1.5

6It is assumed that buyers in the first period have no information regarding which technology will dominate in the future, and which technology will be chosen by future-period buyers. is is possible when both technologies are identical.

CM


55


2


If A has sold 2 units while B has sold 4 units


Note: Subjects are presented with the current WTP for both products. For example, if seller A has sold 2 units and seller B has sold 4 units cumulatively in period 1 and 2, both seller A and B will be presented with the information similar to the last row of Table 1 (not the whole table) in period 3.

ere are seven generations of new buyers who make independent decisions: one generation for each round. e products offered by the same seller across different periods are compatible with each other, while the products offered by different sellers are not compatible with each other. As a result, if a seller can sell more units than another seller, additional buyers’ valuations of the product with more units sold will be higher due to the benefits from product compatibility but the buyers’ valuations of the other good are unaffected. is is a direct network externality or positive consumption externality. e network externality effect is an increasing function of cumulative sales. Network externalities are built up, starting in the second period, based on the sales in the first period. e second period’s WTP depends on the first period’s units sold by the seller, and the third period’s WTP depends on the cumulative sale from the first and second period, and so on. Specifically, WTP of each buyer increases by $0.50 for each unit sold in the past by a given seller. Table 1 presents some examples when sellers have accumulated units sold. Additionally, network externalities are assumed to affect only current-period buyers, so buyers do not form expectations about future network externalities.

Hypothesis and Prediction e purpose of this experiment is to answer whether predatory pricing emerges in the presence of network externalities when firms are identical in terms of technology and cost structure but goods are differentiated and incompatible. In this experiment, since both sellers have the same unit

CM


56


cost and all units costs are the same, a seller cannot price below another seller’s average cost and still make a profit in that period. As a result, a seller who engages in predation in this experiment would ensure himself a net loss in that period which is equivalent to Harrison’s type III defini-tion. Predatory pricing is reclassified for the purposes of this experiment into either a “strong” or “weak” form as follow: 1. Strong predatory pricing is defined as either seller A or B pricing below marginal (and average) cost ($1.00). 2. Weak predatory pricing is defined as seller A or B setting its price between $1.00 and $1.99. is price range guarantees a loss7 to the seller after considering the entry fee of $5.00. As a result, this definition is applied only from the second to the seventh period since neither seller has to pay a first-period entry fee. Since the goal of adopting a predatory pricing strategy is to eliminate competitors, successful predation is defined as when a seller pricing in the range of strong or weak predatory pricing, and a competing seller does not enter in the subsequent period. Nevertheless, with network exter-nalities, sellers may adopt a predatory pricing strategy for the purpose of increasing market shares, resulting in additional profits due to the creation of network externalities. is can be considered a weaker definition of successful predation. e discussion about this game’s equilibrium assumes that pricing low means pricing at $0.00, and pricing high means the single period Nash equilibrium price conditional on the seller’s own and the rival’s prices during earlier periods. e proposed equilibrium for this game is for each seller (A or B) to price low during the first 4 periods and price high during the last 3 periods. It can be shown that this combination of prices is the Nash equilibrium of the complete seven-period game. $6.008 is the equilibrium price for both sellers for each of the last 3 periods based on both sellers’ pricing at $0 during the first 4 periods. With this equilibrium, each seller loses $2.5. To answer the question of why pricing low means pricing at $0, consider the following example. In period 4, if one seller decides to deviate from the equilibrium path pricing by pricing at $0.1, 0.99 or 1.00, this

7If a seller sets the price greater than $1.99 but sells than 5 units, the seller will incur a loss. However, since the seller never knows how many units can be sold when pricing decision is made, $1.99 is the lower bound that guarantees a loss to the seller.8Based on each seller’s best response to another seller’s prices, the single period equilibrium prices in each of periods 5 to 7 are cycling within the boundary of $5.00 and $7.00. e midpoint ($6.00) drawn from this set is chosen here for illustrative purposes. In fact, all equilibrium prices discussed in this section are chosen as the midpoint from set prices which provide the support for the equilibrium.

CM


57


seller will make a total loss of $9.48, 7.52 and 7.50, respectively, rather than just $2.50 if both play the equilibrium. Additionally, in period 4, if one seller prices at $0.00 but another decides to price at $6.00 which is the single period equilibrium price conditional on both sellers pricing at $0.00 during the first 3 periods, the seller who prices at $6.00 will lose $30. is analysis can be extended to periods 1, 2 and 3 by symmetry. Intuitively, the network externalities which can be created from pricing low in period 4 are larger than the period 4 loss due to the low price. Pricing low longer than the first 5 periods is not the equilibrium because both sellers will be worse off in terms of their profits. For example, both lose $17.5 if they decide to price low during the first 5 periods and price high during the last 2 periods. Intuitively, the network externalities that can be created during the last two periods are not sufficiently large to compensate the five-period loss due to pricing low.

Conducting the Experiment e experiment was conducted at the University of South Carolina’s Beam-Lab using a z-Tree program. All subjects had to participate in a “training experiment.”9 Two days later, most of them were invited back to the “actual experiment.” Specifically, thirty-four subjects participated in the training session, while twenty-two subjects were in the actual session. e training experiment lasted approximately 75 minutes, while the actual experiment lasted approximately 2 hours. e training experiment consisted of four rounds. In each round, subjects played a monopoly game with network externalities. ey were seller A during the first two training rounds, and seller B during the last two training rounds, since subjects did not know which seller role would be assigned to them in the actual experiment yet. Moreover, subjects would learn that buyers’ WTP for seller A’s and seller’s B goods which depended on cumulative units sold in previous period(s) were different, although sellers were identical in terms of their marginal cost. ere were seven periods in each of the four training rounds. Each subject received $5 participation fee regardless of his or her performance. e actual experiment consisted of one practice period and six actual rounds. e practice period allowed subjects to refresh their memories regarding the screen layout. Because each subject was assigned the role of either seller A or B throughout all six actual rounds, each type of seller practised with his or her own cost structure in the practice period. Since

9Subjects in Rutstrom (1986), Harrison (1988) and Goeree and Gomez (1998) had prior experiences in monopoly experiments.

CM


58


there was no cost advantage for either seller A or B, role switching was not necessary. Subjects were paired randomly and re-matched between each actual round to avoid repeated game effects. e experiment consisted of three sessions. All three sessions were conducted with the same procedures, but with different numbers of sub-jects and different payment methods. Eight subjects participated in the first session and were informed that one out of six rounds was randomly picked by a lottery and paid for. Eight subjects participated in the second session but three out of six rounds were paid for. Six subjects participated in the third session and one out of six rounds was paid for.

RESULTS

Tables 2a to 2f summarize basic results from all 22 subjects10 by round. Recall that the weak predatory pricing definition does not apply to the

first period because sellers were automatically in the first period without paying an entry fee of $5. e only cost incurred in the first period was the unit cost of $1.

Table 2a: Numbers of Strong and Weak Predatory Pricing Decisions in Round 1.Period Strong Predatory Pricing Weak Predatory Pricing Total

1 1 N/A 12 0 2 23 0 0 04 0 1 15 0 0 06 0 0 07 0 1 1

Table 2b: Numbers of Strong and Weak Predatory Pricing Decisions in Round 2.Period Strong Predatory Pricing Weak Predatory Pricing Total

1 5 N/A 52 0 5 53 1 0 14 0 0 05 0 1 16 0 1 17 1 0 1

10e Mann-Whitney test is applied to prices set by sellers A and prices set by sellers B in all 6 rounds. e test cannot reject the null hypothesis that there is no difference between these two groups at 5% significant level.

CM


59


Table 2c: Numbers of Strong and Weak Predatory Pricing Decisions in Round 3.Period Strong Predatory Pricing Weak Predatory Pricing Total

1 8 N/A 82 4 8 123 3 4 74 1 2 35 0 1 16 0 2 27 1 0 1

Table 2d: Numbers of Strong and Weak Predatory Pricing Decisions in Round 4.Period Strong Predatory Pricing Weak Predatory Pricing Total

1 9 N/A 92 3 10 133 2 2 44 1 3 45 2 0 26 1 1 27 1 0 1

Table 2e: Numbers of Strong and Weak Predatory Pricing Decisions in Round 5.Period Strong Predatory Pricing Weak Predatory Pricing Total

1 16 N/A 162 7 12 193 4 6 104 1 3 45 0 2 26 0 2 27 2 0 2

Table 2f: Numbers of Strong and Weak Predatory Pricing Decisions in Round 6.Period Strong Predatory Pricing Weak Predatory Pricing Total

1 15 N/A 152 8 9 173 3 5 84 1 0 15 0 0 06 2 0 27 1 0 1

CM


60


DISCUSSIONe Existence of Predatory Pricing

Considering the first four periods11 across all six rounds, there were 166 pricing decisions (out of 528 possible decisions = 22 subjects x 6

rounds x first 4 periods) or 31.44% that satisfied either the weak or strong predatory pricing definition. Note that the predatory pricing definition here is actually equivalent to type III predatory pricing definition employed by Harrison which involves a loss for the predating firm. Out of the 166 predatory pricing decisions, 94 could be classified as strong predatory pricing and the other 72 as weak predatory pricing. Most strong predatory pricing decisions occurred during the first period (54 decisions). Considering only period two to seven, most sellers tended to engage in weak predatory pricing (72 decisions) rather than strong predatory pricing (39 decisions). Note that although I did not see a lot of predatory pricing decisions (pricing below $1.00) in the first period, there were 24 pricing decisions at exactly $1.00. ough some sellers were not willing to lose, they were willing to accept a break-even. e $1.00 pricing in the first period was below the duopoly’s profit-maximizing price12 of $4.00 or the monopoly price13 of $3.00 or $4.00. Also, there were 40 decisions of pricing at exactly $2.00 in periods 2 through 4 from all rounds, corresponding to the upper boundary of weak predatory pricing. In order to analyze how effective predatory pricing is in generating profits for sellers in one round, I propose two simple ways to measure predatory intensity. First is to count the number of either strong or weak predations in all seven periods. Second is to calculate the difference between average prices in earlier periods and average prices in later periods. A large difference implies that sellers price low early, obtain network benefits and raise prices later. Based on Table 4, a panel model based on GLS random-effect esti-mation is analyzed with the average profit across seven periods as the de-pendent variable. Independent variables include the number of predations, five dummies representing each of the first five rounds and demographic variables.14 Acronyms of all variables and their descriptive statistics are

11Based on the Nash equilibrium, (weak or strong) predatory pricing is predicted to occur during the first four periods.12In the first period, by pricing at $4.00, each of them makes $6.00 profit, which is the maximum profit.13If there is only one seller in the market, the profit-maximizing price is $3.00 or $4.00 and the profit is $6.00.14Due to the small sample size of 22 subjects, demographic variables here are used for the purpose of controlling the random variations among sessions rather than for the analysis purpose.

CM


61


presented in Table 3. e number of predation has a negative coefficient and is significant at the 5% level in all four models. is implies that the more predations sellers employed in a round, the less profits they make. In this sense, predation was not successful.

Table 3: Variables and Descriptive Statistics.=Variable Explanation Mean Std Dev N

Tprofit Earnings in one round (7 period summed profit) for each subject.

19.74 35.76 132

Tpredate Number of predation in one round. 1.42 1.47 132Dif1457 e difference between average prices in

period 1 to 4 and 5 to 7.3.19 1.93 131

Gender Male = 1, Female = 0. 0.50 0.50 132Major Business or economics major = 1, Other

majors = 0.0.59 0.49 132

GPA GPA choices from the questionnaire. 2.41 1.27 132Risk Risk attitude. 0.05 1.51 120

Note: GPA = 1 means GPA between 3.75 and 4.00, GPA = 2 means GPA between 3.25 and 3.74, GPA = 3 means GPA between 2.75 and 3.24, GPA = 4 means 2.25 and 2.74, GPA = 5 means GPA between 1.75 and 2.24, GPA = 6 means GPA between 1.25 and 1.74, GPA = 7 means GPA less than 1.25.Risk = “negative” means risk loving, “0” means risk neutral, “positive” means risk averse. Before the session began, all subjects were tested for their risk attitudes. Each subject was basically asked to choose between a fixed payoff of $2.50 or variable earnings which depended on the outcome of a roll of a ten-sided dice.

CM


62


Table 4: Panel Model Predicting Profit Outcome Based on the Number of Predations in One Round.

Variable

Round1 -1.839(10.241)

-2.962(10.929)

Round2 -4.524(9.833)

-1.574(10.463)

Round3 -6.658(9.282)

-4.255(9.946)

Round4 -8.881(9.269)

-9.408(9.932)

Round5 -7.375(9.297)

-7.353(9.999)

Gender 22.929*(8.691)

23.042*(8.693)

Major -13.934(9.750)

-14.731(10.022)

GPA -4.246(3.476)

-4.178(3.480)

Risk 1.947(2.941)

2.097(2.973)

Tpredate -9.095*(2.007)

-8.629*(2.525)

-9.176*(2.187)

-8.612*(2.742)

Constant 32.622(5.057)

36.842(8.851)

39.200(13.289)

42.906(15.152)

F-Test (p-value) 0.932 0.009 0.097Note: N = 132 for the first two models. N= 120 for the last two models. Tprofit is the dependent variable. Numbers in parentheses are standard errors. * P-value < 0.05. F-test compares each of the last three models to the first model.

Consider a similar model but with the difference between average prices in periods 1 to 4 and in periods 5 to 7 as dependent variable, the model is presented in Table 5. e average price difference variable and dummies representing the first two rounds have positive coefficients and are significant at the 5% level in all four models. is implies that the larger the price difference, the higher the profits in a round. In other words, if sellers start a round by pricing very low and obtaining network benefits, and raise prices later, they increase profits in that round. In summary, there is a positive relationship between profit outcomes and the loss a seller is willing to take during earlier periods by pricing low, but there is a negative relationship between profit outcomes and a number of times a seller engages in predatory pricing in one round.

CM


63


Table 5: Panel Model Predicting Profit Outcome Based on Price Difference in One Round.

Variable

Round1 15.840*(6.878)

20.040*(7.122)

Round2 14.176*(6.906)

18.796*(7.133)

Round3 1.415(6.886)

6.184(7.110)

Round4 1.543(6.905)

4.049(7.143)

Round5 -9.370(6.878)

-6.293(7.102)

Gender 11.871*(5.835)

11.666*(5.848)

Major -11.479(6.265)

-11.245(6.272)

GPA -4.197(2.271)

-4.249(2.284)

Risk 1.132(1.916)

1.078(1.923)

Dif1457 12.926*(1.149)

13.011*(1.099)

12.974*(1.217)

13.349*(1.150)

Constant -21.316(4.517)

-25.560(6.548)

-11.966(9.448)

-20.288(10.701)

F-Test (p-value) 0.002 0.009 0.000Note: N = 132 for the first two models. N= 120 for the last two models. Tprofit is the dependent variable. Numbers in parentheses are standard errors. * P-value < 0.05. F-test compares each of the last three models to the first model.

Successful Predation Predatory pricing can be considered unsuccessful in eliminating competitors in these games. From all six rounds, there were only 13 times when either seller A or seller B decided not to participate in a particular period.15 ere were two sellers who decided not to participate at least two periods consecutively. ese two can be considered the only successful instances of predation. e first one did not participate in any of periods

15In some other cases, I observed a few sellers who did not participate in a particular period but then decided to reenter in the next period. Among these, sellers who decided not to enter a particular period generally had priced higher than another competing seller in the previous period. I do not consider this to be a successful predation.

CM


64


4, 5, 6 and 7 in one round. is seller priced at $1.00, $2.00 and $2.00 while the competitor priced at $0.00, $0.49 and $1.98 in period 1, 2 and 3, respectively. is seller lost a total of $8.00 during periods 1 through 3, so he decided not to continue the round. e competitor’s cost was $11.20 from the predatory pricing. e second one (different subject from the first) did not participate in periods 6 and 7 in one round. is seller (seller 1) priced at $0.99, $1.98, $0.50, $12.00 and $2.25 while his competitor (seller 2) priced at $2.00, $0.25, $2.00, $2.20 and $4.00 in period 1, 2, 3, 4 and 5, respec-tively. Seller 1 lost $17.07 during periods 1 through 5. In this case, I would consider the predatory pricing by seller 1 periods 1 and 2 unsuccess-ful but the predatory pricing by seller 2 in period 2 successful. Seller 2’s monopoly position in periods 6 and 7 allowed him to make up for this and achieve a net profit of $74.25 which is greater than the equilibrium prediction of $16.88.

Learning Subjects seemed to learn how to engage in predation as the rounds went on. e numbers of both types of predatory pricing decisions were 3, 11, 30, 31, 49 and 41 in round 1, 2, 3, 4, 5 and 6, respectively. If consider only the last two rounds, almost all sellers were engaged in predation during the first two periods. ere were 90 predatory pricing decisions out of 176 possible decisions (22 subjects x 2 rounds x first 4 periods) or 51.13% during the last two rounds. Learning seems to provide a significant impact on predation behavior. Figures 1 through 7 present medians and predictions of price offers in each period.

Figure 1. Mean and Prediction of Price Offers by Seller A and B in Period 1.Note: SD(Upper) and SD(Lower) are one standard deviation from the mean in each round.

-1 .0 0 0

0 .0 0 0

1 .0 0 0

2 .0 0 0

3 .0 0 0

4 .0 0 0

1 2 3 4 5 6

R o u n d

Pri

ce (

$)

MeanPredicted

SD(Upper)

SD(Lower)

CM


65





0 .0 0 0

1 .0 0 0

2 .0 0 0

3 .0 0 0

4 .0 0 0

1 2 3 4 5 6

R o u n d

Pri

ce (

$)

MeanPredicted

SD(Upper)

SD(Lower)

0 .00 0

1 .00 0

2 .00 0

3 .00 0

4 .00 0

5 .00 0

1 2 3 4 5 6

Ro u n d

Pri

ce (

$)

MeanPredicted

SD(Upper)

SD(Lower)

0 .0 0 0

2 .0 0 0

4 .0 0 0

6 .0 0 0

8 .0 0 0

1 2 3 4 5 6

R o u n d

Pri

ce (

$)

MeanPredicted

SD(Upper)

SD(Lower)

CM


66


Figure 5. Mean and Prediction of Price Offers by Seller A and B in Period 5Note: SD(Upper) and SD(Lower) are one standard deviation from the mean in each round.



Based on Figures 1 through 7, the mean and median price offers in periods 1 through 4 of all 6 rounds were greater than predicted while the median price offers in periods 5 through 7 of all 6 rounds were around predicted. Sellers were reluctant to price at $0.00 during the earlier periods. However, there was evidence that sellers were willing to incur losses during periods one and two, but not during all of the first four periods. Generally, the median price offer in the second period was in the range of the weak predatory pricing definition, not the strong one.

0 .0 0 0

2 .0 0 0

4 .0 0 0

6 .0 0 0

8 .0 0 0

1 2 3 4 5 6

R o u n d

Pri

ce (

$)

MeanPredicted

SD(Upper)

SD(Lower)

0 .0 0 0

2 .0 0 0

4 .0 0 0

6 .0 0 0

8 .0 0 0

1 0 .0 0 0

1 2 3 4 5 6

R o u n d

Pri

ce (

$)

MeanPredicted

SD(Upper)

SD(Lower)

0 .0 0 0

2 .0 0 0

4 .0 0 0

6 .0 0 0

8 .0 0 0

1 0 .0 0 0

1 2 .0 0 0

1 2 3 4 5 6

R o u n d

Pri

ce (

$)

MeanPredicted

SD(Upper)

SD(Lower)

CM


67


A binomial sign test is employed for each of the first four periods when pricing below $1.00 is considered to be predatory in period one and pricing below $2.00 is considered to be predatory in periods two through four. From the sign tests16 of the first period’s price from rounds 1 through 6, from rounds 2 to 6, and from rounds 3 to 6, I cannot reject17 the hypothesis that prices are above strong predatory prices, i.e., the median price is strictly less than $1.00. For the first period of rounds 4 through 6, I can reject the hypothesis that prices are above strong predatory prices. is suggests that it took three rounds for subjects to learn how to engage in predatory pricing in the first period. Based on the second period observations across all six rounds, I find that the median price is significantly less than $2.00. For periods 3 and 4, the sign tests fail to reject the null hypothesis even when some earlier rounds have been excluded. I can therefore conclude that sellers appeared not to be very aggressive during earlier periods and therefore not to predate very strongly. Sellers were also reluctant to price very high during later periods. e median prices in periods 5, 6 and 7 were in the range of $4, $5 and $6 compared to the predicted price of $6.00 in each of the last 3 periods which is conditional on both sellers’ staying on the equilibrium path by setting their prices at $0.00 during the first 4 periods.

CONCLUSION

The observations provide a strong support for the existence of preda-tory pricing especially during the first two periods of a round. When

demand and cost structures of sellers are symmetric, sellers were willing to lose during earlier periods for network benefits and consequently future profits in later periods. Weak predatory pricing decisions were observed more frequently than strong ones. Predatory pricing was not effective in getting rid of another competitor. When both sellers were present almost throughout all seven periods, prices in later periods were kept lower than what the Nash equilibrium predicted. Sellers were reluctant to price too high in later periods because another seller might undercut the price and get ahead with network benefits. Profit outcome is positively linked to the aggressiveness of predation which is measured in terms of how much a seller is willing to lose during earlier periods.

16Hypothesis: Ho: median of the first period’s prices from round...to round...≥1.00 Ha: median of the first period’s prices from round...to round...<1.0017e p-value is close to 0 not only for this test but also for other tests employed later.

CM


68


ACKNOWLEDGEMENTS

I thank Elisabet Rutström, Henry W. Chappell, Eric Johnson and the journal’s reviewer for helpful comments. I also thank Greg George for

proofreading the first draft of this paper and helping me get started with z-Tree. Any errors are the responsibility of the author.

REFERENCESAreeda, P. E., and D. F. Turner. 1975. Predatory pricing and related practices

under Section 2 of the Sherman Act. Harvard Law Review 88: 697-733.

Cabral, L., D. Salant, and G. Woroch. 1999. Monopoly pricing with network externalities. International Journal of Industrial Organiza-tion 17: 199-214.

Capra, M. C., J. K. Goeree, R. Gomez, and C. A. Holt. 2000. Predation, asymmetric information and strategic behavior in the classroom: An experimental approach to the teaching of industrial organization. International Journal of Industrial Organization 18: 205-225.

Farrell, J., and G. Saloner. 1986. Installed base and compatibility: Innova-tion, product preannouncements, and predation. American Economic Review 76: 940-955.

Goeree, J. K., and R. Gomez. 1998. Predatory pricing in the laboratory. Draft, University of Virginia.

Gomez, R., J. K. Goeree, and C. A. Holt. 1999. Predatory pricing: Rare like a unicorn? In C. Plott, and V. Smith (eds), Handbook of Experimental Economics Results. Forthcoming.

Harrison, G. W. 1988. Predatory pricing in a multiple market experiment: A note. Journal of Economic Behavior and Organization 9: 405-417.

Isaac, M. R., and V. L. Smith. 1985. In search of predatory pricing. Journal of Political Economy 93: 320-345.

Jung, Y. J., J. H. Kagel, and D. Levin. 1994. On the existence of predatory pricing: An experimental study of reputation and entry deterrence in the chain-store game. RAND Journal of Economics 25: 72-93.

Katz, M. L., and C. Shapiro. 1986. Technology adoption in the presence of network externalities. Journal of Political Economy 94: 822-841.

Rutström, E. 1986. In search of a reconciliation of results in predatory pricing experiments. Draft, University of Western Ontario.

CM


69


Rainer vom Hofe*

Understanding Housing as an Economic Stimulus:A Regional Computable General Equilibrium (CGE)

Model for New York State Housing

University of Cincinnati, College of Design, Architecture, Art and Planning, School of Planning, P.O. Box 210016, Cincinnati, OH 45221-0016, U.S.A.*Corresponding author. E-mail:[email protected]

ABSTRACT

This article examines the potential impacts of changes in housing variables, such as mortgage rate, property tax rate and public

housing investment on regional economic growth. A growing body of research focuses on the intersection between housing and regional economics because regional housing markets do not operate in a vacuum and because housing market interventions have far-reaching impacts on the overall economy, as well as on key macroeconomic indicators. Using the example of the New York State housing market, this research applies a Comput-able General Equilibrium (CGE) model to analyze the impact of selected housing policies on the New York State. e results clearly indicate the region-wide economic impacts of housing policies on selected macroeconomic key variables and should provide valuable information for housing policy analysts and policymakers.

Key words: Housing, Regional Economic Analysis, Regional Economic Modeling

INTRODUCTION

Housing is one of the most important sectors of the US economy. One-seventh of all personal consumption expenditures goes towards

housing (e.g., construction, rent and maintenance). Residential real estate accounts for slightly more than half of the nation’s fixed capital stock. Approximately, one-third of gross private investment is in residential con-struction.1 erefore, it is, as Baffoe-Bonnie (1998) points out, “important

1For comparison, Paulo Brito and Alfredo Pereira (2002, pp.247) reveal that real estate investment ac-counts for over 50% of total private investment in the U.S. Furthermore, they emphasize that real estate assets represent almost 60% of the nations’ wealth and that nearly 70% these assets are held privately by families.

CM


70


to note that the performance of the housing market has a major impact on the overall performance of the economy”. Indeed, housing starts and building permit data reported by the Census Bureau are used by economic analysts to understand and predict the health of the national economy.2 Due to the intrinsic relationship between housing and the other sectors of the economy, the reverse is also true. Housing markets are vulnerable to changes in the rest of the economy. Developers, homeowners and landlords combine inputs, such as land, materials, labor, capital and infrastructure to provide housing services. Relative prices of these inputs on one side, and prices of housing services on the other side, are among the forces regulating the provision of available housing services (Malpezzi, 1990). For example, in the economic slowdown of 2002-3, the Fed in-directly stimulated a housing market boom by lowering the federal fund rate which had a cascading effect of lowering mortgage rates, increasing housing sales and prices and encouraging building.3

It is therefore crucial that economic and housing policymakers gain a clearer understanding of the relationship between the housing market, the overall economy’s numerous performance indicators (e.g., gross domestic product (GDP), investment, inflation, labor supply, etc.), and the impacts of one on the other. Over the last 30 years, macro and regional economists have developed equilibrium models to simulate and better understand the impacts of external economic shocks, such as lowering the (federal fund) interest rate or increasing the regional investment. Initially developed to evaluate national-level trade polices and simulate structural changes in developing countries,4 equilibrium models are increasingly used at national and (multi-) regional levels for simulating macroeconomic responses to environmental and transportation policies and utility deregulations among others.5 e beauty of these models is that they allow policy analysts to see what happens to selected macroeconomic variables – GDP, inflation and employment –after a potential policy is implemented as well as per-formance of specific sectors such as transportation and housing. Using the example of New York State, this article introduces the

2http://www.census.gov/econ/www/co0200.html. e Census Bureau building permit survey is used by the Bureau of Economic Analysis, the Federal Reserve Board, the Federal Housing Finance Board, and the Department of Housing and Urban Development.3July 2003 Economic Indicators, United States Congress, pp.19, New Construction and New Private Housing, http://www.access.gop.gov/congress/eibrowse/03julbro.html4Early examples include the Adelman-Robinson model of South Korea, the Cameroon model by Robinson, Hanson, and Kilkenny, and the “maquette” model by Bourguignon, Branson, and de Melo; Robinson (1991).5Patridge and Rickman (1998) provide a detailed literature review on regional computable general equi-librium modeling.

CM


71


computable general equilibrium (CGE) framework as a potential policy analysis tool in the context of housing analysis. Embedding the housing market in an economy-wide macro model reflects the environment of the housing market - it does not operate in a vacuum and is profoundly influenced by the overall economic performance of selected regions and vice versa. While the existing literature addresses macroeconomic-housing market interrelationships using econometric models, this article introduces the application of the CGE framework to account for these economic interrelationships. For this purpose, the general equilibrium framework has been adapted to the state level (New York) and extended to explicitly model the housing market in an economy-wide setting. As such, it accounts for the different roles housing can take (e.g., consumption and investment commodity), includes input markets (e.g., capital and labor), incorpo-rates consumers (e.g., households and governments) and takes account of macroeconomic variables measuring regional economic performance, such as the gross state product (GSP), the state-level GDP equivalent. e model is applied to simulate the impact of changes to the regional housing market6 on the economy of New York State. Developed with the assistance of the U.S. Department of Housing and Urban Development (HUD) staff, these simulated housing market changes include: lowering the mortgage rate, reducing the property tax rate for selected household groups and increasing government spending on public housing. ese simulated housing market changes have complex and far-reach-ing impacts on the New York State economy and they vary significantly in how they affect the regional economy. First, housing market stimuli that originate on the housing investment side – mortgage rate reduction and government spending on public housing – have favorable economy-wide impacts. ey promote regional economic growth, increase regional com-modity output, boost the labor market and increase household income. Second, a reduction in property tax results in regional growth stagnation, reduction of regional commodity output and labor demand and negatively affects household income. ird, comparing the property tax reduction scenario across households reveals that targeting high-income household as beneficiaries reduces the negative impact on regional economic activities; also it may have negative impacts in terms of equity. Overall, the significant responses of macroeconomic key indicators to simulated changes in the housing market make a strong argument for the relevance of economy-wide

6I am very thankful to Nandinee Kutty and Alastair McFarlane, US Department of housing and Urban Development (HUD), Office of Economic Affairs for meeting with me and discussing issues related to this modeling approach and simulation of housing market changes.

CM


72


computable general equilibrium (CGE) models for housing policies. In the discussion that follows, I will review the relevant literature. e paper will then introduce the main data base used – the 1996 Social Accounting Matrix for New York State – and will highlight the interrela-tionships of housing and macroeconomic key variables. e next section will introduce the model framework – the computable general equilibrium approach. e final section will discuss the applicability of the CGE approach, highlight strengths and weaknesses and open questions for future research.

Background and literature review is article addresses the relationship between housing economics and macroeconomic modeling. From the housing side, previous studies by housing economists have investigated the significance of macroeconomic variables for explaining changes in the housing market. On the other side, macro and regional economists have modified general equilibrium models in order to more closely examine the response of the economy to changes in specific economic sectors such as environment and transportation. is study introduces a new approach of integrating housing and economic variables in a general equilibrium model, specified for the housing sector. A stream of housing economics literature has focused on relating housing to macroeconomic variables. Initial research in this area related changes in housing prices and other housing variables to changes in indi-vidual macroeconomic variables. In early research, Schwab (1983), Man-chester (1987) and Harris (1989) found a strong positive link between the rate of inflation (e.g., CPI) and the price of housing. Later studies increased the number of explanatory macroeconomic variables and the complexity of the modeled relationship between the overall economy and the housing market. Goulder (1989) used a general equilibrium model to conclude that the Tax Reform Act of 1986 (TRA86) initiatives adversely affected real housing investment and thus, the total value of housing assets. Abraham and Henderschott (1992) included employment, income growth, inflation of real construction costs and changes in real after-tax interest rates as explanatory variables to explain approximately 40% of the observed variation in regional housing prices. While my research focuses on the economy-wide impacts of changes in housing market variables on macroeconomic key indicators, Baffoe-Bonnie (1998) took a different approach by looking at the dynamic impact of macroeconomic aggregates on housing prices and the housing stocks. Conceptually, this study is similar to the present research in that they investigated the interrelationship of housing market variables and selected macroeconomic variables. In particular, Baffoe-Bonnie discussed the

CM


73


response of housing market variables to shocks in employment, inflation and the mortgage rate at the national level. He confirmed earlier research (e.g., Smith and Tesarek, 1991) which recognized that employment growth is positively related to growth in the housing market, but was unable to empirically establish a strong relationship between inflation and housing prices. He empirically showed that monetary policies strongly influence the mortgage rate, thereby causing an immediate response in both regional and national housing markets. Finally, the author concluded his research with the statement that economic indicators alone cannot explain all the fluctuations on the housing market from 1973 to 1994. Among other influential factors, he pointed to regional differences in supply elasticities, turbulences in energy markets, land use regulations and even international trade pattern as potential sources for housing market disturbances outside his modeling framework. A number of studies from the economics literature examine how housing parameters are incorporated into economic growth models. A paper by Brito and Pereira (2002) presented an innovative approach of an endogenous growth model by merging two branches of economic literature. To account for the different economic roles of housing, they combined the literature on dynamic housing modeling – which treats housing as a consumption and investment commodity (i.e., Brueckner, 1994; Turnovsky, 1994) – and the endogenous growth literature – in which endogenous growth is derived from physical and human capital (i.e., Bond et al., 1996). Based on dynamic modeling assumptions, the authors showed that the endogenous growth rate is highly related to growth in the housing market, particularly to housing prices and housing-related production parameters. ey also concluded that the endogenous growth rate is more responsive to shocks in construction – and therefore changes in the housing market– than in manufacturing. While Brito and Pereira’s dynamic approach stressed modeling of the economy over the long term, the present study employs a static modeling approach which focuses on evaluating short-term responses to exogenous shocks in macroeconomic key variables. is study mirrors that of Brito and Pereira in the treatment of the housing market as well as in the use of some of the same key modeling assumptions. e multiple roles of housing– consumption commodity, investment good and intermediate production input – are addressed, while housing and other commodities are assumed to be imperfect substitutes. Finally, economic growth is assumed to depend on regional price levels, including prices of housing. e same strong economy-wide importance of housing prices is indicated by Englund and Ioannides (1997). eir research focuses pri-marily on house price dynamics and its impact on the distribution of

CM


74


economic well-being, household savings and household’s consumption behavior. Applying the standard model of house price dynamics based on the asset-market approach developed by Poterba (1984), the authors found significant predictive power of lagged GDP growth and the real interest rate on housing prices in an international empirical comparison of 15 OECD countries. Other research explored the linkages between housing and labor markets. Mayo and Stein (1995), for example, investigated housing and labor market distortions in Poland. Based on an earning regression model, they concluded that housing market disequilibrium affected wage rates and showed strong evidence that housing policies in Poland have major deleterious consequences for macroeconomic performance, such as foregone high-return investments, depressed savings incentives and labor market and migration distortions. ey further emphasized that restrictive private housing markets, like in Poland, and poorly-structured subsidies dampen households’ incentives to save for housing purchases, and thus they may reduce the supply of domestic savings to the rest of the economy. Overall, they concluded that housing market distortions may result in labor market distortions with further penalties in economic performance. All these studies provide insight into the interrelationships of hous-ing markets to selected macroeconomic key indicators. In my research, I offer an alternative approach to traditional econometrics by merging two branches of the economic literature: housing economics and regional CGE modeling. Using the case of New York State, the present study employs this approach to explore how selected changes in housing market variables affect the performance of regional macroeconomic key indicators.

Data and conceptual approach e center piece of the New York State computable general equilib-rium (CGE) model is the 1996 Social Accounting Matrix (SAM).7 A SAM is a simple and efficient economy-wide framework to organize income flows among sectors of production, factors of production, households, govern-ments and investment sectors. It is a consistent and complete snapshot of the economy at one point in time - consistent in the sense that for every income there exists a corresponding outlay; and complete in that it iden-tifies both the recipient and the payer of each transaction. For instant, a SAM indicates intersectoral linkages (such as between construction and retail industry), records industry payments to factor markets (such as labor

7e New York State SAM was derived using the commercially available ImplanPro software package by the Minnesota Implan Group (MIG), Inc. (www.implan.com).

CM


75


payments by service industries), determines the income distribution by socioeconomic groups (for example, wages and salaries received by house-holds), and how these socioeconomic groups spend their money on final demand consumption.8 Figure 1 represents a simplified abstract presenta-tion of the applied conceptual framework and identifies various markets, economic factors and the circular flow of income (e.g., payments) with emphasis on the integration of housing consumption and investment. With this conceptual policy framework in mind, attention is now paid to describe New York State’s housing market in more detail as it appears in the New York State Social Accounting Matrix. For this reason, Table 1 puts together all housing transactions in the regional economy, from the 1996 New York State SAM, into one condensed table. e fact that the complete SAM contains as many as 20 industry sectors, three factors of production (employed labor income, self-employed labor income and capital), seven household groups, two government sectors, two investment sectors, an enterprise sector, inventory adjustments and two distinct trade accounts (trade with the rest of the U.S. and the rest of the world) make a complete textual description of the complete all income flows impossible. erefore, this paper focuses only on those aspects that are important in the context of comparing selected housing policies.

Figure 1. Housing consumption and investment in a social accounting framework.

Markets for Factorsof Production

Markets for Goodsand Services

Financial Market

Government

Firms

Households

Private Savings

Income Factor Payments

Govt. Purchases

HousingConsumption

TaxesGovt. Deficit

Firm Revenue

HousingInvestment

8For an in-depth discussion on social accounting matrices and their use as a data system and as a conceptual framework for policy and planning purposes, the following references may give some guidance on where to start reading: i) Social Accounting Matrices: A Basis for Planning, edited by Graham Pyatt and Jeffery I. Round, ii) Social Accounting Matrices and Social Accounting Analysis by Erik orbecke, and iii) Input-Output Tables, Social Accounting Matrices, and Multipliers by Elisabeth Sadoulet and Alain de Janvry.

CM


76


Table 1: New York State housing market (in Millions of 1996 Dollars)9

Residential Construction

Owner-occupied Housing

Real Estate Grand Total

Intermediate Inputs 5,981.9 - 24,498.3 30,480.3

Households - 59,343.4 10,159.7 69,503.1

HHLow1 - 2,385.8 1,753.7 4,139.5 HHLow2 - 3,736.9 1,760.5 5,497.4 HHMed1 - 4,932.5 1,787.6 6,720.1 HHMed2 - 6,677.3 1,660.8 8,338.2 HHMed3 - 6,668.7 1,117.9 7,786.5 HHHigh1 - 14,710.4 1,251.6 15,962.0 HHHigh2 - 20,231.8 827.7 21,059.5Government 10.9 - 1,003.1 1,014.0

Federal 10.9 - 19.6 30.5 State/Local - - 983.4 983.4Investment 5,514.8 - 2,263.5 7,778.3

Private 5,097.0 - 2,263.5 7,360.5 Public 417.8 - - 417.8Grand Total 11,507.7 59,343.4 37,924.5 108,775.7

Referring to Table 1, about 28 percent of regional housing supply ($30.5 billion) is used by industries as intermediate inputs. Here, industry demand for housing services by the real estate sector is highest with $24.5 billion (80.4%). e demand for residential construction of $6.0 billion (19.6%) is substantially less and is exclusively used by owner-occupied housing ($3.7 billion), real estate ($1.9 billion) and government ($0.4 billion) services. Annual household demand for housing totals $69.5 billion in New York State for 1996. At the aggregated level, households spend $59.3 billion (85.4% of total housing spending by households) on owner-occupied housing and $10.2 billion (14.6%) on services by the real estate sector, including rental housing. It is apparent that with increasing income, the

9Households are broken down intoo seven groups, depending on annual income:Low-income household: HHLow1: less than 10K, HHLow2: 10-20KMedium-income household: HHMed1: 20-30K, HHMed2: 30-40K, HHMed3: 40-50KHigh-income households: HHHigh1: 50-70 K, HHHigh2: more than 70KIntermediate inputs represent an aggregated total. Internal in the model, there are as many as 20 distinct industry sectors.

CM


77


consumption of owner-occupied housing increases, while the expenditures on rental housing decline. In particular, the four lowest-income groups (out of seven income categories) spend a combined $11.1 billion for real estate ownership, while the four highest-income groups spend a combined $48.3 billion on owner-occupied housing. Further, each of the two highest-income household groups spend more on their own housing than the lowest three household groups combined, with $14.7 billion and $20.2 billion, respec-tively. Regional household spending on owner-occupied housing ranges from $2.4 billion (57.6% of total housing spending by household category) for the lowest-income category to as much as $20.2 billion (96.1%) for the highest-income category. Conversely, rental spending can be as much as $1.8 billion (42.4%) for the lowest-income category and $0.8 billion (3.9%) for the highest-income category. Total government housing consumption for New York State is $1.0 billion. With some minor demand from the Federal Government, this consists mainly of state/local government demand of $983 million. Of the total investment demand of $7.8 billion, government investment in residential construction is 0.4 billion (5.4%). is includes maintenance expenditures for public housing programs. e remaining investment demand comes from households and amounts to $7.4 billion. Private investment in residential construction, which covers repair and maintenance work, alterations and remodeling, and new residential construction totals $5.1 billion. Significantly less is invested in the real estate sector, totaling $2.3 billion including all investment in existing dwellings.

e analytical modeling framework As mentioned already, the presented regional CGE model closely follows the conceptual framework of the above-described social accounting approach. Having already described the conceptual framework in more detail, this section’s attention will then mainly focus on: i) why the chosen and more-advanced CGE approach improves the more-simplistic analytical accounting framework, in general, and ii) how the housing market can be modeled within this analytical framework to account for the existence of strong economic links between the housing market and the rest of economy in particular.

Why the CGE? Generally, applied regional CGE models combine features from different types of economic models. e present model is largely based on the socioeconomic structure of the 1996 New York State SAM and reflects all the multisectoral and multiclass disaggregation of the NYS SAM. It acts as a multimarket model in that labor, capital and commodity markets

CM


78


reconcile supply and demand decisions, driven by a model internal price mechanism. It further integrates numerous macroeconomic components, such as investment and savings, balance of payments and government budgets. Combining all these different features, it is a well-suited policy analysis tool when socioeconomic structure, economic interrelationship, prices and macroeconomic phenomena all prove important. e state-level economy-wide model captures all payments recorded in the NYS SAM in either a set of simultaneous equations or by treating them exogenously. Most equations describe non-linear relationships, with a set of constraints that relate the equations to factor and commodity markets and macroeconomic equilibria, such as savings – investment, government budget and supply and demand of primary factors of production (e.g., labor and capital). Potential policy scenarios are simulated as exogenous shocks to the model (e.g., change of the mortgage rate). Once hit by a simulated “shock”, the model is solved such that the regional economy returns to equilibrium in that all markets clear through price adjustments. is model is static by nature, meaning that all equilibria are solved for a single-time period. In this case, the CGE model provides a comprehensive snapshot of the circular flow of all payments in New York State’s economy for 1996. Comparing the CGE model to the more-simplistic SAM approach, the substantial advantage lies in its capability to: i) incorporate an endogenous price mechanism which allows the simultaneous determination of prices and quantities to achieve required factor and commodity market equilibria ii) model all economic interactions in a non-linear framework iii) account for economic structures by including non-market activi-ties, substitution effects and government interventions

Integrating housing into the CGE As mentioned earlier, the starting point for the integration of housing-related activities into the CGE framework is their appearance in the 1996 NYS SAM. In particular, the NYS SAM identifies three housing-related industry sectors. ese are: i) residential construction and maintenance, ii) owner-occupied housing, and iii) real estate, which contains rental pay-ments made by tenants.10 Together, these three industry sectors represent the supply side for housing within the model. On the demand side (also

10e definition of residental construction and real estate sectors closely follows the 1987 Standard Industrial Code (SIC). For reference please see: http://www.osha.gov/oshstats/sicser.html.

CM


79


indicated in Table 1), one must differentiate between intermediate demand by industry sectors (e.g., residential construction and real estate) and final demand by households, governments and investment sectors. e remainder of this section will take a closer look at how exactly the New York State housing market is embodied in this economy-wide macro model; particularly, the approach to integrate housing consumption and housing investment into the modeling framework.11 Given that the CGE model uses the SAM as its main data source, the CGE conceptual framework retains structural similarities in terms of mapping interrela-tionships between consumers, industries and various markets. Before the housing market is described in further detail, it is important to point out certain assumptions of the model: i) Households have no final demand towards residential construc-tion. As such, households consume either real estate, which contains rental payments, or owner-occupied housing. ii) Only households consume owner-occupied housing. iii) e federal government has almost no housing expenditures and state/local governments have only expenditures towards the real estate sec-tor. iv) Both public and private entities invest in residential construction and real estate. As indicated earlier, a set of nonlinear equations capture households’ final demand for housing services and regional private and public investment behavior in residential construction and real estate. In the case of hous-ing consumption, preference is given to the standard utility maximization approach, such that households maximize their utility by choosing between a housing consumption bundle and a consumption bundle consisting of all remaining goods and services (Figure 2). e driving force in the household decision process is changes of market prices and other model-specific parameters.12

11Overall, the regional New York State CGE model contains almost 800 individual equations. For this reason, a more detailed description of the equation system is for beyond the scope of this paper.12A constant elasticity of substitution (CES)-type utility function is applied for each of the seven household groups. is functional form allows incorporating parameters, which account for the fact that despite price changes in housing, only few households will take advantage of these price changes. Not every tenant moves into a bigger apartment just because the rental price becomes more attractive.

CM


80


Figure 2. Households with different indifference curves (I) and budget lines (B).

On the investment side, investors make their choice according to a two-level nested investment tree as indicated in Figure 3. At the aggregated level, regional investment is the composite of total private and total public investment. Public investment through governments is set exogenous due to its use as a policy variable in multiple simulations. With public invest-ment fixed, the main focus is on private investment.

Figure 3. Two-level nested regional investment tree.

At the first and aggregated level, private investors face the choice of investing in housing or in an aggregated non-housing commodity bundle. At the second level, investors can then choose among real estate and resi-dential construction investment on one side, and among all remaining individual non-housing commodities on the other side. Again, the invest-ment maximization process follows a set of utility functions which responds mainly to changes in commodity prices, including the mortgage rate and predetermined parameters.13

I 2

I 1

I3

B2

B1

X

Y

Z

All Other Goods

Housing

RegionalInvestment

TotalPrivateInvestment

TotalHousingInvestment

ResidentialConstruction

Real Estate

Investmentin All OtherGoods andServices

TotalNonhousingInvestmentTotal

GovernmentInvestment

13Here, preference is given a Cobb-Douglas utility function.

CM


81


It becomes apparent that the functional approach of modeling housing consumption and investment relies heavily on how the 1996 New York State Social Accounting Matrix maps these housing-related income flows. is in turn may be crucial in interpreting the outcomes of selected hous-ing policy simulations. For example, the definition of housing investment is critical. In the present framework, housing investment represents a final demand activity. Because of this definition, an increase in investment in residential construction will not necessarily result in a significant increase in household demand for housing services. Although housing investment will, in the long-run, increase the available housing stock (e.g., housing supply), the short-run character of the present model and the treatment of housing investment as a final demand activity prevent housing supply from changing by a large amount. However, this is not contrary to real-world expectations as housing supply is inelastic in the short-run. us, an immediate demand response cannot be expected.14

Although this section’s primary focus is on housing, it is nevertheless important to emphasize that the CGE approach presented here captures all interrelationships mapped and outlaid in the New York State SAM and as such, the housing market is only a small part of the regional economy-wide model. As Reichert (1990) and Baffoe-Bonnie (1998) correctly point out, the housing market does not work in isolation and its interrelationships with the rest of the economy play an important role in determining hous-ing and macroeconomic key variables. Referring once again to the selected body of housing literature, the described CGE model contains many of these macroeconomic variables, including among others: i) Household incomes (e.g., wages and salaries) and household savings, ii) Market for factors of productions (e.g., capital and labor), iii)Tax payments by industries and households, iv) Inflation, v) Regional economic output by industries. e model follows the literature with respect to housing in that it treats housing as a consumption and investment commodity and includes

14Within the CGE framework, only the actual commodity output from regional industries, adjusted for imports and exports, defines the supply of housing, while all final consumption by households, govern-ments, and investment sectors make up the demand-side. Housing investment is defined as part of housing demand, rather than supply. Referring to the physical housing stock as described in the relevant literature (i.e., Hirsch 1997), investment activities in housing increase the available housing stock and thus can also be interpreted as supply-side activities.

CM


82


housing prices and the mortgage rate. Having described the applied modeling framework and having established its place within the existing housing literature, the following section will finally turn to the appropriateness of the New York State computable general equilibrium (CGE) model as a policy simulation tool for evaluating hypothetical housing policies.

e selected housing market simulations Generally, each change in a housing market variable simulates an “external shock” to the existing economic system. For example, using federal money to rehabilitate the existing public housing stock would be an external shock to the regional economy. e main assumption is that the federal monetary injection into the public housing stock originates from outside the region. In return, the regional economy adjusts to such an external shock, returning to its initial equilibrium condition by follow-ing all on the model imposed constraints, such as: • price adjustments guarantee commodity, labor and capital market clearing • balance of savings and investment and regional trade In simple terms, the requirements embedded in the accounting framework, whether input-output table, social accounting matrix or, in this case, the CGE framework are that for all industry sectors, institutions and value-added and trade accounts receipts must equal expenditures. Now, let us come back once more to the main data base on which this study is based. As mentioned earlier, the way economic interrelation-ships are mapped in the SAM directs, to a large extent, how the housing market is modeled in a CGE framework. Of further importance for the simulated housing market shocks is that the framework constrains the implementation of selected changes in housing market variables. Needless to say that in reality, housing policies can combine very complex structures of incentives, eligibility criteria and beneficiary participation. However, in this study, the selected simulations are kept simple to guarantee the appli-cability of the CGE framework. In selecting the possible housing market interventions, I focused on exploring potential economy-wide impacts and particularly which housing market interventions hold the greatest promise for state-wide economic development. In this article, three conceptually-different housing market inter-ventions are evaluated and compared with each other. For comparability across these selected housing simulations, a common exogenous expenditure of $1.0 billion is assumed for all scenarios presented. In this sense, the expenditure acts as an external injection into the New York State economy.

CM


83


With the exception of the mortgage rate reduction scenario, which attracts capital from private investors, the initial external financial resources are provided by the Federal Government. In particular, the $1.0 billion is transferred from the Federal Government to the New York State/Local Government and then spent on the hypothetical corresponding housing program.

Scenario One: Twelve percent reduction of the mortgage rate. is scenario is not directly comparable with the remaining scenarios which assume an external inflow of $1.0 billion as part of the simulation. In the mortgage rate reduction scenario, the simulated external shock occurs solely via the selected price variable, the mortgage rate. e mortgage rate is set at a level which would lead to a $1.0 billion increase in housing investment and is derived from the model internal equation system. For all remaining simulations, the mortgage rate is endogenous and subjected to change.

Scenario Two: Property tax reduction e described accounting framework lumps property tax payments made by households together with all other tax payments by the corre-sponding household group to state/local governments.15 Assuming that a property tax reduction lessens the overall tax liability by households, the corresponding tax rates are computed and reduced under this simulation to reflect an overall decrease of property tax collections by state/local gov-ernments of $1.0 billion. e property tax reduction simulation was run twice. Once (scenario 2a) assumes a property reduction for households with an annual income of $70,000 plus and once (scenario 2b) targeted towards households with an annual income of $30,000-40,000.

Scenario ree: Public housing programs Under this supply-side housing simulation, public authorities spend a total of $1.0 billion on the construction of new public housing and on rehabilitation and maintenance of the existing public housing stock. is government expenditure is treated as an increase in the demand for industry output from the residential construction sector. Results of selected housing market simulations are listed in Table

15e household tax payments are broken down into tax payments to the federal governments and tex payments to state and local governments. e former include income tax, estate and gift tax, nontaxes (fees and fines). State and Local Government revenue include income tax, estate and gift tax, nontaxes (fees and fines), moter vehicle license, property taxes, and other taxes (fishing, hunting). MIG IMPLAN Technical Report TR-98002, www.implan.com.

CM


84


2. Reading down the column shows chosen macroeconomic indicators. e base run indicates regional economic balances without any exogenous shocks and therefore simply repeats summary results from the 1996 New York State SAM. For example, real gross state product (GSP) in New York State was $616.4 billion in 1996. Household spending contributed $335.7 billion, government consumption $76.7 billion and investment $47.2 billion. e difference to complete the GSP of $156.9 billion comes from the state’s positive trade balance (e.g., exports net of imports). Combined regional household income (B) for all seven different household groups amounts to $464.1 billion and the region’s aggregated commodity output (C) values $983.5 billion.

Table 2: Selected simulation results in million dollars (in deviations from the base run).

A) FD Consumption/ Base Run Mortgage Rate Reduction

Tax Reduction toHH 30-40K HH>70K

Publich Housing

Real GSP 616435.5 222.6 (266.6) (201.9) 193.2

Household 335,661.3 11.3 862.5 514.4 9.8 Housing 69,503.1 1.7 36.3 5.7 1.5 Nonhousing 266,158.1 9.6 826.2 508.7 8.3Government 76,656.7 0.0 0.0 0.0 0.0Investment 47,230.8 1,155.6 (2,124.5) (1,530.3) 1,003.1 Inventory 1,232.3 2.5 (1.4) (1.3) 2.1 Private Invest. 32,886.0 1,153.1 (2,123.1) (1,529.0) 0.9 Housing 7,360.5 1,043.3 (164.4) (156.3) 24.0 Nonhousing 25,525.5 109.8 (1,958.7) (1,372.7) (23.1) Public Invest. 13,112.5 0.0 0.0 0.0 0.0Export 282,303.4 (456.7) 378.8 334.2 (396.5)Import 125,416.7 487.6 (616.5) (479.8) 423.2B) Household Income 464,148.9 338.2 (193.1) (148.2) 293.5

C) Commodity Output

983,475.4 482.6 (764.0) 552.0) 419.0

e base run results as well as the simulations results listed in Table 2 are at a highly-aggregated level. Household income, for example, could be broken down by different household groups as well as by different income sources (e.g., wages and salaries, social benefits, etc.). e reader can easily imagine that there would be an almost unlimited number of individual results to be reported. However, for the purpose of introducing the CGE framework in the context of housing policies, I will only list aggregated macroeconomic variables which show the trend of different housing poli-

CM


85


cies for regional economic development. e next step then is running the above-described three different policy scenarios and taking a closer look on how these simulated “exo-genous shocks” change final demand consumption, household incomes and industry output in New York State. e reported results indicate deviations from the base run. Altogether, the results suggest a wide difference among the economic impacts of different simulations. Housing programs, such as housing invest-ment strategies (i.e., public housing) and mortgage rate reduction, positively impact the regional gross state product (GSP), suggesting a potential for promoting regional economic growth. e returns to capital and labor as factors of production increase and improve the overall income situation for all household categories. Depending on the magnitude of increase in household incomes, household consumption spending increases. By and large, reducing the property tax shows superior impacts with respect to the consumption level of households, and therefore has potential to improve the vertical equity among households. However, this large increase in private consumption spending negatively impacts regional investment spending which leads to a decline in the level of industrial pro-duction. What follows is a decline in the regional GSP and in household incomes. From here it becomes obvious that the large increase in house-hold consumption – which follows an increase in disposable household income through a tax reduction – is solely met through a sharp decline on investment. In addition, the decline in investment spending is further magnified by reduction in industry output, which logically follows the drop in investment and a large negative trade balance. Overall, the increase in household consumption following the tax cut leads to a substitution of household demand for fixed investment demand, which inhibits economic growth and leads to a decline in the gross state product. e remainder of this section will now summarize the selected housing policies, their potential for fostering regional economic growth in general and their influence on selected macroeconomic indicators in particular.

Scenario One: Twelve percent reduction of the mortgage rate. Simulation of the mortgage rate reduction scenario indicates superior results in terms of investment, economic growth and composite commodity demand. Although the mortgage rate is not exclusively under government control and therefore, usually not an alternative for federal housing programs, its importance for the dynamics of the housing market makes its inclusion as a potential policy scenario useful. Following the mortgage rate reduction, private housing investments increase by $1.04 billion (14.2%) which in turn stimulates industry production. Aggregated

CM


86


regional commodity production increases by $482.6 million (0.05%) and household consumption increases by $11.3 million (<0.01%). Despite large increases in housing investment, observed changes in housing con-sumption are relatively small. e CGE framework treats both investment and consumption spending as final demand expenditures. is implies that changes in investment spending do not directly relate to changes in consumption spending. Rather, changes in the level of investment impact the level of regional commodity supply which, in turn, influences the level of private consumption. e mortgage rate scenario further indicates a negative regional trade balance of $944.3 million (0.6%) as exports decline while imports simultaneously increase. Real GSP increases by $222.6 million (0.04%), mostly due to the increase in investment spending. Driven by the rise in commodity outputs, labor and capital payments increase accordingly and households benefit, as their aggregated income, goes up by $338.2 million (0.07%).

Scenario Two: Property tax reduction Generally, the results from this scenario point out a sharp increase in household consumption, although household income declines. is is explained by the way in which tax cuts affect household income and by the overall decrease of investment spending. A tax reduction of $1.0 billion represents a cut on the household expenditure side (which includes taxes) rather than additional income in form of wages and salaries or transfer pay-ments from other institutions. As household consumption increases, overall investment expenditures decline sharply as (refer to Figure 1) households and the investment sector compete for goods and services from the com-modity market. As a result, the large increase in household consumption is compensated for by the decline in investment spending.16 is then leads to a decline of commodity production and thus to a reduction of house-hold incomes in form of wages and salaries. is explains why aggregated household incomes decrease in all tax reduction simulations (Table 2). e across-household comparison shows that the simulated tax cut increases household consumption for higher-income households by $514.4 billion versus $862.5 billion for the lower-income household group. Starting with higher levels of annual income and consumption, wealthier households do not necessarily need to use the tax savings for consump-tion purposes as may households with lower annual incomes. Rather, they

16e result is similar to the crowding out effect of investment, which follows expansionary government consumption.

CM


87


have the potential for saving larger amounts of the tax cut and making more money available for investment which helps explain the compara-tively lower drop in regional investment expenditures of $1,530.3 billion, compared to $2,124.5 billion through the lower-income households. In the end, targeting higher-income households through this tax reduction dampens the decrease of the regional GSP to a total of $201.9 billion, compared to $266.6 billion through the lower-income household group and lessens the loss in aggregated household income and commodity output of $148.2 billion and $552.0 billion versus $193.1 billion and $764.0 billion, respectively.

Scenario ree: Public housing programs Comparable in its economy-wide impacts to the mortgage rate reduction is the public housing program which assumes an increase in government investment in residential construction (new construction and maintenance on existing structures) of $1.0 billion. e level of changes in selected key economic indicators is similar to the mortgage rate reduc-tion scenario. is might be expected as the simulations are similar in that both stimulate regional housing investment. By increasing regional spending in public housing by $1.0 billion, household consumption ($9.8 million), regional gross domestic product ($193.2 million), household income ($293.6 million) and commodity output ($419.0 million) all in-crease. And again, the trade balance decreases by a large amount ($819.7 million) to compensate for the large increase in investment consumption which cannot be met by an increase in local commodity production.

Summary of findings I conclude this paper with the findings that exogenous changes to the New York State housing market have complex and far-reaching impacts on the state-wide economy and that these selected housing market stimuli vary significantly how they affect the regional economy. First, housing market stimuli that originate on the housing investment side – mortgage rate reduction and government spending on public housing – have favorable economy-wide impacts. ey promote regional economic growth, increase regional commodity output, boost the labor market and finally add towards household income. Second and contrarily, a reduction of the property tax indicates regional growth stagnation, leads to a reduction of regional com-modity output and labor demand, and therefore, negatively affects house-hold income. ird, comparing the property tax reduction scenario across households reveals that targeting high-income households as beneficiaries reduces the depressing impact on regional economic activities; also it may have negative impacts in terms of equity. Overall, the significant responses

CM


88


of macroeconomic key indicators to changes in the housing market make a strong argument for the relevance of economy-wide computable general equilibrium (CGE) models in clarifying the impacts of housing policies. In reality, many housing programs are implemented simultaneously and the housing market is much more complex than presented here in the New York State CGE model. In this first approach of explicitly accounting for housing market transactions in a computable general equilibrium model, especially with respect to the definition of housing consumption and hous-ing investment, I draw heavily upon the literature of housing economics and CGE modeling. e housing consumption equation block literally is taken from relevant work in urban economics; the housing investment equation block relies on relevant CGE literature, demonstrating the use of investment equations. Based on these assumptions which cannot completely capture the complexity of the housing market, this framework allows a first comparison of different potential housing policies and provides valuable information on how these programs affect economic variables/indicators other than housing. But it also raises the following suggestions for further research: i) Supplementing investment and housing data in the New York State database and redefining housing sectors. While the housing data were taken from the ImplanPro Social Accounting Matrix, one focus of future research would be to supplement and, if necessary or available, exchange housing data with data from specific housing focused data sources, such as HUD, American Housing Survey (AHS) and Fannie Mae. e accounting frame-work further allows supplementing existing data through data from these external sources to account for the complexity of regional housing markets and to allow more sophisticated policy simulations. ii) Expanding the model’s framework by including financial transac-tions which allow for investment alternatives and redefining the investment equation block by allowing for alternative investment choices. e modifica-tions should also consider expanding the model towards a financial CGE approach. For many investors, housing provides an investment alternative to investing in bonds, stocks and mutual funds. Allowing the model to consider investment alternatives and incorporating expectations of future investment tendencies would enrich the model’s applicability to housing issues and improve the simulation results. iii) Rephrasing the policy simulations. Regional macroeconomic models, such as the described New York State CGE model, investigate the impacts of potential policy changes at highly-aggregated levels. e aggregated nature of many key macroeconomic variables in return often constrains the applicability of the model in that policy simulations cannot be correctly specified. For example, the property tax reduction cannot isolate property

CM


89


tax payments from other state/local household tax payments (e.g., income tax). erefore, the simulated property tax reduction program would be identical in its outcome to an income tax reduction. Here, a disaggrega-tion of household tax payments into its individual components would provide more insight in the dynamics of a property tax reduction and would therefore lead to more precise simulation outcomes. However, any disaggregation requires additional equations as each new variable needs to be determined endogenously, unless fixed exogenously. is research clearly indicates that evaluating housing policies in a computable general equilibrium framework adds valuable information to the decision-making process for housing policy analysts and policy-makers. e research shows the regional economic impacts of each housing program with respect to regional economic growth, household income and other selected key variables. e research further suggests opportunities for future model development, focusing on improving the way the housing market is modeled internally in the model.

REFERENCESAbraham, J. M., and P. H. Henderschott. 1992. Patterns and determinants

of metropolitan housing prices, 1977–1991. p. 18-42. In Real Estate and the Credit Crunch, Federal Reserve Bank of Boston Conference Series, Boston, MA.

Baffoe-Bonnie, J. 1998. The dynamic impact of macroeconomic aggregates on housing prices and stock of houses: A national and regional analysis. Journal of Real Estate Finance and Economics 17(2): 179-197.

Bond, E. W., P. Wang, and C. K. Yip. 1996. A general two-sector model of endogenous growth with human capital and physical capital: Balanced growth and transitional dynamics. Journal of Economic eory: 68(1): 149-173.

Brito, P., and A. Pereira. 2002. A recent paper on housing and endogenous long-term growth. Journal of Housing Economics 51: 246-271.

Brueckner, J. 1994. Housing ownership and the business cycle. Journal of Housing Economics 3(3): 165-185.

Devarajan, S., J. D. Lewis, and S. Robinson. 1991. From stylized to applied models: Building multisector CGE models for policy analysis. Working Paper No. 616, Department of Agriculture and Resource Economics, University of California at Berkeley.

Englund, P., and Y. M. Ioannides. 1997. House price dynamics: An in-ternational empirical perspective. Journal of Housing Economics 6(2): 119-136.

CM


90


Goulder, L. 1989. Tax policy, housing prices, and housing investment. Regional Science and Urban Economics 19: 281-304.

Harris, J. 1989. e effect of real rates of interest on housing prices. Journal of Real Estate and Finance and Economics 2: 47-60.

Henderson, V. J., and Y. M. Ioannides. 1986. Tenure choice and the demand for housing. Economica 53: 231-246.

Hirsch, W.Z. 1997. Urban economics. Pearson Custom Publishing, Bos-ton, MA.

Koh, Y. K., D. F. Schreiner, and H. Shin. 1993. Comparisons of regional fixed price and general equilibrium models. Regional Science Per-spectives 23: 33-80.

Malpezzi, S. 1990. Urban housing and financial markets: Some interna-tional comparisons. Urban Studies 27: 971-1022.

Manchester, J. 1987. Inflation and housing demand: A new perspective. Journal of Urban Economics 21: 105-125.

Mayo, S., and J. Stein. 1995. Housing and labor market distortions in Poland: Linkages and policy implications. Journal of Housing Eco-nomics 4(2): 153-182.

Minnesota IMPLAN Group, Inc. 2002. IMPLAN Professional 2.0: Eco-nomic Impact Assessment Software System. Stillwater, Minnesota.

Minnesota IMPLAN Group, Inc. 1998. Elements of the Social Accounting Matrix. Technical Report, TR-98002, Stillwater, Minnesota.

Muth, R. F., and A. C. Goodman. 1989. e economics of housing markets. Hardwood Academic Publishers, Chur.

Partridge, D. M., and D. S. Rickman. 1998. Regional computable general equilibrium modeling: A survey and critical appraisal. International Regional Science Review 21(3): 205-248.

Poterba, J. M. 1984. Tax subsidies to owner-occupied housing: An asset market approach. e Quarterly Journal of Economics 99: 729-752.

Pyatt, G., and J. I. Round. 1985. Social Accounting Matrices: A basis for planning. e World Bank, Washington, D.C.

Reichert, A. K. 1990. e impact of interest rates, income, and employment upon regional housing prices. Journal of Real Estate and Finance and Economics 3: 373-391.

Robinson, S. 1991. Macroeconomics, financial variables, and comput-able general equilibrium models. World Development 19(11): 1509-1525.

Sadoulet, E., and A. de Janvry. 1995. Quantitative development policy analysis. Baltimore, e John Hopkins University Press.

CM


91


Smith, B. A., and W. P. Tesarek, 1991. House prices and real estate cycles: market adjustments in houston. Journal of the American Real Estate and Urban Economics Association 19: 396-416.

Schwab, R. 1983. Real and nominal interest rates and the demand for housing. Journal of Urban Economics 13: 181-195.

Turnovsky, S. 1994. Taxes, housing, and capital accumulation in a two-sec-tor growing economy. Journal of Public Economics 53: 245-267.

orbecke, E. 1997. Social Accounting Matrices and social accounting analysis. p. 281-332. In W. Isard (ed). Methods of interregional and regional analysis, Ashgate Publishing, Aldershot, U.K.

CM


93


Xu Xianming*

A Study of First-Year College Students’ MetacognitiveAwareness of Reading Strategies

Institute of Foreign Languages, Yuxi Teachers’ College, Yunnan Province 653100, People’s Republic of China.*Corresponding author. E-mail: [email protected]

ABSTRACT

Metacogniton refers to the ability of learners to be aware of and monitor their learning process. is study was designed to inves-

tigate the metacognitive awareness of reading strategies that college students used in reading English as a foreign language (EFL) in their academic reading. Data discussed in this paper were collected from a questionnaire, classroom observation and interviews. e findings provided evidence that the participants had awareness and a reasonable control of metacognitive strategies while reading academic texts.

Key words: first-year college students, metacognitive awareness, English reading strategies

THE RATIONALE AND RESEARCH QUESTIONS

Reading is a fundamental and necessary skill for anyone learning English as a foreign language. Learners use reading as a tool for learning and

acquiring English and to gather information for the professions of medicine, science, technology and law, etc. Strategies are vital in English reading. Students use a variety of strategies to assist them with the acquisition, storage and retrieval of information. Students who pursue their studies at the university level especially need to read English texts because the study of English is included in the national curriculum. e objectives of college English teaching are to encourage students to develop more efficient reading abilities as well as good listening, writing and speaking abilities. We need to enable students to obtain information in English concerning their own majors and to lay a solid foundation for further English study. Clearly, students have to deal with a lot of English in their college studies. ey are expected to read and understand various kinds of texts from paragraphs, passages, simplified articles and business letters and operation manuals for tools, equipment, and machinery. e ability to

CM


94


cope with different kinds of texts is considered an important stepping stone for students in choosing their careers and continuing their studies. It is obvious that reading plays an increasingly important role in learning English, especially in higher education and in the instruction of the students who propose to enter professions in science and technology. Since the early seventies, research in this area has, for the most part, focused on the best way to teach English learners to use a variety of strategies in order to read better. e current focus of research in foreign language reading has begun to concentrate on readers’ strategies. Oxford and Crookall (1989) define “Strategies as learning techniques, behaviors, problem solving or study skills which make learning more effective and efficient”. Rigney (1978) states that reading strategies are of interest for what they reveal about the way readers manage their interaction with written text and how these strategies are related to text comprehension. Research in second language reading suggests that strategies improve read-ing comprehension. Comprehension or reading strategies indicates how readers conceive of a task, how they make sense of what they read and what they do when reading comprehension is difficult (Singhal, 2001). According to Nuttall (1996), reading is considered a process of decoding, deciphering, identifying, articulating, pronouncing, understand-ing and responding. It can be understood as cognition and metacognition and involves the use of cognitive and metacognitive strategies to increase efficiency as readers strive to reach a desirable outcome. Cognition is “the intellectual processes through which information is obtained, transformed, stored, retrieved and used. It is also the mental process or faculty of knowing, including awareness, perception, reason-ing, language, memory and judgment” (Brown and Baker, 1983 cited in Webster’s Ninth New Collegiate Dictionary, 1983). Metacognition is an important concept in cognitive theory. It is “Knowing How To Learn” (Nelson, 1998), and figuring out which strate-gies work best. It consists of two basic processes occurring simultaneously: monitoring your progress as you learn and making changes and adapting your strategies if you perceive that you are not doing well (Winn and Snyder, 1998). According to Garner (1987), metacognitive strategies involve both knowledge about learning (metacognitive knowledge) and control or regulation over learning (metacognitive strategies or experiences). According to Block (1992), metacognitive strategies include the activities undertaken by readers to plan, to predict outcomes and to control, monitor and evaluate their comprehension. e author points out that metacognition involves the ability to think about and control our own learning. Readers need to use a variety of strategies to facilitate comprehen-

CM


95


sion and learning. ey must analyze the reading task before them and reflect on what they know or do not know about the material to be read. Finally, they need to devise plans for successfully completing the reading and to constantly evaluate and check their progress in accomplishing these various tasks (Vacca and Vacca, 1999). In spite of the fact that many of the previous studies have obtained information about learners’ strategies and the reading process, few of these studies have examined readers’ metacognitive awareness of reading strategies (Singhal, 2001), specifically the studies that investigated Chinese university students how to use the metacognitive strategies in their EFL academic reading. erefore, this specific research is designed to investi-gate the metacognitive awareness of reading strategies at university level of Chinese students. It will examine “where, when, how and why” they use metacognitive reading strategies in reading English. Obviously, this study cannot cover all the studies which have been conducted in this area. It will not attempt to distinguish all the categories of metacognitive reading strategies, nor will it investigate how much meta-cognitive awareness can aid reading comprehension, nor will it examine what strategies successful and unsuccessful readers apply in their reading or how to define positive or negative strategies, etc., but this study will discuss the following questions: 1). What metacognitive reading strategies are first-year college students aware of when they read academic materials such as textbooks? 2). What are their observed metacognitive reading strategies when they read academic materials? 3). How do they use these strategies?

RESEARCH DESIGNe Population and the Target Group

The population was first-year students of Yuxi Teachers’ College, enrolled in the first semester of the 2003 academic year. ey were

64 classes, 1245 students distributed in different majors. Normally, the students are divided into two groups according to the performance on the placement test, Group A and Group B. e performance of Group A is better than Group B. e subjects were 74 students, who had been chosen randomly from 11 majors: Chinese Language and Literature (8 students), Applied Math-ematics (8 students), Ideology and Politics (4 students), Law (8 students), Physics (8 students), Chemistry (6 students), Geography (8 students), Fine Arts (4 students), Music (6 students), Physical Science (6 students)

CM


96


and Computer Technology (8 students) majors, who were from 22 classes which cover two Groups (Group A and Group B). e researcher hoped that the participants who came from different majors and groups would more accurately reflect the general situation and provide more reliable data.

e Research Instruments ree kinds of instruments have been used in this study. ey are questionnaire, classroom observation and interviews. Questionnaire used in this study is known as the Survey of Reading Strategies (Mokhtari and Sheorey, 2002). It includes Global Reading Strategies (GLOB), Problem Solving Reading Strategies (PROB) and Support Reading Strategies (SUP). Passive participant observation, checklist and note-taking (Lynch, 1996) were used to explore what really happened while reading academic text. Interviews were used to discover the “when, where, why and how” of the students’ use of strategies. e interview questions consisted of 10 items (see Appendix 3). Most of them were open-ended questions.

Data Analysis and Findings Data Analysis Charts, columns, percentages, note-taking, tape-recording and an observation checklist were used to analyze the data. e charts, columns scoring and percentage analysis were used to interpret the students’ meta-cognitive awareness of reading strategies and quantify the frequency of using metacognitive reading strategies. Note-taking and the observation checklist were used to measure the students’ metacognitive actions or strategies in the interaction between the text and the teacher or peer.

• e Analysis of Questionnaire e questionnaire(see Appendix 1) was completed by 74 students. It was completed with the following conditions. 1). e total time given was 15 minutes. 2). After explaining the purpose of the inventory, the researcher directed the students to read each statement and indicated how they used the strategy described in that statement, using the 5-point Likert scale provided after each statement. 3). Researcher reminded the students that their responses were to refer only to the strategies they used when reading school-related materials, not leisure materials such as newspapers or magazines. 4). Students also were encouraged to respond honestly to each statement in the inventory and to ask questions about any aspect of the inventory they did not understand.

CM


97


When students finished answering the questionnaire, the answers to the questionnaire were collected and scored. en, the results were tabulated and decoded the scores by using the following code system: (1). “+” means a high level of metacognitive reading strategy usage (mean of 3.5 or higher). (2). “O” means a moderate level of metacognitive reading strategy usage (mean of 2.5 to 3.4) (3). “ –” means a low level of metacognitive reading strategy usage (mean of 2.4 or lower) (see following section 3.2.1: Table 1. and Table 2.)

• e Analysis of Classroom Observation Data were also collected from classroom observation and interviews. Before observation, a checklist (see Appendix 2) was prepared to facilitate the observation. During the observation, the researcher sat beside the par-ticipants, observed them and took down the strategies they used in their reading.

• e Analysis of Interviews Following-up interviews were given to the students after class or during the break. Most of the interview questions were open-ended (see Appendix 3). For example, when you are reading, if you encounter unknown words or phrases, what would you do? en the participants were asked to explain the usefulness of their strategies. All the interviews were tape-recorded and recordings were transcribed immediately after interviews.

Findings • e Findings of Questionnaire

Table 1: Global Reading Strategies

Global Reading StrategiesNumber of Students

Percentages

1. Planning before reading 62 84%2. Using background knowledge 71 96%3. Skimming and scanning 59 79%4.inking about the purpose of a reading 51 69%5. Noting length and organization 45 61%6. Knowing what to read closely and what to ignore 45 61%7. Using tables, pictures and context clues 55 74%8. Using context clues 70 95%9. Using bold face and italics to identify key infor-mation

51 69%

CM


98


10. Analyzing and evaluating the information 38 51%11. Checking understanding 59 80%12. Guessing meaning from the context 70 94%13. Checking if guesses are right or wrong 49 66%

Table 2: Problem Solving Reading Strategies.

Problem Solving Reading StrategiesNumber of Students

Percentages

1. Reading slowly and carefully 61 82%2. Getting back on track when distracted 61 82%3. Adjusting reading speed 62 84%4. inking about difficulties while reading 47 63%5. Stopping from time to time and thinking 57 77%6. Visualizing information while reading 41 55%7. Re-reading 56 76%8. Guessing unknown words 58 79%

Table 3: Support Reading Strategies.

Support Reading StrategiesNumber of Students

Percentages

1. Taking notes while reading 39 53%2. Underlining and circling 66 89%3. Using dictionary or references 50 67%4. Paraphrasing 38 52%5. Going back and forth in reading 55 74%6. Asking questions about the text 35 47%7. Translating 64 86%8. Using both English and Chinese to think about information

57 77%

9. Reading aloud to help them understand what they read

13 18%


CM


99


Table 4: e ree Subscales of Metacognitive Reading Strategiese ree Subscales of Reading Strategies

Global Reading Strategies

Problem SolvingReading Strategies

Support ReadingStrategies

Levels High (+)

Mod. (O)

Low (–)

High (+)

Mod. (O)

Low (–)

High (+)

Mod. (O)

Low (–)

Number of Students

28 44 2 31 42 1 8 58 8

Percentages 38% 59% 2.7% 42% 57% 1.3% 11% 78% 11%

Table 5: Metacognitive Awareness of Reading Strategies Metacognitive Awareness of Reading StrategiesLevels High (+) Moderate (O) Low (–)Number of Students 18 students 55 students 1 studentPercentages 24% 74% 1.3%

e findings from questionnaire provided evidence that what students used were: background knowledge, context clues, checking understanding, guessing meaning, skimming and scanning, using tables and pictures, adjust-ing reading speed, reading slowly and carefully, guessing unknown words, re-reading, stopping and thinking, underlining and circling, translating and using dictionary. Well, the strategies they didn’t pay much attention to were: noting length and organization, noticing what to read closely and what to ignore, analyzing and evaluating the information, visualizing information, thinking about difficulties while reading, reading aloud to help them understand what they read, asking question of the text and paraphrasing and note-taking. e three subclasses of metacognitive reading strategies (see Table 4) showed that most of them were in the high (28 and 31 students) and moderate (44 and 42 students) levels of Global Reading Strategies and Problem Solving Reading Strategies respectively. ree students were in the low level. But for Support Reading Strategies, most students (58 students) were in the moderate level and only a few of them (8 students each) were in the high and low levels. e findings of metacognitive awareness of reading strategies (see Table 5) showed that the participants had different levels: high, moderate and low awareness of metacognitive reading strategies. But most of them (55 students) were in the moderate level, some of them (18 students) were in the high level and only 1 student was in the low level.

• e Findings of Observation Data were also obtained from observation aided by a checklist and note-taking. After observation, the researcher counted up the slashed and

CM


100


tabulated the findings. e order from most-frequently-to-least-frequently-used metacognitive reading strategies is as follow: 1). Re-reading (124 times) 2). Underlining and circling (111 times) 3). Translating (English words into Chinese) (108 times) 4). Reading questions before starting reading (82 times) 5). Finger-or pen-pointing while reading (71 times) 6). Skimming or scanning (42 times) 7). Using a dictionary or vocabulary list (18 times) 8). Note-taking (17 times) 9). Asking classmates for help (11 times) 10). Lip reading (7 times) 11). Asking the teacher for help (3 times)

• e Findings of Interviews Participants said that they used translation from English into Chinese in order to understand the text. ey used contextual clues and a dictionary to help them check or find the meaning of unknown words or phrases. ey used their experience and general knowledge to help them predict what they read. ey did the exercises after each text to learn the meaning of unknown words and to test the degree of their understanding. ey paid attention to the topic sentence of each paragraph and synthe-sized the main idea of the whole text. ey read slowly and carefully in order to recognize the main points and supporting details. ey also stated that grammatical structures were not big problems in their reading, but unknown vocabulary and phrases or idioms in the text were still the cause of big problems. In most cases, if they knew the meaning of every word in a sentence, they had no problems in figuring out the general meaning of the sentence.

CONCLUSION

The objectives of this study were to investigate first-year Chinese college students’ metacognitive awareness of reading strategies and

try to answer the following questions: 1). What metacognitive reading strategies are first year college students aware of when they read academic materials such as textbooks? 2). What are their observed metacognitive reading strategies when they read academic materials? 3). How do they use these strategies? e findings showed the matacognitive reading strategies that what first-year college students were aware of are: background knowledge, context

CM


101


clues, checking understanding, guessing meaning, skimming and scanning, using tables and pictures, adjusting reading speed, reading slowly and carefully, guessing unknown words, re-reading, stopping and thinking, underlining and circling, translating and using dictionary, etc. e findings from questionnaire have also provided evidence that most students had the strong awareness of using: background knowledge, context clues, guessing meaning from the context, underlining and circling, translation, planning before reading, adjusting reading speed, reading slowly and carefully, getting back on track when distracted, checking understand-ing, guessing meaning, skimming and scanning, using tables and pictures, guessing unknown words, re-reading, stopping and thinking from time to time, going back and forth in reading, and using a dictionary. While the strategies they didn’t pay much attention to were: noting length and organization, noticing what to read closely and what to ignore, analyzing and evaluating the information, visualizing information, thinking about difficulties while reading, reading aloud to help them understand what they read, asking question of the text, and paraphrasing and note taking. eir observed metacognitive reading strategies were: re-reading, underlining and circling, translating (English into Chinese), reading ques-tions before starting reading, finger-or pen-pointing while reading, etc. However, some psychological activities, such as planning, thinking about the difficulty, using background knowledge, etc., could not be observed. erefore it is necessary to use a triangulation to collect the data. e findings from interviews revealed that students effectively or-chestrated various strategies in order to understand the text. ey used the strategies to plan, monitor, control, self-regulate, evaluate and remedy their comprehension. ey also used the strategies to solve problems and meet with their needs or purposes. In this study, participants relied on their native language (Chinese) while they were reading English. ey used Chinese and translated English into Chinese in order to facilitate input comprehension. ey used self-regulate reading strategies to skim the ques-tions before reading if they read for a test. ey saved time and energy by focusing on the important sentences and skipping the unimportant ones. However, if they read for learning, they preferred to read the vocabulary list before they started reading. ey said this helped them consider a new text as a whole. ey needn’t stop to look up the new words in the text while reading. Some students said that they liked to use finger-or pen-pointing while reading because it was convenient to help them underline and circle the important information or unknown words or phrases and sentences. Moreover, it enhanced their concentration.

CM


102


IMPLICATIONS

The findings of this study have provided evidence that the partici-pants had awareness of using a variety of metacognitive strategies to

facilitate self-management, self-regulation and self-evaluation to meet with their needs or purposes in their academic reading. is helps instructors in determining students’ strengths and weaknesses in terms of strategies use. is in turn will help in developing effective and appropriate strategy instruction that should permeate the whole curriculum. erefore, the following pedagogical suggestions might be provided from this study. First, exploring what reading strategies the students use at the be-ginning of instructions is a prerequisite. is benefits instructors get to know the students’ strengths and weaknesses so that they might provide students with opportunities to employ new strategies by direct and explicit instructions. is is especially helpful for the students who are struggling in reading. Second, for college level English studies, the participants who study college English also have a limitation of vocabulary. erefore, direct vocabulary instruction is still necessary in academic text reading. Instruc-tors could combine incidental and intentional vocabulary instructions to form an effective method to assist students to learn and retain vocabulary. Instructors could try associative metacognitive reading strategies such as guessing unknown words from context (prefix, suffix, antonym, given and new information, etc.) to enlarge students’ vocabulary. ird, the results have indicated that the participants were less aware of the following reading strategies: noting length and organization, noticing what to read closely and what to ignore, analyzing and evaluating the information, visualizing information, thinking about difficulties while reading, reading aloud to help them understand what they read, asking questions of the text, paraphrasing and note-taking. All of these involve in the strategies such as problem solving and decision making. erefore, for college students, critical reading strategies should be focused on read-ing instructions, which might strengthen students to solve problems by utilizing the strategies appropriately. To help students to be more aware of their metacognitive reading strategies plays an important role in developing students reading strategies. Only when students are aware of what they read makes sense, can they become more independent readers and gain confidence in reading.

CM


103


REFERENCESBlock, E. 1992. See how they read: comprehension monitoring of L1 and L2

readers. TESOL Quarterly, 26, p 319-342.Brown, A., and L. Baker, 1983. e role of metacognition in reading and

studying. p.45-75. In J. Orsany (ed). Reading comprehension: From research to practice. Hillsdale, NJ: Erbaum.

Garner, R. 1987. Metacognition and reading comprehesion. Norwood, NJ: Ablex.

Lynch, B.K. 1996. Language program evaluation. Cambridge, Cambridge University Press.

Mokhtari, K., and R. Sheorey, 2002. Measuring ESL students’ awareness of reading strategies. Journal of Developmental Education 25, no.3

Nelson, T.O. 1998. Forward: metacognitive food for thought in educational theory and practice, p1-24. In D. J. Hacker, J.Dunlosky and A. Graesser (eds.), Metacognition in educational eheory and eractice. Mahwah, NJ: Lawrence Erlbaum Associates.

Nuttall, C. 1996. Teaching reading skills in a foreign language. Oxford: Heinemann

Oxford, R., and D. Crookall. 1989. Research on language learning strate-gies: Methods, findings, and instructional issues. Modern Language Journal, 73, 404-419.

Rigney, J.W. 1978. Learning strategies a theoretical perspective. In H.F. O’Neil (Ed.), Learning strategies. New York: Academic Press.

Singhal, M. 2001. Reading proficiency, reading strategies, metacognitive awareness and L2 readers. e Reading Matrix, 1, 2001.

Vacca, R.T., and J.L. Vacca, 1999. Content area reading: Literacy and learning across the curriculum. (6th ed.). U.S.: Longman.

Winn W., and D. Snyder. 1998. Cognitive perspectives in psychology 112-142. In D. H. Jonassen, ed. Handbook of research for educational communications and technology. New York: Simon and Schuster Macmillanm.

CM


104


Appendix 1: Survey of Reading Strategies (SORS) Name: __________ Code: __________ Section: _________Faculty: __________ Major: __________ Tel. No: __________

Instructions: e purpose of this survey is to collect information about the various techniques you use when you read academic materials in English (e.g., reading textbooks for homework or examinations, reading journal articles, etc.). All the items below refer to your reading of college-related academic materials (such as textbooks, not newspapers or magazines). Each statement is followed by five numbers, 5, 4, 3, 2 and 1, and each number means the following: ‘5’ means that I always or almost always do this ‘4’ means that I usually do this. ‘3’ means that I sometimes do this (About 50% of the time.). ‘2’ means that I do this only occasionally. ‘1’ means that I never or almost never do this. After reading each statement, put a tick in the box that shows the number (5,4,3.2, or 1) which applied to you. Note that there are no right or wrong responses to any of the items on this survey.

Statement 5 4 3 2 1

1. I have a purpose in mind when I read.2. I take notes while reading to help me understand what I read.3. I think about what I know to help me understand what I read.4. I take an overall view of the text to see what it is about before reading it.5. When text becomes difficult, I read aloud to help me understand what I read.6. I think about whether the content of the text fits my reading purpose.7. I read slowly and carefully to make sure I understand what I am reading.8. I review the text first by noting its characteristics like length and organization.9. I try to get back on track when I lose concentra-tion.10. I underline or circle information in the text to help me remember it.

CM


105


11. I adjust my reading speed according to what I am reading.12. When reading, I decide what to read closely and what to ignore.13. I use reference materials (e.g., a dictionary) to help me understand what I read.14. When text becomes difficult, I pay closer attention to what I am reading.15. I use tables, figures and pictures in text to increase my understanding.16. I stop from time to time and think about what I am reading.17. I use context clues to help me better understand what I am reading.18. I paraphrase (restate ideas in my own words) to better understand what I read.19. I try to picture or visualize information to help remember what I read.20. I use typographical features like bold face and italics to identify key information.21. I critically analyze and evaluate the information presented in the text.22. I go back and forth in the text to find relation-ships among ideas in it.23. I check my understanding when I come across new information.24. I try to guess what the content of the text is about when I read.25. When text becomes difficult, I re-read it to increase my understanding.26. I ask myself questions I like to have answered in the text.27. I check to see if my guesses about the text are right or wrong.28. When I read, I guess the meaning of unknown words or phrases.29. When reading, I translate from English into my native language.30. When reading, I think about information in both English and my mother tongue.

ank you very much for your cooperation!

CM


106


Appendix 2: Observation Checklist of Metacognitive Reading Strategies

Name:__________ Date:_________ Class:______________

Global Strategies Number of the students’ reading actions (Use a slash if you see one )

SkimmingScanningPreview the textLook at the titleLook at the questions in the textRead closelyIgnore some of sentencesLook at print conventionsUse table, figure and pictureUse context clueProblem solving strategies

Read slowly and carefullyRe-readAdjust speedGo back on trackStop from time to time and thinkUsing a dictionaryAsking a teacher for helpAsking a classmate for helpTaking noteReading aloudUnderlining and circlingTranslating

Appendix 3: Interview Questions

1. When you are reading, if you encounter unknown word or phrases, what would you do?

2. When you are reading, what do you focus on most?3. Why do you translate what you read into Chinese?4. When do you use a dictionary?5. How do you use the context clues to help you guess the meaning of

the unknown words?

CM


107


6. Do you use your experience or general knowledge to help you com-prehend the content of the text?

7. What do you think about the grammatical structures?8. Please tell me about the problems in your reading.9. Why do you go back and re-read the whole passage? 10. Why do you want to read the questions before reading?

CM


109


Sumalee Sawatdirakpong*

Analysis of Dimensionality Underlying EFL Proficiencyof ai Students

Department of English, Faculty of Humanities, Chiang Mai University, Chiang Mai 50200, ailand*Corresponding author. E-mail:[email protected]

ABSTRACT

This study is part of a project to develop an EFL test bank and the main concern is test validity. For a test to be valid, it must

measure what it purports to measure with minimal measurement errors. Toward that end, EFL test papers, each consisting of three parts were developed, based on one set of blueprint. One of the EFL test papers (later referred to as the seed paper) was first administered to the same group of ai university students taking TOEFL. Various portions of the seed paper were later used as part of other EFL test papers as a means to equate them. e item data from the administrations of three different “equated” test papers were made into parcels, and the product-moment correlation matrices of the parcel scores were factor analyzed. It was found that in all three test papers, irrespective of analyzing each part separately, two parts combined, or three parts combined, three different factors corresponding to the three parts of the tests were consistently identified: those underlying grammar in context, indirect writing and reading. Certain reading tests, however, seemed to be bidimensional. ose of lower p-values tended to share the same underlying factor with item parcels made to measure writing, while those of higher p-values tended to share the same underlying factor with item parcels made to measure grammar in context. It can be concluded that the technique works quite well in checking that different sets of test items (made from the same blueprint) measure comparably the same underlying language competence, and this practice is useful in making parallel tests from a pool of test items statistically calibrated and stored.

Key words: Dimensionality, EFL, Proficiency, Valid, Item Data, Equate, Item Parcels, Factor Analyzed

CM


110


INTRODUCTION

Preparing test papers to measure second/foreign language proficiency from a pool of test items can be quite challenging, as there is need to

make sure that the test scores are reliable and valid. For the test scores to be reliable, measurement errors should be minimal or constant. e test scores should be consistent across test administrations, test forms, raters and other measurement contexts. An individual taking different test papers that have been equated should get comparably the same scores, when the test taking times are not too far apart and no special attempts have been made to change his/her level of language proficiency. Different test takers of similar levels of language proficiency should also get similar scores, no matter which papers they take or where they take the test. For a test to be valid, it must possess the quality that allows meaning-ful, appropriate and useful interpretations, inferences or decisions on the basis of the test scores (Hopkins et al., 1990). e test scores must reflect the same dimensionality (underlying traits/constructs measured by the test items), and this dimensionality should correspond to the theoretical framework of what it means to learn a second or foreign language. For any test to have validity, it should ideally measure only what it purports to measure and nothing else. e term dimensionality refers to the number of dimensions needed to summarize a data matrix that results from interactions of a set of test items and a group of test takers (Reckase, 1990). e word “dimension,” as used in this paper, is equivalent to the statistical or psychometric dimen-sion or factor. What does it mean to know or learn a language? Current theories of language proficiency take into account the notion of communicative competence which was first proposed by Hymes (1972) to denote the sociolinguistic features of communication, and later refined by others such as Canale and Swain (1980), Savignon (1986) and Bachman (1990). Language competence, including second and foreign, is theorized as com-prising separate but related dimensions. To communicate effectively in a language, one needs to be competent in these dimensions. He/she needs not only knowledge of the language, but also skills and strategies to perform appropriate communicative acts in various sociolinguistic contexts. Following theoretical models of communicative language compe-tence, language proficiency is by and large regarded as equivalent to com-municative competence or ability. It is viewed as integrative, involving many distinct but related traits. is has changed second/foreign language teaching and testing dramatically during the past decade. Language teach-ing has become more communicative in nature and language testing has

CM


111


moved toward authentic language use. All English as a Foreign Language (EFL) proficiency tests developed in this research consisted of three parts: grammar in context, indirect writing and reading. e test tasks were designed in such a way that they approximated authentic language use as much as possible, without compromising ease of mass scoring. Many factors can play a role in how well, or not well, a test taker performs on a test (Cziko, 1984), but the test scores should best reflect what the test purports to measure. In addition, when any test is administered on a regular basis, it is important to be able to justify that each parallel test paper used measures comparably the same dimensionality. erefore, in this ongoing project where EFL proficiency tests have been developed and administered on a regular basis, there is need to check if all test papers measure the same underlying dimensions of language proficiency (Acker-man, 1989). e objective of this research was twofold: to investigate if principal factor extraction with oblimin rotation could be used as a means to identify the dimensionality underlying the EFL test scores, and if the obtained dimensionality reflected the test blueprint and remained constant across administrations of equated test papers.

METHODSPreparation of Test Papers and Test Administrations

One EFL test paper (later referred to as “seed paper”) was developed, based on a well-established set of blueprint. (All test development in

this research went through all the meticulous procedures of test develop-ment: peer review, expert judgment, tryouts and improvements.) It was then concurrently administered with TOEFL (Institutional Paper) to the same group of test takers on two consecutive days. On Day 1, Part 1 of the seed paper was administered along with Section 2 of TOEFL. On the following day, Part 2 and Part 3 of the seed paper was administered along with Sections 3 (only the reading) of TOEFL. More EFL papers were developed, based on the same set of blueprint and with a portion of the seed paper in each of the three parts. Inclusion of the portions of the seed paper helped in the equating process of various test papers. ree “equated” EFL papers were selected for further research to identify the dimensionality of the test. Each of the three equated EFL test papers was administered to a different group of graduate students of Chiang Mai University. One paper was given in May 2000, another in June 2001 and the other one in June

CM


112


2002, assuming a group of students of similar profile taking the test at approximately the same time of each year.

Data Analysis From the item level data which were coded into zero and one, descriptive statistics at both the test-level and item-level were computed. en, to avoid the possibility of unreliability of item level data and to avoid the many drawbacks of using either phi or tetrachoric correlation matrices in factor analysis (Knol & Berger, 1991), the items in each part of the EFL tests were grouped into parcels. e items which were based on the same dialogue, text or passage were made into the same parcels, and the matrices of the product-moment correlation of the parcel scores were analyzed via principal components analysis (with unities in the diagonals) to obtain eigenvalues. Plots of eigenvalues were then examined to decide on the num-ber of factors to extract. e distinctive “elbow-bend” point in a plot was used as a rough indication as to how many factors to extract. With specified numbers of factors to extract, each matrix was again analyzed using principal axis factor analysis with iteration until the communalities converged. Because the correlated psychological constructs may underlie language proficiency (Bachman, 1990), an oblique rotation (Oblimin) of the factors was performed to aid in the interpretation of the factors. e final decision as to the best solution and the number of substantial factors was made using a composite of criteria (Gorsuch, 1983). Only the loadings of at least an absolute value of .3 were considered substantial, and only a factor with at least 2-3 substantial loadings was further considered for substantialness.

RESULTSItem and Test Descriptive Statistics

For all three tests the proportion-correct p-values of most items were below .5, suggesting that overall the tests were a bit challenging for

the test takers. is was also evidenced by the low mean scores and the positively skewed score distributions (as shown in Table 1). e biserial correlation coefficients of most of the items were acceptable (higher than .3). e (coefficient alpha) reliability estimates of all three parts of the test and the total test, corrected for attenuation, were all moderately high (ranging from .76 to .87, as shown in Table 2), considering that the num-bers of test takers were between 246-267. e mean values of all statistics were not significantly different across the three test papers. e Pearson

CM


113


product-moment correlations among part scores and total scores were all large and positive, and the highest correlations were found between the reading part and the total test for all test papers.

Table 1: Descriptive Statistics for Test Scores for Each Part of the Three Tests. Part 1 (k=20) Part 2 (k=30) Part 3 (k=20)

Test 1 (n=246) Mean 13.06 12.87 17.72sd 4.07 6.27 6.04alpha .71 .67 .64skewness .45 .51 .41

Test 2 (n=252) Mean 14.17 14.26 18.5sd 4.67 5.94 6.61alpha .73 .68 .74skewness .31 .43 .32

Test 3 (n=267) Mean 14.94 14.12 18.27 sd 4.63 5.62 6.72 alpha .74 .69 .81skewness .41 .45 .44

n=number of examinees in each group k=number of items in each part of the test

Table 2: Pearson Product-Moment Correlation Coefficients Among Part and Total Scores for each Testa.

Test 1 Part 1 Part 2 Part 3 TotalPart 1 (.71) .56 .53 .63Part 2 (.67) .60 .66Part 3 (.64) .68Total (.76)



a Coefficient alpha reliabilities are in parentheses along the diagonal for each part. All correlations were corrected for attenu ation.

CM


114


At a One-Part Level When analyzing one part of the tests separately, plots of eigenvalues strongly suggested a one-factor solution for each of the three parts of the tests. Despite such a clear indication of one single factor, two-factor extrac-tion with an oblique rotation was performed for confirmation. Clearly, two-factor solution analysis of each part separately revealed that all parcels in each of the three parts loaded substantially on a single factor (as shown in Table 3). is indicated that all item-parcels in Part 1 of all three test papers measured grammar in context, all item-parcels in Part 2 measured indirect writing, and all item-parcels in Part 3 measured reading, as speci-fied in the test blueprint.

Table 3: Oblimin Rotated Factor Matrices, Communalities and Related Statistics. Part 1

Parcel Factor 1 Factor 2 Communalities Test1 P11 .83 -.07 .60

P12 .78 .01 .63P13 .79 .00 .64P14 .75 .01 .58P15 .68 .13 .61P16 .64 .13 .56

eigenvalue 5.74 .21% of variance 57.40 2.10

Test2 P11 .77 -.04 .57P12 .74 .06 .58P13 .73 -.02 .52P14 .79 -.12 .58P15 .75 .00 .56P16 .75 -.13 .52

eigenvalue 5.23 .15 % of variance 52.30 1.50

Test3 P11 .74 -.15 .57P12 .77 -.11 .60P13 .74 -.09 .56P14 .73 -.09 .55P15 .73 .05 .54P16 .66 .01 .43


CM


115


Part 2

Parcel Factor 1 Factor 2 Communalities Test1 P21 .75 .10 .58

P22 .71 .08 .52 P23 .62 .25 .48 P24 .71 -.18 .52 P25 .77 .03 .60


Test 2 P21 .80 .04P22 .62 -.13 .53 P23 .73 .03 .50 P24 .80 .09 .54 P25 .73 -.08 .63


Test3 P21 .67 .17 .55 P22 .70 .05 .51 P23 .57 .29 .51 P24 .66 .02 .44 P25 .78 -.02 .60


Part 3

Parcel Factor 1 Factor 2 Communalities Test1 P31 .72 -.01 .51

P32 .77 -.02 .58 P33 .76 -.17 .55 P34 .77 -.14 .56


Test2 P31 .93 -.18 .64 P32 .63 .16 .57 P33 .75 .03 .60 P34 .70 .09 .59



CM


116


Test3 P31 .73 -.18 .56 P32 .76 .03 .58 P33 .73 -.09 .55 P34 .78 -.16 .63


At a Two-Part Level At a two-part level, plots of eigenvalues suggested more than one factor. us, both two-factor and three-factor solutions were carried out, but two-factor solution yielded more meaningful results. For Parts1 and 2 combined, two-factor solution revealed two separate factors, one underlying all item-parcels in Part 1 and the other underlying all item-parcels in Part 2. is was true for all three test papers. For Parts 1 and 3 combined, two-factor solution also revealed two separate factors for all three test papers. For Parts 2 and 3 combined, all item-parcels in Part 2 substantially loaded on one single factor, but item-parcels in Part 3 did not. Parcels at the end of the test tended to deviate, and similar patterns were found for all three tests.

At a ree-Part Level Like at the two-part level, both two-factor and three-factor solutions were carried out. For all three tests, two-factor solution cleanly separated the item parcels into two factors, one underlying Part 1 and the other underlying Parts 2 and 3 together. However, in three-factor solution, all item-parcels in Part 1 substantially loaded on a single factor, and all item-parcels in Part 2 substantially loaded on another single factor. It was the item-parcels in Part 3 (reading) that deviated: one with low average p-values tended to group with item-parcels in Part 2 (indirect writing) and one with high p-values with item-parcels in Part 1 (grammar in context).

DISCUSSION AND CONCLUSION

The mean test statistics and mean item statistics were not significantly different across the three test papers. e moderately-high correla-

tion coefficients between parts of the test and the whole test supported the theoretical model that language competence comprises separate but related dimensions. Assuming a similar profile of the test takers, there was no need to be concerned about how functional the three test papers were in terms of reliability.

(Table 3. Continued).

CM


117


Based on the results above, the three EFL test papers consistently shared the same overall loading pattern, as summarized in Table 4.

Table 4: A summary of factor loadings of the three EFL tests under investigation.Test Factor 1 Factor 2 Factor 3

Test 1 Part 1 Part 2 Part 3Test 2 Part 1 Part 2 Part 3Test 3 Part 1 Part 2 Part 3

It is important, however, to stress that the number of factors to extract while analyzing all three parts combined could significantly affect the number of factors or dimensions identified. In this research both 2-factor and 3- factor solution were attempted. While the three-factor solu-tion, which was more meaningful considering a set of multiple criteria, resulted in three different factors, the two-factor solution resulted in only two factors. Part 2 and Part 3 turned out to share the same underlying dimension. How could indirect writing and reading share the same underlying dimension, when the two skills are very different—one is productive, the other receptive? Perhaps, it had a lot to do with the test format used in assessing indirect writing, which was cloze procedure. When doing cloze procedure, the test taker had to read actively in order to understand the text so that he/she could complete the blanks correctly (Bachman,1985). What the test taker brought into reading the cloze passage could well approximate what he/she brought into normal reading. Obtaining the same or different underlying constructs of indirect writing and reading was not an issue in this study because the results were consistent across the three tests. Using two-factor solution, the two parts shared the same underlying construct, and using three-factor solution they separated into two factors. However, since two-factor solution of each part separately and two-factor solution of two parts combined consistently yielded a separate factor for each part, for the purpose of checking construct validity, it would be justified for this ongoing project of test development to adopt three-factor solution when analyzing three parts combined. With some powerful statistical packages widely available, validation studies of EFL proficiency tests via principal factor extraction can be done quite conveniently and cost effectively. ere are other complicated tech-niques, but those are unnecessarily more expensive and difficult to use.

CM


118


REFERENCESAckerman, T. A. 1989. An alternative methodology for creating parallel

test forms using the IRT information function. Paper presented at the annual meeting of the National Council on Measurement in Education, San Francisco, CA.

Bachman, L. 1985. Performance on cloze tests with fixed-ratio and rational deletions. TESOL Quarterly, 19, 535-556.

Bachman, L. 1990. Fundamental considerations in language testing. Oxford: Oxford University Press.

Canale, M., and M. Swain. 1980. eoretical bases of communicative approaches to second language teaching and testing. Applied Lin-guistics, 1, 1-47.

Cziko, G. A. 1984. Some problems with empirically-based models of com-municative competence. Applied Linguistics, 5, 23-38.

Gorsuch, R. L. 1983. Factor analysis. (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum Associates.

Hopkins et al. 1990. Educational and psychological measurement and evaluation. Englewood Cliffs, New Jersey: Prentice-Hall, Inc.

Hymes, D. 1972. On communicative competence. p. 269-293 In J. B. Pride and J. Holmes (eds), Sociolinguistics. Harmondsworth, Eng-land: Penguin.

Knol, D. L., and P. F. Berger. 1991. Empirical comparison between factor analysis and multidimensional Item Response models. Multivariate Behavioral Research, 26, 457-477.

Reckase, M. D. 1990. Unidimensional data from multidimensional tests and multidimensional data from unidimensional tests. Paper pre-sented at the annual meeting of the American Educational Research Association, Boston, MA.

Savignon, S. 1986. e meaning of communicative competence in rela-tion to the TOEFL program. p. 17-30. In C. W. Stansfield (ed.), Toward Communicative Competence Testing: Proceedings of the Second TOEFL Invitational Conference (TOEFL Research Report 21). Princeton, NJ: Educational Testing Service.

CM


119


Nantarat Puengpang1*, Vantipa Roadrangka1 and Bronwen Cowie2

Development of an Instructional Set for a Laboratoryin Biology Course for First-Year ai Science

Student Teachers

1Faculty of Education, Kasetsart University, Bangkok, 10900, ailand2CSTER, University of Waikato, Hamilton, New Zealand*Corresponding Author. E-mail: [email protected]

ABSTRACT

The objectives of this case study research are (1) to study the in-structor’s regular practice and beliefs in teaching and learning

in the Laboratory in Biology Course and investigate first-year ai science student teachers’ biology concepts, the nature of science concepts, science process skills and attitudes towards biology after they have participated in the course, (2) to create the guiding principles to develop an instructional set based on a learner-centered approach for the Laboratory in Biology Course from the findings of the first objective of the study, the National Education Act, 1999, and a review of documents and related research reports, and (3) to develop an instructional set for the Laboratory in Biology Course for science student teachers based on the guiding principles generated from the second objective. e research study was conducted from June 2003 to October 2005. e findings revealed that the regular practice in the Laboratory in Biology Course was teacher-centered and emphasized the promotion of biology concepts. e participant instructor believed that science student teachers must have correct concepts before doing laboratory activities. After the science student teachers participated in the course, their biology concepts were mostly in form of partial understanding and specific misunderstandings were found in every topic. e science student teachers’ nature of science concepts were mostly appropriate. Most science student teachers had bet-ter skills in interpreting data and drawing a conclusion than identifying problems, formulating hypotheses, identifying variables and designing an experiment. eir attitudes towards biology were mainly found to be posi-tive. From these findings, a consideration of the National Education Act, 1999 and a review of documents and related research reports, the guiding principles for the development of an instructional set for the Laboratory in Biology Course based on a learner-centered approach were generated. e instructional set comprised an instructor manual, a student activity

CM


120


book and a student thinking book. e instructional set was evaluated by experts and implemented by an instructor experienced in teaching in the Laboratory in Biology Course during November 2005 to March 2006. Initial findings highlight the role of instructors’ beliefs, science student teachers’ motivation, content coverage and laboratory preparation.

Key words: biology, instructional set, laboratory, a learner-centered approach, science student teachers

INTRODUCTION

Effects of ailand’s economic crisis in 1997 revealed an important problem with ai education that ai citizens lacked logical and

critical thinking skills and had insufficient understandings of the natural world. is affected all levels of education, especially in science subjects. In response to the crisis, the ai government stipulated the National Education Act in 1999 (Office of Education Council [OEC], 2001). e Act emphasizes on learning reform as being at the heart of educational reform. A learner-centered approach is regarded as an approach which may possibly yield productive teaching and learning. e Act describes learners as most important participants in the learning process who should have an opportunity to develop themselves at their own pace and to the best of their potential. Several later studies of teaching and learning in ai tertiary education (see for example, OEC, 1999; Polsaram and ephasdin Na Ayudhaya, 2000; Areekul, 2005) also revealed many problems, including inflexible and out-of-date curriculum, unproductive teaching and learn-ing activities dominated by a teacher-centered approach, insufficient and low-quality textbooks, and inadequate assessment strategies. From this, it can be inferred that the National Education Act, 1999, has not yet been put into general practice in classrooms. A learner-centered approach, delineated in the Act, served as a guid-ing principle of an instructional set for the Laboratory in Biology Course. First-year science student teachers were participants as they were regarded as crucial personnel for the future implementation of the National Education Act recommendations. e research study consists of two phases: Phase I: the exploratory phase and Phase II: the development, implementation, and evaluation of the instructional set. Phase II data analysis is yet to be completed. is report describes only the results of Phase I and the development and implementation of the instructional set and preliminary findings of Phase II.

CM


121


Research Objectives e study is designed to achieve the following objectives: 1) to study the instructor’s regular practice and beliefs in teaching and learning in the Laboratory in Biology Course and investigate first year science student teachers’ biology concepts, the nature of science concepts, science process skills and attitudes towards biology after they had partici-pated in the course, 2) to create the guiding principles to develop an instructional set based on a learner-centered approach for the Laboratory in Biology Course from the findings from the first objective of the study and the National Education Act, 1999, and a review of documents and related research reports, 3) to develop an instructional set for the Laboratory in Biology Course for the science student teachers based on the guiding principles generated from the second objective.

Methodology e research aims at studying the regular practice and finding instructor’s beliefs in teaching and learning in the Laboratory in Biology Course and gathering data about the learning process of the science student teachers as well as their conceptions, science process skills and attitudes which are viewed as learning outcomes in science. A case study was em-ployed as the design of this research. e case was situated in a laboratory class included twenty-nine science student teachers and an instructor. e case focused on the phenomena which occur in the teaching and learning during the academic semester of the implementation of the instructional set. e first researcher participated in the classroom activities as a teacher assistant and observed every class session during the semester in which the instructional set was implemented.

Phase I: Exploratory Phase Exploring Regular Practice and Instructors’ Beliefs in Teaching Regular practice in the Laboratory in Biology Course was investi-gated by classroom observation in the first semester of the 2003 academic year. e science student teachers who participated in the course in the semester were asked to respond to a questionnaire about their perceptions of teaching and learning in the course. is questionnaire consisted of two parts. e first part was a rating scale asked them to indicate the quality of teaching and learning activities, laboratory apparatus, class environ-ment, and instructor’s help. e second part was in an open-ended format asked about the science students teachers’ feelings of, and experiences from the Laboratory in Biology Course and suggestion to improve the

CM


122


teaching and learning. During the semester, a semi-structured interview was used to elicit the instructor’s beliefs in teaching. Questions in this semi-structured interview asked about instructor’s aims in teaching in the Laboratory Course, how she taught and assessed science student teachers’ learning and what she thought about how they learned. e questions in the questionnaire and the semi-structured interview were checked by one science educator.

Finding out Concepts, Skills and Attitudes of Science Student Teachers who Completed the Laboratory in Biology Course To find out science student teachers’ biology concepts, the nature of science concepts, science process skills and attitudes towards biology, laboratory learning and the Laboratory in Biology Course, two question-naires and an inventory were developed: the Biology Concept and Science Process Skill Questionnaire, the Nature of Science Concept Questionnaire, and the Attitude towards Biology Inventory. e Biology Concept and Science Process Skill Questionnaire con-sisted of 10 items with two types of question formats, multiple-choice and open-ended. e first eight items asked about basic biology concepts taught in the Laboratory in Biology Course. ese concepts included cell structure and function, prokaryotes and eukaryotes, osmosis, cell divi-sion, reproduction, photosynthesis, respiration and ecology. e last two items, developed by the first researcher, asked about integrated science process skills; formulating hypotheses, identifying and controlling variables, designing an experiment, interpreting findings and drawing conclusions. e biology concept part of the questionnaire was developed by reviewing the manual of the Laboratory in Biology Course and the literature. e majority of the questions (Questions 1, 2, 4, 6, 8, 9 and 10) were generated by the first researcher. All of these questions were in contexts relevant to the ai science student teachers. e other questions were based on research papers: Question 5 was adapted from Lewis et al. (2000) and Question 7 was adapted from Barker (1985). Question 3 was developed by adapting the knowledge question type, typically found in general biology textbooks, and sought to assess the concept without a context. e Nature of Science Concept Questionnaire was developed to assess the science student teachers’ views of four aspects of the nature of science: 1) the definition of science, 2) the characteristics of scientific knowledge and methods, 3) the characteristics of scientists and 4) the influence of science and society. is questionnaire was adapted from Views-on-Sci-ence-Technology-Society (VOSTS) developed by Aikenhead et al. (1987). irteen statements about the nature of science were selected from the VOSTS statements. Two items assessed conceptions of the definition of

CM


123


science, 5 items assessed the characteristics of scientific knowledge and methods, 4 items assessed the characteristics of scientists, and 2 items assessed the influence of science and society. e questionnaire asked science student teachers to read each statement and consider whether they ‘agree’, ‘not sure’ or ‘disagree’ with the statements and also to provide an explanation in support of their ideas. e Attitudes towards Biology Inventory sought to measure science student teachers’ attitudes towards biology, laboratory learning and the Laboratory in Biology Course. ere were 10 positive and 10 negative items, totaling 20 items. e items were adapted from a number of research papers (for example, Ebenezer and Zoller, 1993; Moore and Foy, 1997; Francis and Greer, 1999; Jesky-Smith, 2002; Salta and Tzougraki, 2004). Science student teachers were asked to select their answers from a scale consisting of the choices ‘strongly agree’, ‘agree’, ‘not sure’, disagree and ‘strongly disagree’ to indicate their attitudes towards each statement. ese two questionnaires and the inventory were translated into the ai language and reviewed by experts; four university science instructors and four science educators, and revised according to the feedback. e Biology Concept and Science Process Skill Questionnaire was piloted with 15 second-year science student teachers who had completed the Labora-tory in Biology Course in the first semester of the 2003 academic year. e Nature of Science Questionnaire and the Attitudes towards Biology Inventory were piloted with 19 third-year science student teachers. e questionnaires and the inventory were improved again to change the word-ing of some questions to make them clearer to the respondents. en, the questionnaires and inventory were administered to 36 first-year science student teachers who had participated in the Laboratory in Biology Course in October 2004. e science student teachers’ responses of biology concepts were analyzed by content analysis using predetermined categories similar to the categories of student understanding developed by Simpson and Marek (1988). e categories comprised scientific understanding (SU), partial understanding (PU), partial understanding with specific misunderstand-ings (PS), specific misunderstandings (SM) and no answer (NA). e responses to the science process skills and the nature of science concepts were all read carefully and similar responses grouped and presented as a percentage. e attitude responses were analyzed by rating-scale analysis for mean score and standard deviation.

CM


124


e Results of Phase I Regular Practice and Instructors’ Beliefs in Teaching e objectives of the course, as outlined in the course syllabus, were to promote science student teachers’ understanding of biology concepts; to promote observational skills and creativity to initiate own experiments, and to give science student teachers experiences of learning from real speci-mens. In every laboratory session, the instructor gave a lecture to revise the biology concepts. en the procedures of the laboratory activities were presented, demonstrated and explained. Next, the science student teachers completed the activities by following the laboratory manual. At the end of the session, the instructor usually gave the results of the laboratory activities and asked questions to check if all science student teachers got the same correct results. While the science student teachers were doing the experiments, they were often confused about the procedural steps even though the instruc-tor had given instructions before they began the experiments. ey often waited to copy the correct results from the instructor and did not pay much attention to the appropriate way of doing the experiments and using the equipment. At the beginning of some classes, the instructor used a quiz for a formative purpose to assess the science student teachers’ understandings of concepts taught in the previous lessons. She also believed that these quizzes would help the science student teachers in paying attention to review the lessons. e quizzes did not count for the science student teachers’ final mark. e midterm and final examinations counted for 50% and 50% of the final mark, respectively. e quiz and examination items were in a short-answer form and asked about biology concepts and the steps to do the laboratory activities that the science student teachers had learned in the course. e instructor’s aims for her teaching were to allow the science student teachers to experience science from real things and be able to pass the examinations. e laboratory activities were for reinforcement of the lectures. She believed that science student teachers should have the correct concepts before they did laboratory activities. e science student teachers’ perceptions of the learning objectives were consistent with those of the instructor. ey expected the laboratory to help them understand biology concepts better, to help them to remember knowledge for tests and to learn from real things. A majority indicated that the laboratory apparatus were in good conditions and given in an adequate number for each group. With the instructor’s sense of humor and hospitality and good help, most of them were pleased with the laboratory activities, even though they had difficulties with some of the practical skills and in taking the examinations.

CM


125


Biology Concepts Percentage of the science student teachers’ responses to the biology concept part of the Biology Concept and Science Process Skill Question-naire is shown in Table 1.

Table 1: Summary of Responses of the Science Student Teachers’ Biology Concepts (N=36).

ConceptsCategories of Responses (%)

SU PU PS SM NA

1. Cell structure & function 6 64 19 3 82. Prokaryotes and Eukaryotes 2.1 Prokaryotes 2.2 Eukaryotes

3028

3953

116

1711

33

3. Osmosis 33 17 6 28 174. Application of Osmosis 33 25 6 33 35. Cell division 5.1 Organisms having mitotic cell division 5.2 Organisms having meiotic cell division 5.3 Mitosis: genetic information in daughter cells 5.4 Meiosis: genetic information in daughter cells

-

-

-

-

58

39

25

47

6

19

-

-

14

22

19

28

22

19

56

25

6. Reproduction 6.1 Sexual reproduction 6.2 Asexual reproduction

-67

56-

86

617

3111

7. Photosynthesis: starch test 6 75 - 3 178. Ecosystem: meaning - 72 - 25 3

1SU = Scientific Understanding2PU = Partial Understanding3PS = Partial Understanding and Specific Misunderstanding4SM = Specific Misunderstanding5NA = No Answer Results shown in Table 1 indicated that a majority of the 36 re-spondents had a partial understanding (PU) of most concepts. No scien-tific understanding (SU) was found in three topics: cell division, sexual reproduction and the meaning of an ecosystem. Responses with specific misunderstandings (SM) were found in every topic. e science student teachers were likely to give no responses (NA) for osmosis, cell division and sexual reproduction because they could not explain their responses but rather repeated what was already given in the question. Details of the

CM


126


findings of the first-year science student teachers’ biology concepts were presented in Puengpang et al. (2006).

Science Process Skills To assess their science process skills, the science student teachers were asked to design an experiment to test the effectiveness of three bird flu medicines on a group of 1200 infected chickens. ey were given a blank to fill anything they considered necessary to do the experiment. None of them posed a question as a basis for doing any experiments. ey listed the steps of the procedure. Very few of them (8%) formulated hypotheses and those that were given were not appropriate or illogical. Most of the experimental designs lacked a control group and experimental replication. e majority of the science student teachers did not clearly detail the data collection methods and how to draw a conclusion. A typical response was to say, ‘collect the data and conclude the results’. e science student teachers were also given graphs illustrating the results of an experiment to investigate the rate of activity of four enzymes at different pH levels. ey were asked to pose a question, identify the variables, results and conclu-sion for these graphs. It was found that about 50% of the science student teachers posed an appropriate question, correctly identified dependent and control variables. Sixty-seven percent of them could formulate an appro-priate hypothesis for the experiment. A majority correctly responded to sub-questions that asking about the results (78%) and the conclusion (75%) of the experiment. In contrast, only 5% identified the correct independent variables.

e Nature of Science Concepts e results of the science student teachers’ responses to the Nature of Science Concept Questionnaire can be summarized as follows: 1) 98% of them thought science was understandable because sci-ence was from human understanding, thinking processes and experimenta-tion 2) 98% of them thought science could change over time 3) 91% of them agreed that scientific discoveries could be from accident 4) 86% of them agreed that society influenced the direction of scientific discoveries in that discoveries were made by demand from society and fitted with the tradition and beliefs of people in a society 5) 83% of them thought that imagination and creativity could be used in getting results and 14% asserted that further verification must be made 6) 75% of them believed that most scientists were logical, objec-tive, unbiased and open-minded while 22% thought that a minority of

CM


127


scientists held strong personal beliefs and did not accept others’ ideas 7) 72% of them expressed views showing a belief in the limitations of science whereas 28% thought science could explain everything, inclu-ding the supernatural 8) 72% of them noted that scientific methods could not always provide valid, clear and logical results in that errors could occur during experimentation and data interpretation 9) 69% of them thought curiosity motivated scientists to continue their work, 14% thought curiosity and societal issues motivated them, and 8% thought scientists worked only for the society, not themselves 10) 67% of them believed that personal feelings and moral values influenced scientists’ decision making while 25% thought scientists rely on only evidence or truths, not feelings and values 11) 58% of them viewed science as a product in that science was whatever could be logically explained, while 42% viewed science as process and product in that they noted science was also related to experiments and thinking processes 12) 36% of them thought that possible harms that could come from scientific discoveries were the responsibility of the scientists, 14% noted that those who used the discoveries were responsible, and 25% did not identify who was responsible but rather discussed the morality of using scientific discoveries As the results shown above, a majority of ai science student teachers in the study had proper views of the nature of science in most aspects. However, the view that science was only product, scientists must be responsible for possible harms from their discoveries, and personal feelings and values could not be part of scientists’ decision making, were also held by more than 25% of them.

Attitudes towards Biology, Laboratory Learning and the Course e results from the Attitude towards Biology Inventory showed that the science student teachers had positive attitudes towards biology in that the majority of them thought biological knowledge was regarded as important and useful for solving everyday life problems and for country development. Everyone in the society should have biological knowledge for better living and biology knowledge could be used in their own lives. ey also had positive attitudes about laboratory learning in that they thought doing laboratory work was useful and enjoyable. ey would prefer to do experiments to answer their questions rather than asking from other people and the experiments would help them learn more than the instructor’s teaching. e Laboratory in Biology Course was found to be useful to help the science student teachers understand biology concepts and

CM


128


promote the confidence to use scientific methods and apparatus. However, the science student teachers felt that handling scientific apparatus made biology diffi-cult and they were very uncomfortable with doing a biology experiment individually. In addition, they viewed biological knowledge as a quite serious cause of many world problems.

Guiding Principles for Development of the Instructional Set e guiding principles for the development of the instructional set for the Laboratory in Biology Course were generated from the literature review of learner-centered and constructivist teaching approaches, labora-tory teaching and learning as well as the main findings of phase I. ese principles can be presented as follows. 1) All learners should be encouraged to develop their competence in science according to their interests and to the best of their potential. 2) Productive science learning can be promoted by multiple, active and challenge learning activities with various instructional materials. 3) Laboratory work can be assessed by both summative and forma-tive assessment. 4) Possessing correct scientific concepts, understanding of the nature of science, acquisition of science process skills and positive attitudes towards science are regarded as characteristics of a scientifically literate person. 5) e understanding of nature of science and acquisition of science process skills are effectively taught by an explicit approach. 6) Prior knowledge is important for learners in learning of new knowledge 7) Social interaction facilitates successful science learning, hence, communicating of ideas and group work are considerably cooperated in the learning process. 8) Instructor is the creator of an active, challenging and collaborative learning environment and scaffolds learners’ thinking skills. 9) Finding out the learners’ concepts, skills and attitudes after learn-ing benefits in designing appropriate teaching and learning activities for the Laboratory in Biology Course.

Phase II Development, Implementation and Evaluation of the Instructional Set

In Phase II, the instructional set for the Laboratory in Biology Course was developed based on the ten guiding principles presented in the previous section. It was implemented from November 2005 to March 2006 and its effectiveness was evaluated. e outline of Phase II is shown in Figure 1

CM


129


Figure 1: Outline of Phase II. Development of the Instructional Set To develop the instructional set, twelve lesson plans were created based on the ten guiding principles. e topics and the sequence of content of the lesson plans followed the Laboratory in Biology Course’s manual. e teaching and learning activities have been adapted from many sources such as the laboratory manual, and science education websites. Some activities were created by the first researcher. Various teaching strategies were incorpo-rated; practical laboratory activities, planning and critique of experimental designs, the conduct of an investigation, presentations, the exchange of ideas and discussion, reflection, and debates. e instructional materials used were samples of organisms, wet mount and permanent slides, word cards, articles to be read, CD-ROMs, and interactive animations. e instructional set was comprised of an instructor manual, a student acti-vity book and a student thinking book. e student thinking book was designed so that at the end of every laboratory session the science student teachers would be able to write down what they had learned, their feelings about their learning, and how they could improve their learning.

Evaluation of the Instructional Set by Experts A meeting was held to evaluate the instructional set in October, 2005. e purpose of the meeting was to consider whether the instruc-

Guiding principles for the instructional set forthe Laboratory in Biology Course

Development of the instructional set

Evaluation of the instructional set by experts

Improvement of the instructional set

Implementation of the instructional set

Evaluation the instructional set duringand after the implementation

Data from Phase I Guidelines inThailand's NationalEducation Act, 1999

Information fromthe literature review

CM


130


tional set were appropriate and could be possibly used in the real labora-tory classroom. Two instructors experienced in teaching the Laboratory in Biology Course, the course coordinator, one science educator and the first two researchers were participants at the meeting. e instructional set, as well as the underlying theories used to frame them, were presented and explained. A summary of the guidelines in the National Education Act, constructivist perspectives of learning, and the expected science con-tent, the nature of science concepts, skills and attitudes outcomes were presented. e researchers answered questions related to ideas underlying the development of the instructional set. An overview of the instructional set was presented and then each lesson was discussed in turn. From this meeting, beliefs about teaching as well as comments and suggestions about the instructional set from the experienced instructors and the course coordinator were captured. Possible problems and obstacles were discussed. e teaching and learning activities were then improved and adjusted according to the instructor’s suggestions and researchers’ de-liberation. e revised instructional set was presented and discussed again before implementation. is meant that the final draft of the instructional set drew from current research in science education, the guidelines in the ailand National Education Act and the experiences of those involved in the Laboratory in Biology Course.

Preparation of Instructional Materials before the Implementation Although most of the instructional material for the instructional set was available for regular activities in the course, the first researcher made sure this was prepared and assembled. For instance, she checked the num-ber and condition of the materials. Other preparations concerned working with the laboratory staff to check the condition of light microscopes, the microscope linked with the screen in front of the class, and the process for setting up a computer with Internet access within the laboratory classroom. e extra materials mentioned in the instructional set were gathered and prepared such as cardboard for presentations, pictures in transparencies, and word cards. e first researcher also tested out some of the samples to be used in laboratory activities such as the investigation of chlorophyll pigments in plant leaves, and the classification of C3, C4 and CAM plants from the inner structure of a leaf section.

Implementation of the Instructional set e instructional set was implemented in the Laboratory in Biology Course which is a one-credit course. e course consists of one three-hour laboratory session per week. Twenty-nine science student teachers participated in the lab section which was studied for this research. e instructional set was implemented by an instructor who is experienced in

CM


131


teaching the course. e instructional set was implemented in the second semester of the 2005 academic year (November 2004 – March 2005) in a university in Bangkok. During the implementation of the set, the first researcher was a participant-observer as a teacher assistant. Data were ga-thered by observation and interviews with a focus on describing the science student teachers’ learning and the instructor’s teaching. Video and audio recorders were used to capture the data.

Ongoing Evaluation of the Instructional Set during and after Imple-mentation After the implementation of each lesson plan, the data from observa-tions of the teaching and learning activities, responses in worksheets the student thinking books were discussed in the weekly meeting between the instructor and the first researcher. is data were used to improve the lesson plan for the coming week to match with the situations and the level of science student teachers. Unstructured interviews with the instructor were used to elicit the instructor’s views of the implemented set and to provide suggestions for the improvement of the teaching and learning activities of the implementation of next lesson plan. Science student teachers were selected at random to be interviewed at the end of each class session. ey were asked to describe their feelings during the learning and for their sug-gestion to improve the instructional set. e improvement of each lesson plan was negotiated between the instructor and the first researcher with a check by two science educators. After the course finished, eight science student teachers were selected to be interviewed to express and summarize their ideas and experience of learning during the course. Within the semi-structured interviews, these science student teachers elaborated on the responses they had written at the end of the last laboratory session. Fifteen science student teachers were asked to complete the three questionnaires, constructed in Phase I to compare their achievement with those of the science student teachers who completed the regular course. Data from the observations, interviews and related documents were also analyzed to determine the effectiveness of the instructional set.

e Preliminary Results of Phase II Only some of the data has been analyzed at this time. e aspects which might possibly influence teaching and learning that have emerged from the data and contexts are as follows:

e Role of Instructors’ Beliefs Teachers’ beliefs had been found to be important in the planning and practice of teaching and learning activities (Cronin-Jones, 1991; Bryan,

CM


132


2003; Veal, 2004). is was found to be the case in this research study. e instructor began to implement the instructional set with doubts in her mind about whether she would be able to follow the lesson plans. For her, the activities seemed to take to a lot of time. She did not necessarily ask questions to elicit science student teachers’ prior knowledge or discuss their experimental results. She considered these activities would encroach on the time she had to run her brief lecture. e first researcher asked her to risk trying the new learning activities and she was open-minded enough to try. By the end of the course, she considered the instructional set had helped her reduce the time needed for the lecture at the beginning of the class and provided her an opportunity to listen to science student teachers’ ideas that she had not considered before.

Science Student Teachers’ Motivation Both the instructor and the science student teachers agreed that marks and grades strongly affected the motivation in learning. Science student teachers considered learning activities or work, which counted as part of a final grade was more important. A representative comment about the importance of marks follows: R: When you do activities in group, if a member does not partici-

pate and fails to help other members of the group, what would you do?

S1: e instructor should tell the science student teacher that there will be a test. When mentioning about a test, everyone will pay attention to study.

R: If there were no test, you won’t pay attention? S1: It’s not like that. S2: It’s just motivation. R: Is there anything else which can motivate you to learn? Marks

and...? S1: Marks and. Marks only.(R: researcher, S: science student teacher) e instructor also supported this idea, as she said in the third weekly meeting, “Ninety-nine percent of students think marks are very important. Generally, they are likely to enroll in the subjects which give grade A easily. Because grades are important to apply for a job.” Science student teachers’ views of the instructional set appeared to be influenced by the content of the examination; all of the marks in the examination were to do with content knowledge. e activities in the instructional set, class attendance and completion of the student activity books did not count as part of their grades. An interview with science student teachers after the midterm examination revealed that they thought

CM


133


the questions in the examination did not ask about the procedures they had done in the laboratory, nor learning activities. ey claimed it was not necessary to come to the class and that they could study using the notes of those science student teachers who attended the class. After this, some science student teachers became less active and did not complete the laboratory activities. Some of them did not hand in their activity books and thinking books as they had before the midterm examination.

Content Coverage Content overload has been found as an obstacle to productive teaching and learning activities (Geddis, et al., 1993). is research study found similar results. e instructor, with the responsibility to cover all of the content in her mind, was worried that she could not present all the content and do the activities. e volume of content left insufficient time for science student teachers to think in depth about their learning and obstructed their creativity and enjoyment in learning.

Laboratory Preparation e quality of samples and apparatus could be part of successful learning outcomes (Authors, under review). e poor quality of the razors provided to the science student teachers when they worked on an activity about leaf structure meant that none of them could prepare an adequate slide and so they could not learn from observing the samples. is sug-gests the need for an instructor to check the samples and apparatus before class. is could help in reducing time to simply complete the technical aspects of laboratory activities and leaving more time for thinking. Also, it could also reduce the science student teachers’ feeling of frustration and failure in doing laboratory activities which might motivate them better to learn.

SUMMARY

This paper presents the development, implementation and a brief evalu-ation of an instructional set for the Laboratory in Biology Course

designed to promote the first year ai science student teachers’ biology concepts, the nature of science concepts, science process skills and posi-tive attitudes towards biology. e instructional set was drawing on the guidelines for a learner-centered approach set out in the ailand National Education Act, 1999, and a review of research on constructivist teaching approaches as well as a survey of the regular practice, instructor’s beliefs and science student teachers’ learning outcomes after they participated in the regular course. e instructional set was validated by experts and negotiated with the instructor and implemented during November 2004

CM


134


to March 2005. Data on instructor’s and science student teachers’ perspec-tives and experiences were collected during and after the implementation. Initial findings highlight the role of instructor beliefs, science student teachers’ motivation, content coverage and laboratory preparation. e findings to date suggest that the study will have implications for the design of laboratory courses.

REFERENCESAikenhead, G., R.W. Fleming, and A.G. Ryan. 1987. High-school graduates

beliefs about science-technology-society: Methods and issues in monitoring student views. Science Education. 71: 145-161.

Areekul, S. 2005. Higher education for country development. Available: http://www.onec.go.th/publication/4009001/index.htm, March, 17, 2005.

Barker, M. 1985. Teaching and learning about photosynthesis. Working Paper. No. 220-229. Hamilton, NZ: e University of Waikato.

Bryan, L. A. 2003. Nestedness of beliefs: examining a prospective elementary teachers’ beliefs system about science teaching and learning. Journal of Research in Science Teaching. 40: 835-868.

Cronin-Jones, L. L. 1991. Science teaching beliefs and their influence on curriculum implementation: two case studies. Journal of Research in Science Teaching. 28: 235-250.

Ebenezer, J. V., and U. Zoller. 1993. Grade 10 students’ perceptions of and attitudes toward science teaching and school science. Journal of Research in Science Teaching. 30: 175-186.

Francis, L. J., and J. E. Greer. 1999. Attitude toward science among secondary school pupils in northern ireland: relationship with sex, age, and reli-gion. Research in Science & Technological Education. 17: 67-74.

Geddis, A. N., B. Onslow, C. Beynon, and J. Oesch. 1993. Transformation content knowledge: learning to teaching isotopes. Science Educa-tion. 77: 575-591.

Jesky-Smith, R. 2002. Me, teach science?. Science and Children. 39: 26-30.Lewis, J, J. Leach, and C. Wood-Robinson. 2000. Chromosomes: the

missing link-young people’s understanding of mitosis, meiosis, and fertilization. Journal of Biological Education. 34: 189-199.

Moore, R. W., and H. Foy. 1997. e Scientific attitude inventory: a revision (SAI II). Journal of Research in Science Teaching. 34: 327-336.

Office of Education Council. 1999. Higher educational reform in ac-cordance with the national education act, 1999. Available:http:// www.onec.go.th/publication/4309026/dereformudom.pdf, March, 17, 2005.

CM


135


Office of Education Council. 2001. e National Education Act, 1999. (2nd Ed.). Bangkok: Siamsport Syndicate Company Limited.

Polsaram, P., and W. ephasadin Na Ayutaya. 2000. e Development of bachelor’s degree learning process. Bangkok: Aroon Press Ltd.

Puengpang, N., V. Roadranka, B. Cowie, and C. Eames. 2006. First-year science student teachers’ biology concepts. Paper presented at the International Science Education Conference (ISEC), Singapore, November 22-24.

Salta, K., and C. Tzougraki. 2004. Attitudes towards chemistry among 11th grade students in high school in Greece. Science Education. 88: 535-547.

Simpson, W. D., and E. A. Marek. 1988. Understandings and miscon-ceptions of biology concepts held by students attending small high schools and students attending large high schools. Journal of Research in Science Teaching. 25: 361-374.

Veal, W. R. 2004. Beliefs and knowledge in chemistry teacher development. International Journal of Science Education. 26: 329-351.

CM


137


Tepkanya Promkatkeaw1, Michael Forret2 and Judy Moreland2

A Case Study of a ai In-service Primary Teacher’sUnderstanding of the Nature of Science

1Ph.D. candidate, the Program to Prepare Research and Development Personnel for Science Education, Department of Education, Kasetsart University, Bangkok, 10900, ailand. 2Centre for Science and Technology Educational Research, the University of Waikato, Private Bag 3105, Hamilton, New Zealand.Corresponding authors. E-mail: [email protected]

ABSTRACT

This is an exploratory phase of a research study for the develop-ment of an in-service primary teacher’s professional development

program. e program focuses on enhancing teachers’ understanding and teaching of the Nature of Science. e Nature of Science is referred to the values and assumptions inherent in science, scientific knowledge, and the development of scientific knowledge. e conceptions of the Nature of Science are generally represented in statements about science. ere are five major aspects of these conceptions which composed of; definition of science; charac-teristics of scientific knowledge; development of scientific knowledge; character-istics of scientists; and interrelation between science, technology, and society. Understanding of the Nature of Science on each conception can be interpreted as temporary view, contemporary view, and alternative or inappropriate conception. To investigate teacher’s understanding of the Nature of Science, a Grade 1 teacher from one public primary school was interviewed and asked to complete the Nature of Science Instruction Questionnaire. is teacher did not major in science but she had to teach science. e findings showed that she had a contemporary view of science that included science as a systematic process of knowledge inquiry and scientific knowledge as empirical, theory-laden and tentative. However, she did not like to teach science because of her lack of scientific knowledge, experiences in science teaching and understanding of scientific process and relationship between science and technology. e findings suggested that the professional develop-ment program should focus more on developing these aspects of the Nature of Science.

Key words: Understanding of the Nature of Science, In-service Primary Teacher

CM


138


INTRODUCTION

In ai science education, an understanding of the Nature of Science has been recommended as one of the learning objectives for science in

basic educational levels similar to those of many other countries (AAAS, 1994; Austin, 1997; McComas and Olson, 1998; NSTA, 1998; IPST, 2003). It has been appreciated that an understanding of the Nature of Science could assist a person become literate in contemporary science. By learning about the Nature of Science, students will be able to critique their everyday situations concerning scientific issues, make effective arguments and apply basic scientific concepts to many situations. Understanding the Nature of Science also enhances students’ learning of science content, awareness of moral and ethic values and decision-making (Driver et al., 1996; McComas et al., 1998). Teachers play a vital role in improving students’ understanding of the conception of the Nature of Science. All teachers, including primary teachers, are required to have an adequate understanding of the concep-tion and know how to help students construct such understanding. ey must possess the Nature of Science Pedagogical Content Knowledge (NOS PCK) (Abd-El-Khalick and Lederman, 2000). In the primary school science classrooms, it has been suggested that primary teachers should not avoid conveying the appropriate conceptions of the Nature of Science to their students. is is because students should be developing some conceptions of the Nature of Science in lower primary levels (Harlen, 1996; Etheredge and Rudnitsky, 2003; Skamp, 2004). From the pilot interview in this study, it was found that the Nature of Science was quite new to many ai primary teachers. is was because it had just been explicitly mentioned in the newest national standards for teaching science. Science in primary school level just had been separated out of the integrated group of subjects of Life Experiences. Because of this separation, science had clearer and more specific content strands, learning standards and intended learning outcomes. Primary teachers were required to pay more attention in teaching science by using an inquiry learning approach and other hands-on and minds-on learning activities in order to achieve the aims and objectives of science education and to help students build an understanding of the Nature of Science (IPST, 2003). In spite of the requirement of science educational reform, most ai primary teachers who teach science did not have a degree in science teaching. Most of them lacked an understanding of scientific process skills, had negative attitudes toward science and teaching science and lacked confidence in teaching science (ONEC, 2001; Soydhurum, 2001). us, ai primary teachers needed professional development to help them better

CM


139


understand and teach of science and the Nature of Science. Before reporting the methodology and results, the definition and elements of the Nature of Science and the philosophical perspectives on the Nature of Science are discussed below to provide a context for this inquiry.

Definition and Elements of the Nature of Science ere is no one absolute definition of the term: ‘Nature of Science’. e conceptions of the Nature of Science are also considered by science educators as tentative and dynamic as they have changed “throughout the development of science and systematic thinking about its nature and workings” (Lederman, 1998; Akerson et al., 2000: 298). However, it is necessary to have operational definitions for conducting the research. us, this research adopted the definition term from Lederman’s (Lederman, 1992; Lederman et al., 1998) and defined the Nature of Science as the values and assumptions inherent in science, scientific knowledge and the development of scientific knowledge which represent unique characteristics of science as describing and explaining what the science is, how it works and how it is different from other disciplines, what the scientist has done in the society along the history and how science and scientist interact with the society. For this research, various conceptions of the Nature of Science were grouped into five main aspects. ey were: definition of science; characteristics of scientific knowledge; development of scientific knowledge; characteristics of scientists; and interrelation between science, technology and society. ese conceptions were commonly represented in the form of statements concerning about science. e components of each element of the Nature of Science were identified and shown in the form of ‘the Nature of Science Conceptual Scheme’ (see Appendix A). is conceptual scheme was used as the framework for developing assessment instruments including the interview and the questionnaire on the Nature of Science. It was also used for the classification of responses.

Philosophical Perspectives on the Nature of Science is research considers that the conceptions of the Nature of Science can be distinguished into two major groups. e first major group is the conceptions which relate to philosophical perspectives. ese conceptions can be distinguished into two sub-groups called traditional view of science and contemporary view of science. Traditional view of science relies heavily on two philosophical perspectives, the realist’s view of scientific theories and explanations and the empiricist’s view of scientific knowledge and

CM


140


knowledge development. Contemporary view of science relies heavily on other two philosophical perspectives; the instrumentalist’s view of scientific theories and explanations and the constructivist’s view of scientific knowl-edge and knowledge development. Conceptions of the Nature of Science based on these two different sets of philosophical perspectives are usually represented by different or opposite statements. For example, a person holding traditional view of science may state that: “scientists discover theories because the theories are there in nature and scientists just have to find them, and scientific knowledge corresponds directly to reality” while a person who has contemporary view of science may mention differently that: “scientists invent theories because theory invention comes from the mind, and scientific knowledge is our understanding of reality, not reality as it is.” (Nagel, 1961; Munby, 1983; Tobin and Tippins, 1993; Hung, 1997; Hammerich, 1998; Haidar, 1999) e second major group is the conceptions which do not relate to philosophical perspectives. ese conceptions are held by people who view the Nature of Science different from what science really does or by those who do not understand what science really does. For example, some people are confused with the definitions of science and technology that science is equated with technology (Clough, 1997). ey can not distinguish some technical terms of scientific method and scientific knowledge used in science out of those terms used in daily life or school life, e.g., theories, laws, experiments, hypotheses, observations and inferences. is research used these ideas of distinguishing conceptions of the Nature of Science based on philosophical perspectives and non-philosophi-cal perspectives in the analysis and interpretation of participant’s responses to interview and questionnaire. In the science education context, teacher and students have been promoted to construct their understanding of the contemporary view of science rather than the traditional view. is is because conceptions based on traditional view do not suit with the requirements of contemporary science curriculum for the twenty first century (Abell and Smith, 1994; Clough, 1997; Eick, 2000; Gess-Newsome, 2002). e misunderstandings of conceptions which have no relationship with philosophical perspectives are also a concern for many science educators and researchers (McComas, 1998; Akerson et al., 2000). Many research studies, either, based on paper-pencil standardized instruments or qualitative research, found that teacher understanding of the Nature of Science is generally limited, simplistic, unclear and inadequate for teaching the Nature of Science according to the contemporary science curriculum. Both pre-service and in-service science teachers normally have

CM


141


similar traditional views regarding the Nature of Science. For example, they tended to see science as searching for the truth of the world and relied heavily on objective scientific observation, to view scientific method as one fixed set of steps and universal and to lack concern for the social aspects of science (Rubba and Harkness, 1993; Abell and Smith, 1994; Tobin and McRobbie, 1997; Haidar, 1999; Murcia and Schibeci, 1999; Tairab, 2001; Tsai, 2002). However, it was found that among the scientists, secondary science teachers and elementary teachers, the elementary teachers held the most appropriate views of the Nature of Science. ey held less tra-ditional views and a more non-traditional, contemporary view of science. e reasons suggested were that elementary teachers may have developed their alternative views of science from their views about children’s learning, their own personal practice of teaching and construction of knowledge, and participation in in-service workshops emphasizing a constructivist’s approach to science (Pomeroy, 1993).

OBJECTIVE

The objective of this research paper is to investigate participant teacher’s understanding of the Nature of Science. e findings from

a case study of one participant teacher was presented and discussed. is research paper is a part of the exploratory phase of a research study for the development of an in-service primary teacher professional development program on the Nature of Science instruction. e findings of participant teacher’s understanding of the Nature of Science would be later used in the development of an in-service professional development program and its training manual and materials for the participant teacher.

METHODOLOGY

This study used a qualitative case study research design based on an interpretive research strategy as the framework for the collection and

analysis of data. It aimed to understand Miss Dara (pseudonym)’s under-standings of the Nature of Science in her context. It also needed to gather in-depth information of her understandings which would be later used in the development of an in-service professional development program and its training manual and materials for her and her peer teachers.

Participant and School context Miss Dara is a Grade 1 classroom teacher who taught science in academic year 2004-2005 in one middle-size public primary school, under the Office of the Basic Commission, located in Nontaburi province in the

CM


142


central region of ailand. She was introduced to the first researcher by her school administrator who concerned about the need for teacher pro-fessional development of science teaching and allowed the first researcher to conduct the research with Miss Dara and other lower primary school teachers in this school. Miss Dara’s school has similar context as other general public primary schools in ailand. First, her school lacks science specialist teachers, especially in lower primary level. Second, teachers who taught science were both general teachers and classroom teachers at the same time. ey were not majoring in science and lack of experiences and training in teaching science. ird, each classroom teacher in this school has to teach almost all eight core subjects including science. Fourth, teachers also have other teachers’ loads besides the teaching loads similar to primary teachers in other schools. Last, there are approximately 35 students per classroom.

Instruments e Semi-structured interview on the Nature of Science understand-ing e semi-structured interview was used to investigate Miss Dara’s understandings of the Nature of Science conceptions. e first part of the in-terview was designed to find out her general background information related to science teaching. is included questions about; academic background, teaching experiences, teaching load and non-teaching responsibilities in school, experiences in attending any training programmes, preference and satisfaction on teaching science, supervisions for science teaching in school and school supporting in science teaching. e second part was designed to investigate Miss Dara’s understandings in all five major aspects of the Nature of Science according to the Nature of Science conceptual scheme. e first researcher applied question items from various open-ended ques-tionnaire instruments which included a questionnaire developed by Abell and Smith (1994), and View of Nature of Science questionnaire (VNOS) developed by Lederman et al. (2002). e interview question items from these sources were applied and translated into ai language. ey were also reviewed and suggested for improvement of the content validity by two ai science educator experts and three ai research supervisors. en, the semi-structured interview was modified and was tried out with three lower primary teachers who were in similar context of the study before being used in this study.

e Nature of Science Instruction Questionnaire (NOSI question-naire) is instrument was developed after the try-out of the semi-struc-tured interview. is was for diminishing one major limitation found in

CM


143


the usage of the first instrument. It was found out that the semi-struc-tured interview could not cover every specific detail in each aspect of the Nature of Science and neither did the teacher responses cover all of these aspects. e researcher suspected that it was because the questions were too broad and lacked a consideration of the teachers’ contexts. is question-naire was developed firstly in English language under the suggestions and supervisions of two experienced science educator researchers. en, it was translated into ai language and was reviewed and suggested to improve its face and content validity by two ai science educator experts and three ai research supervisors. e questionnaire was modified and was trialled by other teachers before being used in the study. e final form of NOSI questionnaire contained three parts (see Appendix B). e first and second parts composed of check lists and open-ended questions to find out Miss Dara’s general information and her understanding of the teaching of science which related to conceptions of the Nature of Science. e last part consisted of 29 rating-scale statements on the ideas of the Nature of Science to find out Miss Dara’s understanding of specific aspects of the Nature of Science. ese statements of the Nature of Science were applied from various assessment instruments (Aikenhead and Ryan, 1992; Cobern and Loving, 1998; Hammerich, 1998; Nott and Wellington, 1998; Haidar, 1999; Tairab, 2001).

Data collection In the beginning of the first semester of academic year 2004, Miss Dara was interviewed for 45 minutes with an audio recorder. e interview aimed to find out general background information and gather primary information about her understandings of the Nature of Science. Miss Dara also completed the Nature of Science Instruction Questionnaire (NOSI questionnaire) for finding out more specific data of her understanding of the Nature of Science.

Data Analysis Miss Dara’s interview responses were transcribed and categorized into two major groups: teacher’s general information and background and understandings of the Nature of Science. e understandings of the Nature of Science were also categorized into five major subgroups which included: definition of science; characteristics of scientific knowledge; development of scientific knowledge; characteristics of scientists; and interrelation between science, technology and society. Her responses to the first and second part of the NOSI questionnaire were coded and categorized into the same two major groups and five major subgroups as those of the interview responses. Her responses to the rating-scale part of the NOSI questionnaire were scored and calculated for finding out her mean score. en, these responses

CM


144


were analyzed and categorized into the range of strongly-traditional view to strongly-contemporary view of science and into the range of “do not have understanding” to “have correct conception” of the Nature of Science. Data from both interview and questionnaire responses were grouped and triangulated to ensure the validity of interpretation.

RESULTS

Miss Dara’s profile and her understandings of the Nature of Science are described under two major topics as follows:

Teacher’s Profile Miss Dara (pseudonym) is a Grade 1 classroom teacher and takes responsibility in the department of evaluating and assessing student performance of her school. She graduated from a faculty of education, majoring in teaching ai language. She has experiences more than 15 years in teaching all groups of subjects, including Life Experience group of subjects which science had been integrated within. According to the new ai science curriculum and recent educational reform policy as well as lacking of science specialist teacher in her school, in the academic year 2004, Miss Dara had to teach science as one of an individual core subject for the first time. She had never attended any training courses for science teaching. However, she has tried to improve her teaching by observing her peer teachers’ teaching science. From the questionnaire, she responded that she felt neutral with science because she did not have good background of studying science. She also did not like to teach science because of lacking of expertise in teaching and lacking of ability in “transmitting” knowledge of science (from Miss Dara’s written statement in open-ended part of the questionnaire). However, from an interview, she mentioned that she tried to make it fun when teaching science for her students would love and enjoy studying science.

Teacher’s Understandings of the Nature of Science From the statistical analysis of the whole rating scale part of NOSI questionnaire, Miss Dara’s mean sore of the whole question items was 3.696. is could be inferred that on the whole, she had contemporary view on the conceptions of the Nature of Science. Her responses to 29 rating-scale items of the Nature of Science statements are shown in the Table 1 below. ‘X’ refers to a position of her agreement on the rating scale. It is also inferred to her point of view on each statement in the range of traditional to contemporary view or in the range of misconception to right conceptions.

CM


145


Table 1: Miss Dara’s Responses to the Nature of Science Statements.

Statements

Stro

ngly

agre

e

agre

e

Neu

tral

disa

gree

Stro

ngly

disa

gree

1 Science is only a body of knowledge which explains concrete things, natural phenomena, and the world around us.

Traditional Contemporary <-------------X------------------------>

2 Science is a process of exploring the unknown and discovering new things about our world and universe and how they work.

Contemporary Traditional <-------------X------------------------>

3 Science is inventing and designing things and finding out ways to make people’s lives better.

Misconception Naïve Right <-------------X------------------------>

4 Scientific knowledge is discovered by scientists who make observations of the world and carefully record and organize the facts or truths in a systematic way.

Traditional Contemporary <--X----------------------------------->

5 Scientific knowledge is generated by scientists who attempt to explain and predict the natural phenomena based on the light of empirical evidences and logical reasoning.


6 We can trust in scientific knowledge because it is confirmed by scientific method, scientific experiments and ap-proved by scientific community.


7 e words facts, concepts, principles, laws and theories in science have the same meaning as using in our daily lives.

Misconception Naïve Right <------------------X------------------->

8 Scientific facts come from observation and experiments, and then, always become scientific theories that in turn become laws.

Misconception Naïve Right <-------------X------------------------>

9 We can not have questions about scientific theories and laws. ey are 100% true according to the reality and unchangeable. We have to remember and use them.

Traditional Contemporary <---------------------------X---------->

10 When scientists make arguments, their personal bias does not have influence.

Traditional Contemporary <-----------------------------------X-->

11 Scientific knowledge can change if there is more evidence.

Contemporary Traditional <--X----------------------------------->

CM


146


Statements

Stro

ngly

agre

e

agre

e

Neu

tral

disa

gree

Stro

ngly

disa

gree

12 We make senses of the world by careful and systematic study for making rea-sonably explanation of phenomena.


13 e world and universe occur in con-sistent patterns, we can discover these patterns by using the intellect and aid of instruments.

Traditional Contemporary <-------------------X------------------>

14 Science is the most powerful way of understanding and discovering the truth of the natural world.


15 Some matters cannot be examined use-fully in a scientific way.

Contemporary Traditional <--------Do not understand ------->

16 Science can not completely explain the world and how it works. But, it can give accurate approximations.

Contemporary Traditional <----------------------------X--------->

17 Scientists have no ideas about results before doing observations or experi-ments.


18 Scientists do experiments to confirm scientific knowledge as proving the true or false of this knowledge.


19 Every scientist always does scientific investigations in the same ways as a fixed set of steps to gain their scientific knowledge.


20 Scientists try to explain and predict phenomena as the most accurate to the reality, although they know that they can not discover the absolute truth.


21 We can not use imagination and creati-vity to get scientific knowledge.

Traditional Contemporary <-------------------X------------------>

22 Scientists need to communicate their scientific results and require peer re-view.


23 Scientists always work individually in laboratory and social affairs outside do not affect their work.


24 Scientists are people like every body and participate in community both as specialists and as citizens.


CM


147


Statements

Stro

ngly

agre

e

agre

e

Neu

tral

disa

gree

Stro

ngly

disa

gree

25 Science and technology affects society on the one hand, society affects science and technology on the other hand.


26 Technology gets ideas from science and science gets new processes and instru-ments from technology.


27 Scientists can do scientific research by do not have to consider much about ethical principles.


28 Science is only scientists’ work. Other group of people in society can not participate.


29 Today scientists develop scientific knowledge from previous knowledge according to history.


In the following sections, the findings of Miss Dara’s understand-ing from both interview and NOSI questionnaire are described in details within five topics of each major aspect of the Nature of Science.

1. Definitions of Science In the semi-structured interview, Miss Dara was asked that: Based on the Nature of Science or science subject, what is the meaning or definition of science? Her answer was:

…Science, it must make students learn by emphasizing process, right? Learning in process… Students must know about observa-tion, analysis and synthesis… [in science] has observation, analysis synthesis and application.

From the answer, she seemed to view science as the process and she also gave the answer by defining science education or gave the definition of science from teacher perspective. From Table 1, Miss Dara’s responses to items 1 and 2 of the NOSI questionnaire show that Miss Dara agreed with both traditional and con-temporary view of the definitions of science. us, it could be inferred that she had mixed view or blended view of the definition of science. She seemed to view science as both body of knowledge and process to under-stand the world. However, according to the interview, she seemed to adopt a contemporary view of science as process into her classroom teaching

CM


148


rather than a traditional view of science as a body of knowledge. e questionnaire response also found that Miss Dara agreed that science is inventing and designing things and finding out ways to make people’s lives better (item3). is could be inferred that Miss Dara had alternative conception of the different definition between science and tech-nology. She seemed to view science in the term of technology. It could also be inferred that she might have utilitarian view which thought that the aim of science was to help finding ways to make people’s lives better.

2. Scientific Knowledge 2.1 Source of Scientific Knowledge In the semi-structured interview, Miss Dara was asked that: How could scientists gain scientific knowledge? Her answer was:

…It [scientific knowledge] may firstly come from learning by oneself, from experiences. at is experiences sometimes teach us automati-cally… As children, sometimes they haven’t learnt from us yet, but they can tell some stories or something to us. is is because they have experiences, their own direct experiences. ey would learn science unconsciously. I think that scientists may be like that, learn-ing from experiences…

It could be seen that Miss Dara emphasized that scientists might gain scientific knowledge primarily from learning by oneself and from direct experiences similar to how young children learn from their own experiences. From Miss Dara’s responses to questionnaire items 4 and 5 of the Table 1, it showed that she agreed with both traditional and contemporary view of the source of scientific knowledge whether it came from discovery or generation by scientists. is could mean that Miss Dara strongly agreed that scientific knowledge was discovered by scientists through observation and scientific method, at the same time she agreed that scientific knowledge was also generated by scientists through their process of thought based on empirical evidence. 2.2 Nature of Scientific Knowledge ere are three sub-points of the nature of scientific knowledge focused in this research. 2.2.1 Validity of Scientific Knowledge Miss Dara’s response to a statement about validity of scientific knowledge on the NOSI questionnaire is shown in item 6 of the Table 1. It was found that she agreed with contemporary view which emphasized that validity of scientific knowledge was based on both confirmation

CM


149


of empirical evidence through scientific methods and experiments and approval of scientific community. 2.2.2 Definitions of Different Kinds of Scientific Knowledge In the semi-structured interview, Miss Dara was asked that: Do you have heard about scientific terms which include: facts, concepts, prin-ciples, laws and theories before? In your understanding, what are they or what do they mean? When she was asked about these, she looked hesitant to answer and tried to explain her ideas with low confidence. Her answer was:

…To give definitions … based on my understanding? A law should be, what should it be? Explaining from ai language, law in ai language is a regulation, an agreement, right? Being regulation, rule, standard, order … which we have created … Fact … it should be different … Law, sometimes we set it but we cannot do [or follow] it, but the fact? Um… it [fact] should have been analyzed as being correct. [Fact] should pass analysis before … as we told students to think of a classroom rule. en, students told us but they may not be able to do. But, what they could have told should be a law or order … eory, I do not teach older students. ere are not many theories for [young students] to think or search … I have never taught older students in this subject. So, I rarely explain about theories. It quite hard for me to explain about what theory is … I have to tell, ‘wait and let me search out first’ similar to students searching about them … I need time to search. I cannot remember everything.

is interview responses show that Miss Dara tried to explain defini-tion of scientific laws based on her perspective of ai language teacher. She thought that laws were something which had been set by people as regulation or agreement to follow. About scientific facts, she thought that scientific facts should come after analysis. In other words, scientific facts should have passed the verification. About scientific theories, it seems like Miss Dara could not give any ideas about this and she gave the reasons for her excuse that this was because she taught in lower primary level and there were not many scientific theories for students to learn. Miss Dara’s responses to statements about different kinds of scientific knowledge on the NOSI questionnaire are shown in items 7 and 8 of the Table 1. From the table, Miss Dara chose neutral choice to item7. is could mean that she had naïve understanding about the difference between using words: facts, concepts, principles, laws and theories in science and using these words in daily lives. is finding was correlated with her own interview response which she had tried to explain these scientific words from her classroom practices. Miss Dara’s agreement to item8 could be

CM


150


inferred that she had misconception that scientific facts could be simply developed to become scientific theories which then turn to be scientific laws. 2.2.3 Specific Characteristics of Scientific Knowledge In the semi-structured interview, Miss Dara was asked that: How could you distinguish which knowledge is scientific knowledge or not? Or how could you know what is scientific knowledge? When Miss Dara was asked about these, she took times to think before expressing her ideas. She gave her brief answer as follows:

… It [Scientific knowledge] should be higher-order thinking. Analy-sis, analysis thinking is probably be science… Scientific knowledge should have thinking, analysis thinking, [and] reasoning. It should have rational thinking.

us, an interview response showed that Miss Dara seemed to emphasize that scientific knowledge was different from other kinds of knowledge because it has higher order thinking, analysis thinking and rational thinking. Miss Dara’s responses to statements about specific characteristics of scientific knowledge on the NOSI questionnaire are shown in items 9 to 11 of the Table 1. It was found that Miss Dara disagreed and strongly disagreed with items 9 and 10 and strongly agreed with item11. ese could be inferred that Miss Dara had contemporary view on the aspect which recognized that scientific knowledge is tentative, questionable, em-pirical-based and subjective.

3. Development of Scientific Knowledge 3.1 Scientific World View and Basic Assumption of the Nature Miss Dara’s responses to statements about scientific world view and basic assumption of the Nature on the NOSI questionnaire are shown in items 12 to 16 of the Table 1. From the table, Miss Dara strongly agreed with item12. is could mean that she had contemporary view that the world could be understandable by human intellect and their construction of reasonable explanation. Miss Dara chose neutral choice to item13. is could be inferred that she might have mixed view between traditional and contemporary view. She might neither agree nor disagree with the tradi-tional view that the world and universe occurred in consistent patterns which we could discover these patterns. Miss Dara’s responses to items 14 and 16 clearly showed that she had traditional view which emphasized that science could provide absolute certainty or discover the truth of the natural world with lack of concerning of its inherent limitations. Miss

CM


151


Dara’s response to item 15 as “do not understand” could be inferred that she might not know about limitation of scientific inquiry before. 3.2 Scientific Method and Process In the semi-structured interview, Miss Dara was asked that: Do scientists have any processes? What are their processes? Her answer was:

It would be process about where it came from, how it came. It has steps from observation everyday to see how it start, develop and change. It would be process. [Scientists] may not know [every con-cept] all at once … Yes, [scientists] have their process of thinking.

e answer showed that Miss Dara thought that scientific process started from observation cooperated with thinking process to describe or explain the processes and development of observed phenomenon. Miss Dara’s responses to statements about scientific method and process on the NOSI questionnaire are shown in items 17 and 18 of the Table 1. It was found that Miss Dara strongly disagreed with item17. is means that she had contemporary view which emphasized that scientific method and process had theory-laden. Her response to item18 showed that she strongly agreed with the traditional view, which emphasized that scientific experiments or scientific method are the means to generate and verify truth in scientific knowledge as true or false. 3.3 Characteristics of Scientific Inquiry Miss Dara’s responses to statements about characteristics of scientific inquiry on the NOSI questionnaire are shown in items 19 to 22 of the Table 1. Miss Dara’s “strongly agree” answer to item19 could mean that she had traditional view which scientists used only a single scientific method, or other sets of orderly and logical steps, as tools for their research. Her response to item 20 showed that she had contemporary view which em-phasized that science was an attempt to explain and predict phenomena as the most accurate to the reality with concerning of its limitation to discover or reveal the absolute truth. Miss Dara chose neutral choice for item 21. is could mean that she had mixed view between traditional and contemporary view on the role of imagination and creativity in scientific inquiry. In other words, she might think that scientists either used or did not use imagination and creativity in their work. For item 22, Miss Dara chose to respond as “do not understand”. is could be inferred that she might not know that communication of scientific results attributed scientific inquiry.

CM


152


4. Interrelation between Science, Technology and Society For this major aspect, Miss Dara’s understanding was investigated as the whole and as particular points divided into: interrelation between science and technology and interrelation between science and society. For the whole aspect of interrelation between science, technology, and society, Miss Dara was asked in the semi-structured interview that: What is the importance of science? How is it important to technology and society? Her answer was:

… Science should relate mostly to our daily life because nowadays science is very important. If children do not have scientific knowl-edge, they would live in suffering. It is hard to live in society because now our lifestyle, environment, food, even cloth are relating with technology. Since we wake up until sleep, they are all relating to science. It is about how to protect ourselves, care for ourselves and what are the importance and the need of our lives.

From the interview, it showed that Miss Dara recognized the impor-tance of science in daily life. She knew that science related to all aspects of living and related to technology. She seemed to emphasize the use of scientific knowledge in protecting and caring of oneself. Miss Dara’s responses to statements about interrelation between science, technology and society on the NOSI questionnaire are shown in item 25 of the Table 1. It showed that Miss Dara strongly agreed with this item which means that she had contemporary view in recognizing the interrelation between science, technology and society. For particular points, Miss Dara’s responses to questionnaire are shown and discussed in two sub-topics as follows: 4.1 Interrelation between Science and Technology Miss Dara’s response to a statement about interrelation between science and technology on the NOSI questionnaire is shown in item 26 of the Table 1. Miss Dara chose to respond as “do not understand” to this item. is could be inferred that she might not know about the interrela-tion between science and technology. is lack of understanding seemed to relate with her misunderstanding to distinguish between science and technology as had been discussed in topic 1 (Definition of Science). 4.2 Interrelation between Science and Society Miss Dara’s responses to statements about interrelation between science and society on the NOSI questionnaire are shown in items 23, 24 and 27 to 29 of the Table 1. From the table, it could be seen clearly that Miss Dara had strong contemporary view of the interrelation between science and

CM


153


society as indicated in all of her responses with items 23, 24, 27, 28 and 29. ese findings could be inferred that Miss Dara viewed that science was not separated or in isolation from humanity as she strongly disagreed with items 24 and 28. She recognized the social and cultural context in which scientists worked and social aspect of science in which science was a complex social activity as she strongly agreed with items 23 and 29. She appreciated that scientists could not work with value and ethics free as she strongly disagreed with item 27. 5. Characteristics of Scientists and Scientific Attitudes For this major aspect of the Nature of Science, in the semi-struc-tured interview, Miss Dara was asked that: What are the characteristics of scientists? Her answer was:

Scientists? In my opinion, since I was young, I think scientists are clever people … they are omniscient. ey have knowledge much more than other people … Being observant, thoughtful, researcher, inventor …analyst …everything is all in a scientist.

From her answer, it could be seen that Miss Dara seemed to view scientists as a group of people who are smarter than other general people and appreciate the curiosity and reasoning aspects of scientific attitudes. However, she did not recommend about other important aspects of scien-tific attitudes which included: persevering, trustworthy, circumspect and broad-minded. From open-ended part of the questionnaire, question number 4 was designed to assess teacher’s understanding of the components of scientific attitudes. e question was: What are scientific attitudes or habits of mind that students should gain from learning science? Miss Dara’s written statements responded to this question were: (1) Like to learn science, (2) Pay attention to learning, and (3) Know how to acquire knowledge It was found that her answer did not relate to the question. It could be inferred that she might not understand the question or she might confuse to distinguish between attitudes toward science or science learning and scientific attitudes or habits of minds. Moreover, it seemed that she failed to recognize the importance of scientific attitudes in science teaching and learning while she already had appreciated some of their components.

CM


154



There are at least six issues on ai teacher’s understanding of the Nature of Science which emerged from the findings.

First, teacher who did not have good background on science and teaching science could have contemporary view of the Nature of Science in the whole. is finding was similar to the findings from survey research with junior high school science teachers in the northeastern region of ai-land and the findings with primary teachers in other countries (Pomeroy; 1993; Pholthum, 1997; Khantha, 1998; Suksringarm and Suksringarm, 1998). is understanding of the Nature of Science might come from self-acquiring of scientific knowledge for science teaching and receiving information from various media. Second, teacher in this case defined science and explained several conceptions of the Nature of Science in the light of subject to teach and teaching practice. is was similar to other pre-service teachers who had education view of science found in other research (Abell and Smith, 1994; Gess-Newsome, 2002). It was suggested that teacher defined and explained science in this way because they viewed science from the teaching and learning context. However, it should be considered that teacher might not be able to distinguish science education (school science) from real science (scientists’ science) (Abell and Smith, 1994). ird, teacher in this case had contemporary view of several concep-tions of the Nature of Science. e findings showed that this teacher was different from many teachers who had inappropriate view on the specific characteristics of scientific knowledge and scientific inquiry. Teacher in this case recognized that scientific knowledge was tentative and scientific methods and processes were theory-laden while many teachers lacked concerning of these aspects (Abell and Smith, 1994; Murcia and Schibeci, 1999; Akerson et al., 2000; Eick, 2000). Teacher in this case was similar to other teachers in having more adequate views of the interrelation between science, technology and society. ey agreed that science, technology and society were mutually dependent, but they could not clearly explain them (Rubba and Harkness, 1993; Tairab, 2001). Moreover, this teacher’s view on the primary purpose of science as being for the benefit of humankind seemed to correlate with many teachers’ view (Lederman, 1992; Abell and Smith, 1994). Fourth, teacher in this case had some mixed views between traditional and contemporary view of the Nature of Science in several conceptions. is was similar to the findings from many teachers in other research. is teacher and many teachers view science as both a body of knowledge and process of gaining knowledge. ey viewed that scientific knowledge

CM


155


was gained from both discovery and generation of the knowledge by scientists. ey thought that scientists could use creativity in the some stages of investigations, but could not use it for some stages of scientific knowledge development such as interpretation of data and invention of explanation (Murcia and Schibeci, 1999; Akerson et al., 2000; Gess-New-some, 2002). Fifth, teacher in this case had traditional view of the Nature of Science in particular conceptions similar to many teachers from other research findings. is teacher and many teachers thought that science could provide absolute certainty or discover the truth of the natural world with no limitations. ey agreed that scientific method and process were the means to generate and verify truth in scientific knowledge. ey viewed that scientific know ledge came totally from objective observation. ey thought that scientists used only a single fixed set of orderly and logical steps of scientific method (Pomeroy, 1993; Tobin and McRobbie, 1997; Akerson et al., 2000; Eick, 2000). Finally, teacher in this case had inadequate and inappropriate un-derstandings of several conceptions of the Nature of Science similar to the findings of other research. is teacher and many teachers defined science in the term of technology with utilitarian view. Teachers could not formulate and express the differences between facts, laws, and theories in science and could not explain their roles, functions and characteristics in science. ey had hierarchical view of facts laws and theories in science. ey lacked concerning of inherent limitations in scientific inquiry. ey could simply describe possible methods of data collection and emphasize observation, explanation and experimentation as crucial part of the scientific process. ey lacked concerning of communication role in attributing scientific inquiry. ey had difficulty in clear explanation of relationship between science, technology and society. ey could not distinguish between sci-ence and technology and lacked understanding of interrelation between science and technology. ey viewed scientists as unfamiliar extraordinary beings. ey could not distinguish between attitudes toward science or science learning and scientific attitudes or habits of minds (Rubba and Harkness, 1993; Murcia and Schibeci, 1999; Akerson et al., 2000; Eick, 2000; Tairab, 2001; Akerson and Abd-El-Khalick, 2003). It could be concluded that primary teacher who did not have good background on science and teaching science could have contemporary view of the Nature of Science in the whole. is understanding may come from self-acquiring of scientific knowledge for science teaching and receiving information from various media. However, it was found inadequate understandings and traditional views of some particular aspects

CM


156


of the Nature of Science. us, the professional development program should focus more on developing these aspects of the Nature of Science for teacher could have more adequate and appropriate understanding and more contemporary views.

ACKNOWLEDGEMENTS

This research was funded by Graduate School, Kasetsart University and the Institute for Promoting of Teaching Science and Technology

(IPST), ailand.

REFERENCESAbd-El-Khalick, F., and N.G. Lederman. 2000. Improving science teachers’

conceptions of nature of science: a critical review of the literature. International Journal of Science Education. 22: 665-701.

Abell, S.K., and D.C. Smith. 1994. What is science?: preservice elementary teachers’ conceptions of the nature of science. International Journal of Science Education. 16: 475-487.

Aikenhead, G.S., and A.G. Ryan. 1992. e development of a new in-strument: “View on Science-Technology-Society” (VOST). Science Education. 76: 447-492.

Akerson, V.L., F. Abd-El-Khalick, and N.G. Lederman. 2000. Influence of a reflective explicit activity-based approach on elementary teachers’ conceptions of nature of science. Journal of Research in Science Teaching. 37: 295-317.

Akerson, V.L., and F. Abd-El-Khalick. 2003. Teaching elements of nature of science: a yearlong case study of a fourth-grade teacher. Journal of Research in Science Teaching. 40:1025-1049.

American Association for the Advancement of Science (AAAS). 1994. Science for all americans. New York, Oxford University Press.

Austin, L.B. 1997. Teaching and learning about the nature of science. p.175-185. In B.Bell and R. Baker (eds) Developing the science curriculum in Aotearoa New Zealand. Wesley Longman New Zea-land Limited.

Clough, M.P. 1997. Strategies and activities for initiating and maintaining pressure on students’ na?ve views concerning the nature of science. Interchange. 28: 191-204.

Cobern, W., and C. Loving. 1998. e card exchange: introducing the philosophy of science. p.73-82. In W.F. McComas (ed) e nature of science in science education: rationales and strategies. Kluwer Academic Publishers, Netherlands.

Driver, R., J. Leach., R. Millar, and P. Scott. 1996. Young people’s images of science. Open University Press, Buckingham.

CM


157


Eick, C.J. 2000. Inquiry, nature of science, and evolution: the need for a more complex pedagogical content knowledge in science teaching. Electronic Journal of Science Education. 4: 1-14.

Etheredge, S., and A. Rudnitsky. 2003. Introducing students to scientific inquiry: How do we know what we know?. Pearson Education, Inc., Boston.

Gess-Newsome, J. 2002. e use and impact of explicit instruction about the nature of science and science inquiry in an elementary science methods course. Science & Education. 11: 55-67.

Haidar, A. H. 1999. Emirates pre-service and in-service teachers’ views about the nature of science. International Journal of Science Educa-tion. 21: 807-822.

Hammerich, P. 1998. Confronting students’ conceptions of the nature of science with cooperative controversy. p.127-136. In W.F. McComas (ed) e nature of science in science education: rationales and strategies. Kluwer Academic Publishers, Netherland.

Harlen, W. 1996. e teaching of science in primary schools. (2nd Edited). David Fulton Publishers, London.

Hung, E. 1997. e nature of science: Problems and perspectives. Wad-sworth Publishing Company, USA.

Institute for Promoting of Teaching Science and Technology (IPST). 2003. e manual of content of science learning. Kurasapha, Bangkok.

Khantha, S. 1998. A study of understanding the nature of science as held by junior high school science teachers under the general educa-tion department and under the office of education commission in educational region 10. esis of Master Degree in Science Education, Mahasarakham University, Mahasarakham, ailand.

Lederman, N.G. 1992. Students’ and teachers’ conceptions of the nature of science: a review of the research. Journal of Research in Science Teaching. 29: 331-359.

Lederman, N.G.1998. e state of science education: subject matter without context. Electronic Journal of Science Education. 3:1-12.

Lederman, N.G., P. Wade, and R.L. Bell. 1998. Assessing understanding of the nature of science: a historical perspective. p. 331-350. In W.F. McComas (ed) e nature of science in science education: rationales and strategies. Kluwer Academic Publishers, Netherlands.

Lederman, N.G., F. Abd-El-Khalick, R. Bell, and R.S. Schwartz. 2002. Views of nature of science questionnaire: toward valid and meaning-ful assessment of learners’ conceptions of nature of science. Journal of Research in Science Teaching. 39: 497-521.

CM


158


McComas, W.F. 1998. e principal elements of the nature of science: dispelling the myths. p. 53-72. In W.F. McComas (ed) e nature of science in science education: rationales and strategies. Kluwer Academic Publishers, Netherlands.

McComas, W.F., M. Clough, and H. Almazroa. 1998. e role and character of the nature of science in science education. p. 3-40. In W.F. McComas (ed) e nature of science in science education: rationales and strategies. Kluwer Academic Publishers, Netherlands.

McComas, W.F., and J.K. Olson. 1998. e nature of science in inter-national science education standards documents. p. 41-52. In W.F. McComas (ed) e nature of science in science education: rationales and strategies. Kluwer Academic Publishers, Netherlands.

Munby, H. 1983. irty studies involving the “scientific attitude inven-tory”: what confidence can we have in this instrument?. Journal of Research in Science Teaching. 20: 141-162.

Murcia, K., and R. Schibeci. 1999. Primary student teachers’ conceptions of the nature of science. International Journal of Science Education. 21: 1123-1140.

Nagel, E. 1961. e structure of science: Problems in the logic of scientific explanation. Harcourt, Brace & World, New York.

Nott, M., and J. Wellington. 1998. A programme for developing under-standing of the nature of science in teacher education. p. 293-314. In W.F. McComas (ed) e nature of science in science education: ra-tionales and strategies. Kluwer Academic Publishers, Netherlands.

National Science Teachers Association (NSTA). 1998. NSTA standards for science teacher preparation adopted by the NSTA Board of Directors, 1998. Available: http://www.nvgc.vt.edu/nsta-ncate/november98.htm., April 25, 2003.

Office of the National Education Commission (ONEC) 2001. Research report for the development of science education reform policy in ailand. Science Education Reform Policy Group of Staffs: Bangkok. Available: http://www.onec.go.th/publication/sci_res/index_res.htm, May 18, 2005.

Pholthum, C. 1997. A study of junior high school science teachers’ understanding of the nature of science on the educational region 11. esis of Master Degree in Science Education, Mahasarakham University, Mahasarakham, ailand.

Pomeroy, D. 1993. Implications of teachers’ beliefs about the nature of science: comparison of the beliefs of scientists, secondary science teachers, and elementary teachers. Science Education. 77: 261-278.

CM


159


Rubba, P.A., and W.L. Harkness. 1993. Examination of preservice and in-service secondary science teachers’ beliefs about science-techno-logy-society interactions. Science Education. 77: 407-431.

Skamp, K. 2004. Teaching primary science constructively. 2nd Edition. Nelson Australia Pty Ltd., Australia.

Soydhurum, P. 2001. Science education in ailand. Kul Printing Co., Ltd: Bangkok.

Suksringarm P., and J. Suksringarm. 1998. Understandings of nature of science held by junior high school science teachers in the north-eastern region. Mahasarakham University, Mahasarakham, ailand.

Tairab, H. 2001. How do pre-service and in-service science teachers view the nature of science and technology?. Research in Science & Tech-nological Education. 19: 235-250.

Tobin, K., and C.J. McRobbie. 1997. Beliefs about the nature of science and the enacted science curriculum. Science and Education. 6: 355-371.

Tobin, K., and D. Tippins. 1993. Constructivism as a referent for teach-ing and learning. In K. Tobin (ed) e practice of constructivism in science education. AAAS PRESS, USA.

Tsai, C.C. 2002. Nested epistemologies: science teachers’ beliefs of teach-ing, learning and science. International Journal of Science Educa-tion. 24: 771-783.

CM


160


AppendixAppendix A: e Nature of Science Conceptual Scheme

1. Definitions of science 01. Defining science (e.g. Science as both body of knowledge and process to explain about natural phenomena) 02. Defining major disciplines in science (e.g. pure science, applied science) 03. Interdependence of science and other disciplines (e.g. relation-ship between science and mathematic, language, art) 2. Scientific knowledge 01. Source of scientific knowledge (e.g. questioning natural phe-nomena, scientific approach to investigations) 02. Nature of scientific knowledge 1. Validity of scientific knowledge (e.g. knowledge about phe-nomena; knowledge based on observations and measurements; getting accurate data; and careful analysis, knowledge which is checked and ap-proved by other scientists) 2. Kinds of scientific knowledge (e.g. facts, concepts, principles, laws and theories) 3. Specific characteristics of scientific knowledge (e.g. tentative, empirical, theory-laden, perceptual bound/relative-absolute, inferential, imaginative and creative, objective, universal/uniformitarian, precise and uncertain, logical/rational argument, scepticism) 3. Development of scientific knowledge 01. Scientific world view and basic assumptions of nature 1. e world is understandable. 2. e universe occurs in consistent patterns that are compre-hensible through careful and systematic study. 3. People can discover patterns in all of nature by use of intel-lect and aid of instruments. 4. Science cannot provide complete answers to all questions. 5. Science has inherent limitations 02. Scientific method and process (e.g. general scientific methods; questioning, data gathering and analysis, conclusion, and communica-tion) 03. Characteristics of scientific inquiry 1. No one way to do science and many ways to do scientific investigations 2. No universal step-by-step scientific method

CM


161


3. Attempts to explain and predict phenomena (e.g. involving the process of formulating and testing hypotheses or theories, examining of evidence, and generating new questions) 4. Logical-reasoning, imagination, and curiosity are attributes of scientific inquiry 5. Communication of scientific results, attributes of scientific inquiry (e.g. new knowledge must be reported clearly and openly) 04. Scientific process skills 1. Scientists use science process skills to accomplish their work (e.g. observing, measuring, inferring, communicating, predicting, etc.)4. Characteristics of scientists and scientific attitudes (Habits of Mind) 01. Curiosity (e.g. scientists thrive on curiosity) 02. Persevering (e.g. scientists require replicability) 03. Reasoning (e.g. scientists are sceptical about all new theories) 04. Trustworthy (e.g. scientists must be intellectually honest) 05. Circumspect (e.g. scientists require accurate record keeping) 06. Broad-minded (e.g. scientists must be open to new ideas, scientists work collaboratively and require peer review)5. Interrelation between science, technology, and society 01. Interrelation between science and technology (e.g. Science and technology impact on each other, technology draws on science and con-tributes to science) 02. Interrelation between science and society 1. Science is part of intellectual, social and cultural traditions (e.g. scientific ideas are affected by their social and historical milieu, science is a human endeavour, scientists participate in public affairs both as specialists and as citizens) 2. e history of science reveals both an evolutionary and revolutionary character (e.g. Science builds on what has gone on before) 3. People from all cultures contribute to science 4. Science is a complex social activity and is conducted in various institutions 5. ere are generally accepted ethical principles in the conduct of science (e.g. scientists make ethical decisions)

CM


162


Appendix B: Sample of Question Items from the Natureof Science Instruction Questionnaire (NOSI questionnaire)

Part I General InformationPlease answer following questions according to your opinions and experiences. 1. Are you a specialist science teacher? Yes No2. What is your major in?3. How long have you taught science?4. Do you like science? Like Neutral Don’t like Please give your reasons briefly:5. Do you like teaching science? Like Neutral Don’t likePlease give your reasons briefly:

Part II Teaching of SciencePlease answer following questions according to your opinions and experiences. 1. What are your objectives when teaching science?

2. What is the most important knowledge you think students should learn in science?

3. What are the important skills you think students should learn in science?

4. What are scientific attitudes or habits of mind that students should gain from learning science?

5. What are the teaching approaches that you use in teaching science?

CM


163


Part III Ideas about Nature of Sciencee following statements are the ideas about science which students may hold. Consider and decide whether you understand them or not; (1) If you do not understand, please tick () in the first box, (2) If you understand, please tick () in the box that best indicates your views of agree/disagree or neutral.

Responses

Statements

Do

not

unde

rsta

nd

Stro

ngly

agr

ee

agre

e

Neu

tral

disa

gree

Stro

ngly

dis

agre

e

1. Science is only a body of knowledge which explains concrete things, natural phenomena, and the world around us. 2. Science is a process of exploring the unknown and discovering new things about our world and universe and how they work. 3. Science is inventing and designing things and finding out ways to make people’s lives better. 4. Scientific knowledge is discovered by scientists who make observations of the world and care-fully record and organize the facts or truths in a systematic way.5. Scientific knowledge is generated by scientists who attempt to explain and predict the natural phenomena based on the light of empirical evi-dences and logical reasoning.

CM


165


Siew Hock Ow1*, Salimah Mokhtar2 and Roziati Zainuddin3

A Review on the Teaching and Learning Resources for theDeaf Community in Malaysia

1Department of Software Engineering, Faculty of Computer Science and Information Technology, University of Malaya, 50603 Kuala Lumpur, Malaysia2Department of Information Science, Faculty of Computer Science and Information Technology, University of Malaya, 50603 Kuala Lumpur, Malaysia3Department of Artificial Intelligence, Faculty of Computer Science and Information Technology, University of Malaya, 50603 Kuala Lumpur, Malaysia*Corresponding author. E-mail:[email protected]

ABSTRACT

At present, there are more than 26,500 deaf people in Malaysia whose main mode of communication with other hearing people

is via the Malaysian Sign Language. ere is a lack of teaching and learn-ing aids to assist the deaf community in the learning process as well as for those who are interested to learn Malaysian Sign Language. is paper reviews some of the teaching and learning resources that are available for the deaf community in Malaysia. More efforts in publishing and developing teaching materials for the deaf community are required to help bridge the communication gap between them and the hearing community.

Key words: Teaching and learning resources, Deaf community, Malay-sia.

INTRODUCTION

Deafness is the result of damage to any part of the ear. e degree of hearing loss depends on the severity of that damage. e causes of

hearing loss vary from person to person and are related to the individual’s circumstances. It could be caused by many factors which include diseases (e.g.; mumps or meningitis), drugs (e.g.; aspirin in high doses or antibio-tics called aminoglycosides), frequent exposure to loud noise, serious head injury or aging (progressive deterioration of hearing in older people which is a natural part of the aging process). e only way to communicate with this hearing-impaired group is via sign language. In Malaysia, there is only one proper deaf school, the Federation School of the Deaf, located in the state of Penang (Malaysian Federation of the Deaf, 2001). Facilities for the deaf community, specifically the teaching aids, are still lacking and need to be looked into. us, a project was initiated to study the resources that

CM


166


are available to assist the deaf community in the teaching and learning of Malaysian sign language in Malaysia.

DEAF COMMUNITY IN MALAYSIA

Based on the data compiled by the Social Welfare Department Malaysia (SWDM), there were 17,692 registered hearing-impaired in Malay-

sia in 2000 (e-pek@k, 2002). e distribution of the deaf population in Malaysia, according to state and race, is shown in Table 1. It is obvious that the deaf population is on the rise. By the end of 2002, the number of hearing-impaired who registered with the SWDM had increased to 21,981 (Hari Orang Pekak disambut hari ini, 2003). In late 2004, this figure had reached 23,000 and in 2006, it is reported at 26,500 (Bridging the gap for the deaf, 2006). As registration is not mandatory, the actual number could possibly be over 26,500.

Table 1: Registered Deaf Population in Malaysia (2000).

State

Races

TotalMalay Chinese Indian Muslim Bumis

Non-Muslim Bumis

Others

Johor 1,419 679 257 - - 81 2,436Selangor 1,037 560 346 - - 15 1,958Perak 958 579 175 2 2 11 1,727WP Kuala Lumpur

745 680 250 - - 6 1,681

Sarawak 350 672 - - 337 - 1,359Sabah - 11 - - 1,341 - 1,352Penang 534 627 123 - - 4 1,288Kedah 939 154 68 - - 10 1,171Terengganu 1,122 13 2 - - - 1,137Kelantan 1,074 36 1 - - 2 1,113Negeri Sembilan

462 201 97 - 1 3 764

Malacca 423 233 63 - 1 5 725Pahang 526 111 31 - - 6 674Perlis 255 13 3 - - 2 273WP Labuan 17 13 - 1 3 - 34Total 9,861 4,582 1,416 3 1,685 145 17,692

Source: Social Welfare Department Malaysia, 2002.

CM


167


In an article published in the News Straits Time on 28 August 2006, entitled, “Bridging the gap for the deaf” (2006), the deaf community is very much dependent on interpreters to help them bridge the communi-cation and information gap. e services of interpreters are needed in job interviews, medical check-ups and even in court. e Malaysian Federa-tion of the Deaf (MFD) stated that, in general, the majority of the deaf depends on visual communication, that is, the Malaysian Sign Language (Bahasa Isyarat Malaysia in Malay), and they have low proficiency in the Malay language, which is their second language. As a result, there is less opportunity for them in socio-economic development of the country and they also have less access to information due to the communication gap between them and the hearing community (Open your heart, 2006).

EDUCATION FOR THE DEAF IN MALAYSIA

Education is very important for everyone including the deaf community. In Malaysia, education for the deaf started in 1954 when the first fully

residential school, the Federation School for the Deaf, was established in Penang. In 1963, special integrated classes for the deaf were established to accommodate more deaf children nationwide. e current education trend is now moving towards streaming and normalisation in a less restricted setting (History of Deaf Education in Malaysia, 1991). Both the residential and special day schools, together with the special classes in the mainstreamed schools, are run in parallel to cater to the needs of the deaf children. ese special schools fulfill the regional special educational needs of the rural and isolated deaf children. ese schools are also engaged in community education about deafness and in promo-ting communications among the non-handicapped. ey also function as centres for pre-school parent guidance and provide vocational awareness, training and guidance for the deaf adolescents. In 1991, there were 27 special schools and 38 special classes for the deaf (Table 2) (Current Special Education Provision for the Deaf, 1989). Besides the 27 special schools for the deaf, deaf students including those with substantial hearing losses are receiving their basic education in regular primary or secondary schools.

CM


168


Table 2: Special Schools and Special Classes for the Deaf People in Malaysia (1991).

Regular Schools All Types

Special Schools for the Deaf

Special Classes for the Deaf

Primary 6,763 25 23 Secondary 1,292 2 15 Total 8,055 27 38

Source: Schools Division, Ministry of Education, Malaysia, 1991.

MALAYSIAN SIGN LANGUAGE

The Malaysian Sign Language (MSL) is the most-widely used method of sign communication in Malaysia. It is also known as “Bahasa Isyarat

Malaysia (BIM)”. MSL uses both manual and non-manual components: hand shapes and movements, and facial expression and shoulder move-ments. It has many ‘dialects’ that vary from state to state. e Ameri-can Sign Language (ASL) has a strong influence on MSL, but both are different enough to be considered as separate languages. In addition, there are some other variations of the MSL used in Malaysia, which are described below. a. Penang Sign Language (PSL) PSL was developed in Malaysia by the deaf children themselves outside the classroom, when oralism was predominant. It is now mainly used by older people, although many younger people can understand it (Gordon Jr., 2005a). b. Selangor Sign Language (SSL) SSL is also known as Kuala Lumpur Sign Language (KLSL), and was originally based on the American Sign Language (ASL) but has diverged significantly from the latter and it can now be considered as a language on its own. As it is used mainly in the state of Selangor, rather than Kuala Lumpur, the deaf themselves call it Selangor Sign Language. Like Penang Sign Language, it is mainly used by the older people although many younger people can understand it (Gordon Jr., 2005b). c. Manually-Coded Malay It is the only form of sign language recognised by the Malaysian government as the language of communication for the deaf. It is also referred to as Kod Tangan Bahasa Malaysia (KTBM), which is the form found in the Ethnologue. KTBM is adapted from the American Sign Language, with the addition of some local signs and grammatical signs representing affixation of nouns and verbs that are used in the Malay language. It is used in the deaf schools for the purpose of teaching the Malay language (Gordon Jr., 2005c).

CM


169


REVIEW ON THE TEACHING AND LEARNING RESOURCES FOR THE DEAF COMMUNITY IN MALAYSIA

Some non-governmental organizations (NGOs) in Malaysia such as the Malaysian Federation of the Deaf (MFD), the Lions Club, research

institutions and a number of voluntary community services have put in a lot of efforts to publish books on the Malaysian Sign Language (MSL) for the use of the deaf community. As sign language involves hand motions and facial expressions, signs illustrated (drawn) in the books cannot ‘truly’ show the sequence of hand motions. Hence, a number of community proj-ects that were initiated use the electronic media to assist in the teaching and learning of MSL. ese electronic teaching and learning aids include MSL learning packages available in video compact disc (VCD), compact disc (CD) or accessible on-line from the Internet. Table 3 shows a com-parison of some of the MSL teaching and learning resources (printed and electronic) that have been published locally. Besides these resources, there are non-governmental organisations such as YMCA (Community Sign Language Courses for Adults, 2006) which conduct MSL classes for its members only, and Majudiri Y Foundation for the Deaf (Kuala Lumpur) (Bahasa Isyarat Malaysia (BIM) for Educators & those who work with the Deaf, 2006). Every participant pays a small fee (per level) to attend the Malaysian sign language classes.

Table 3: A Comparative Look at Some Teaching and Learning Resources for MSL.Title Contents Price

(RM)Publisher/ Producer

Year Published

Learn Sign Language in Ten Hours (Tan, 1998)

e book is authored by Mr. Tan Yap who was the president of the: Society of Hearing-Impaired of the Selangor and Federal Territory, Interpreters Society of the Selangor and Federal Territory, etc., in 1998. e signs were drawn by Mr. David Tay, a deaf student of Mr. Tan Yap.It is a good and simple teaching and learning aid for the deaf people and oth-ers who are interested to learn MSL.• Contains more than 240 signs that cover more than 600 words. • Covers alphabets (A-Z), numbers (1-23, fractions, 30, 40, 50, 60, 70, 80, 90, 100 and 1000).• Ten chapters: covering signs for everyday use.• Meanings are in Chinese, English and Malay.• e signs are clear and easy to follow.

5.00 Academe Art & Printing Services Snd. Bhd.

First published in 1998

CM


170


Bahasa Isyarat Malaysia (Malaysian Sign Language)(Bahasa Isyarat Malaysia, 2000)

First book initiated by the Malaysian deaf people and contains signs that are used by the deaf community in Malaysia.• Contains more than 1,000 signs and has 30 chapters.• Arrangement of graphics is clear; easy to read.• Small size and easy to carry.

10.00 MDF Firstpublishedin 2000

Lions Book of Signs (Tan, 2001)

e book is authored by Mr. Tan Yap who was the president of the: Society of Hearing-Impaired of the Selangor and Federal Territory, Interpreters Society of the Selangor and Federal Territory and others in 2001. e signs were drawn by Mr. Heng Yih Yeong, a student of Mr. Tan Yap’s sign language class.is book was initiated and sponsored by the Lions Club Kuala Lumpur Cen-tral. About 1,500 copies were printed and distributed for free to Lions Club members for teaching of sign language as a community service to assist the deaf people in Malaysia. It is a good and simple teaching and learning aid for the deaf people and others who are interested to learn MSL.• Contains more than 250 signs that cover more than 680 words. • Covers alphabets (A-Z), numbers (1-23, fractions, 30, 40, 50, 60, 70, 80, 90, 100 and 1000).• Twelve sections: covering signs for everyday use.• e signs are clear and easy to follow.• Meanings are in Chinese, English and Malay.

Free Lions Club, Kuala Lum-pur Central (District 308 B1 Malaysia)

Firstpublishedin 2001

Communi-cation In Sign(Tan, 2003)

e author is Mr. Tan Yap who was the president of the: Society of Hearing-Impaired of the Selangor and Federal Territory, Interpreters Society of the Selangor and Federal Territory and others in 2003. e signs were drawn by Mr. Heng Yih Yeong, a student of Mr. Tan Yap’s sign language class.is book was sponsored by the Lions Club of Serdang as a community project. About 1,500 copies were printed and distributed free to Lions Club members to teach them sign language as com-munity services to assist the deaf people in Malaysia. It is a good and simple teaching and learning aid for the deaf people and others who are interested to learn MSL.• Contains more than 430 signs that cover more than 800 words. • Covers alphabets (A-Z), numbers (1-23, fractions, 25, 30, 40, 50, 60, 70, 80, 90, 100 and 1000).

Free Lions Club of Serdang(District 308 B2 Malaysia)


CM


171


• Eighteen sections: covering signs for everyday use.• e signs are clear and easy to follow.• Meanings are in Chinese, English and Malay.

Malaysian Sign Language Book 2(Buku BIM Jilid 2, 2003)

Suitable as a complete reference for the deaf people and others who are interested to learn MSL.• Contains more than 1,000 signs. • 42 chapters: communication, educa-tion, economic, etc.• Simple and easy to understand sign.• Signs for words in Malay language and English.• Sample sentence for each sign.• Attractive and systematic sign illustra-tion. • Small size and easy to carry.

20.00 MDF Firstpublishedin 2003

ICT Signs(Buku BIM: Isyarat-isyarat ICT, 2003)

Suitable for use as teaching aids for deaf students, deaf adults, teachers and parents to understand basic ICT and to learn MSL.• Contains more than 200 ICT signs. • Signs for ICT terminologies in Malay and English.• Attractive and systematic sign illustra-tion.• Provides ICT illustrations and further explanations.• Small size and easy to carry.

8.00 E-Pekak Unit, MFD


CD-ROM Panduan Bahasa Isyarat Malaysia(Malaysian Sign Language Guide)(CD-ROM Panduan BIM, 2001)

Developed by Asia Pacific Institute of Information Technology (APIIT). • Contains more than 350 words. • Includes alphabets A-Z, integers 0-9 and selected integers up to 99.• Preset fingerspelling and phrases.• Exercises are provided to enhance vocabulary and fingerspelling.• Instruction in Malay language with captions in English (selected words).• Signs are represented using video recording of human model and narration is in text mode.

5.00 APIIT 2001

Malaysian Sign Language Guide VCD (Video CD (VCD) Panduan BIM, 2003)

A 60 minute CD. System requirements include television or computer and a VCD player.• Contains 24 chapters. • Signs are easy to learn and understand.• Humorous presentation of the deaf actor.

20.00 2003

e-pek@k(e-pek@k, 2006)

e-pek@k is a national project to promote ICT awareness among the deafs.is website provides:• General information about the deaf.• Bahasa Isyarat Malaysia (Malaysian Sign Language) which covers Introduc-tion, Alphabets (A-Z), Numbers (2-20, 25, 30, 40, 50, 100, 1000, 10,000, 1,000,000) and BIM dictionary.

Free MDF Since 2001

CM


172


In the process of learning a new language, it is also important that a learner understands and masters the vocabulary well before he is able to use the language correctly. A dictionary is thus needed in the learning process. Similarly, in the learning of sign language, a dictionary is needed to understand the meaning(s) of the word (sign) and to know its pronun-ciation if the learner is not a deaf person (up to 20 dB) or who suffers from mild hearing loss (20-45 dB). A survey of sign language dictionaries shows that there are many foreign sign language dictionaries but there is yet to be available a comprehensive, printed electronic dictionary of the Malaysian Sign Language. e following section presents a review of some examples of foreign electronic sign language dictionaries.

REVIEW OF SOME ELECTRONIC SIGN LANGUAGEDICTIONARIES

Table 4 shows the features of four foreign sign language dictionaries. Some of these dictionaries can be accessed freely on the Internet.

Table 4: Features of Some Sign Language Dictionaries.Title Contents Price

(RM)Publisher/ Producer

Year Published

ASLPRO.COM Dictionaries(2006)

It is a Web-based American Sign Language video dictionary. e signs are presented by a model in a video recording.Features:• Covers over 7000 ASL signs.• Have different models to illustrate the signs.• Consists of 4 types of dictionaries: Main Dictionary (over 4000 signs), Religious Dictionary (over 1200 interdenominational signs), Conversational Dictionary (over 2000 commonly- used phrases) and Babies Dictionary.• Provides controller for the video (play, pause, stop, etc.).• English version only.• Signs cannot be viewed from different angles (no rotation function).• Meaning of the words not provided.• Although it has speed control, the signing speed is rather fast, thus not easy to follow.• No searching by category. • No sound.

Free ASLPRO.COM

Not Available

American Sign Language Dictionary(2006)

It was developed by Martin L.A. Sternberg. It com-bines text, video examples, and animated illustrations to create a rich reference tool for learning ASL. It consists of 2 CD-ROMs and can run on Windows 2000/Me/98/NT/XP.Features:• Has over 2600 signs with video demonstrations and illustrations.• Has fingerspelling games and quizzes.

US$11.95

Multi-media 2000 Inc. (M2K)

Version2.1 wasreleasedin 1996

CM


173


• Has video demonstrations and illustrations• Listen to audio explanations or read the descrip-tions on-screen• Progress through signs by subject.• Searching capabilities in 5 languages.

British Sign Language Dictionary(Welcome to the BSL Dictionary Online, 2006)

It is an online Web application of the British Sign Language Dictionary (BSL). It was developed by Stories in the Air assisted by Dunedin Multi-media, Donaldson’s College, e National Deaf Children’s Society and the City of Edinburgh Council Education Department. Features:• Contains 20 categories of sign: food, clothes, days, feelings, weather, medical terms, animals, colours, etc.• Include facial expression for each sign.• Provides speed control on the animation of the sign at any time.• Cannot search for a sign by entering the word.• Signs cannot be viewed from different angles (no rotation function).• Meaning of the words not provided.• Must install the QuickTime software to watch the sign.• No sound.

Free Stories inthe Air

2006

Ameri-can Sign Language Dictionary-Flexi(Sternberg et al., 2006)

e authors are Martin L. Sternberg, Marin Sternberg, Herbert Rogoff (Illustrator). It is an illustrated abridgment of the most authoritative reference book on sign language, with well-written and easily-understood instructions for the use of each signFeatures:• Has more than 5,000 signs and 8,000 illustra-tions.• ird edition includes more than 500 new signs and 1,500 new illustrations.• Provides a mixture of both English and ASL words

US$19.96

Harper-CollinsPublishers



Presently, there is no dictionary on MSL available for teaching and learning MSL. One such dictionary could be developed based on the

system functions and features of the foreign sign language dictionaries reviewed above. As Malaysia is a multi racial country, the MSL dictionary shall be developed to incorporate elements of the cultural and social norms of the Malaysian society. Basically, the MSL dictionary to be developed should have the following features: a. a bilingual system (provides instructions in Malay and Eng-lish). b. stores 600 signs (words) that are commonly used by the deaf com-munity in Malaysia, and index the signs in alphabetical order and group them according to specific categories such as fruits, occupation, vehicle, etc.

CM


174


c. explains one or more basic meaning(s) of each sign and provides links to other related signs (words) such as synonyms and antonyms. d. the meanings of the signs shall be illustrated in text and in graphics (if suitable pictures are available). e. provides functions which include “Administration of Database” and “Search” function. e “Administration” function is to allow an administrator to maintain the database of the sign language dictionary. f. “Search” function will be provided to allow users to search for any commonly-used signs (words) stored in the dictionary. e target users of this system include people from all walks of life, ranging from the teachers of deaf school and students, medical officers, police officers, court translators to the home-users including families with deaf children and Hence, any user who wishes to learn Malaysian Sign Language. e development of this bilingual Malaysian sign language dictionary will be timely and it should certainly be a very useful tool in the teaching-learning process in the deaf schools. It will certainly help bridge the communication gap between the hearing-impaired people and the general public in Malaysia.

ACKNOWLEDGEMENTS

The authors would like to acknowledge the financial support from the University of Malaya through the Fundamental grant (Project No.:

FP029/2005D).

REFERENCESAmerican Sign Language Dictionary. 2006. http://vesoftware.info/

american-sign-language/. Multimedia 2000 Inc. Date Accessed: 4 August 2006.

ASLPRO.COM Dictionaries. 2006. ASLPRO.COM. http://www.aslpro. com/cgi-bin/aslpro/ aslpro. Date Accessed: 27 July 2006.

Bahasa Isyarat Malaysia. 2000. http://www.epekak.net.my/bim/bim_buku_1.htm. Malaysian Federation of the Deaf. Date Accessed: 11 August 2006.

Bahasa Isyarat Malaysia (BIM) for Educators & those who work with the Deaf. 2006. http://www.mydeafoundation.org/BIMteachersbrochure2.pdf. Majudiri Y Foundation for the Deaf (Kuala Lumpur). Date Accessed: 18 July 2006.

Bridging the gap for the deaf, 2006. http://www.nst.com.my/Current_News/nst/Tuesday/ National/20060822083417/Article/local1_html.

CM


175


Date Accessed: 30 August 2006.Buku BIM: Isyarat-isyarat ICT. 2003. http://www.epekak.net.my/bim/bim_

buku_ ict.htm. Malaysian Federation of the Deaf. Date Accessed: 22 August 2006.

Buku BIM Jilid 2 (Malaysian Sign Language Book 2). 2003. http:// www.epekak.net.my/bim/bim_buku_ 2.htm. Malaysian Federation of the Deaf. Date Accessed: 11 August 2006.

CD-ROM Panduan BIM. 2001. http://www.epekak.net.my/bim/bim_cdrom.htm. Asia Pacific Institute of Information Technology. Date Accessed: 28 September 2006.

Community Sign Language Courses for Adults. 2006. http://www.ymcakl. com/pmy/SL_ courses.htm. Date Accessed: 20 July 2006.

Current Special Education Provision for the Deaf. 1989. http://www.mfd. org.my/Information/ Deaf%20education.htm#current. Date Accessed: 2 August 2006.

e-pek@k. 2002. Registered Deaf Population in Malaysia by End of 2000. http://www.epekak.net.my/info/populasi.htm. Date Accessed: 20 July 2005.

e-pek@k. 2006. http://www.epekak.net.my/. Date Accessed: 12 June 2006.Gordon Jr., R. G. (ed.). 2005a. Ethnologue: Languages of the World,

Penang Sign Language. http://www.ethnologue.com/show_language.asp?code=psg. SIL International. Date Accessed: 30 August 2006.

Gordon Jr., R. G. (ed.). 2005b. Ethnologue: Languages of the World, Selangor Sign Language. http://www.ethnologue.com/show_language.asp?code=kgi. SIL International. Date Accessed: 30 August 2006.

Gordon Jr., R. G. (ed.). 2005c. Ethnologue: Languages of the World, Malaysian Sign Language. http://www.ethnologue.com/show_language.asp?code=xml. SIL International. Date Accessed: 30 August 2006.

Hari Orang Pekak disambut hari ini. 2003. http://www.epekak.net.my/akhbar/2003/10_ hopa_4.htm. Published on BERNAMA, 4 October, 2003. Date Accessed: 15 August 2006.

History of Deaf Education in Malaysia. 1991. http://www.mfd.org.my/Information/ Deaf%20education.htm#introduction. Date Accessed: 2 August 2006.

Malaysian Federation of the Deaf. 2001. E-Pekak. http://www.mfd.org.my. Date Accessed: 11 October 2001.

Open your heart, 2006. http://www.mfd.org.my/MFD/donate.htm. Date Accessed: 27 July 2006.

CM


176


Sternberg, M. L., M. Sternberg, and H. Rogoff. (Illustrator). American Sign Language Dictionary-Flexi. 2006. http://search.barnesandnoble. com/booksearch/isbninquiry.asp? ean=9780062736345&z=y#CRV. Barnesandnoble.com llc. Date Accessed: 18 August 2006.

Tan, Y. 1998. Belajar Bahasa Isyarat Dalam Sepuluh Jam. Kuala Lumpur, Persatuan Orang-Orang Pekak Wilayah Persekutuan Dan Selangor Darul Ehsan.

Tan, Y. 2001. Lions Book of Signs. Kuala Lumpur, Lions Club Kuala Lumpur Central (District 308 B1).

Tan, Y. 2003. Communication In Sign. Serdang, Lions Club of Serdang (District 308 B2 Malaysia).

Video CD (VCD) Panduan BIM. 2001. http://www.epekak.net.my/bim/bim_vcd.htm. Date Accessed: 11 August 2006.

Welcome to the BSL Dictionary Online. 2006. http://www.learnbsl.org/. Stories in the Air. Date Accessed: 26 July 2006.

CM


177


Wacharin Chaitha1*, Chupen Viboonsunti2, Tanattha Kittisopee3, Unchalee Permsuwan2, Ratanaporn Awipan2 and Penkarn Karnjanarat2

Relationships of Service Quality, Patient Satisfactionand Patient Behavioral Intention of Pharmacy Service

of Highly-Active Antiretroviral erapy in Community Hospitals in Chiang Rai Province, ailand

1Chiang Saen Hospital, Ministry of Public Health, Chiang Saen, Chiang Rai 57150, ailand2Department of Pharmaceutical Care, Faculty of Pharmacy, Chiang Mai University, Chiang Mai 50200, ailand3Department of Pharmacy Administration, Faculty of Pharmaceutical Science, Chulalong-korn University, Bangkok 10330, ailand*Corresponding author. E-mail:[email protected]

ABSTRACT

The study examined the construct validity of the pharmacist model of pharmacy service quality of highly-active antiretroviral therapy

(PSQ-HAART) program. e proposed model was tested by identifying the relationships among pharmacy service quality, patient satisfaction and behavioral intention using path analysis. It simultaneously considered the direct effects of service quality and patient satisfaction on patient behavioral intention. It also considered the indirect effect of service quality on patient behavioral intention. Data collection was conducted during April 2005 among a random sample of 613 patients from ‘Access to Care’ clinics at 16 community hospitals in Chiang Rai province. A total of 607 questionnaires were returned (99.0 % response rate). Of these, 594 questionnaires were used for data analysis. e results suggest that service quality and patient satisfaction were directly correlated patient behavioral intention considered all the variables collectively (β = 0.218 and 0.617 respectively, p < 0.001). Service quality was also directly correlated pa-tient satisfaction (β = 0.431, p < 0.001). e impact of service quality on patient behavioral intention was enhanced by the indirect correlation (β = 0.266, p < 0.001). Based on these results, the PSQ-HAART model was confirmed. is model may be of benefit to the pharmacy service of highly-active antiretroviral therapy (HAART) programs. It implied that patients’ perception of service quality must be developed before satisfying the patient. If patients were satisfied with pharmacy service, they would have positive behavioral intention towards the service.

CM


178


INTRODUCTION

Chiang Rai province was one of the top five prevalent numbers of HIV/AIDS patients, with 19,144 cases during September 1984 to

October 2005 (Bureau of Epidemiology Department of Disease Control, 2005). All government community hospitals in Chiang Rai province they established separate Access to Care (ATC) clinics to provide highly-active antiretroviral therapy (HAART) (Horizon Program, 2004). Pharmacy service in these hospitals support the various components of the medica-tion-use process (ordering, dispensing, administering, monitoring, and education) with people living with HIV/AIDS (PLWHA), as part of the HAART program. In March 2005, there were 2,285 patients enrolled in the program in Chiang Rai province (data from the AIDS section, Chiang Rai Provincial Health Office; March, 2005). With rapid scaling up of access to ART in Chiang Rai province, patient adherence to their antiretroviral regimen is an extremely impor-tant behavior to ensure patient outcome and the success of the HAART program. However, studies on actual adherence of HAART patients are difficult to carry out and expensive. As patient behavioral intention is a predisposition to the patient’s future behavior (John, 1991), service quality and patient satisfaction are typically studies linked to behavioral intention. Better quality is related to the higher level of patient satisfaction and posi-tive behavioral intention to use the service. Patients with higher satisfaction levels often make important behavioral changes, including maintaining more stable relationships with providers (Crosby et al., 1990), changing or returning to the same hospital service provider (Woodside et al., 1989), adhering to medical regimen and treatment (Bartlett et al., 1984; Ley, 1988; Scott-Lennox et al., 1998; Peck, 2002), keeping appointments and recommending companies or service providers to others (Woodside et al., 1989). However, a review of the literature indicates that there is a general confusion over the causal ordering of service quality and satisfaction (Bitner and Hubbert, 1994; Rust and Oliver, 1994; Dabholkar, 1995) and as to which construct is a better predictor of behavioral intention (Bolton and Drew, 1991; Cronin and Taylor, 1992; Bitner and Hubbert, 1994). From the service providers’ perspective, clarity of these fundamental issues is necessary before practitioners can determine which construct best captures their customers’ evaluations of the service(s) provided and whether it is both constructs, or one opposed to the other. ey are needed to provide relevant evaluations that can be used for decision-making to improve pa-tient health behaviors.

CM


179


Causality of Global Constructs A review of the service marketing literature reveals several aspects which support the conceptual framework this study which is composed of three constructs. e first perspective suggests that satisfaction is antecedent to per-ceived service quality. e basis of this perspective is that service quality (similar to an attitude) (Parasuraman et al., 1988), is an overall evaluation and accumulation of satisfaction judgments, resulting in a global perception of service quality. Although this perspective has been supported theoreti-cally within the literature (Bolton and Drew, 1991), there is little empirical research to support this conceptual (Anderson and Sullivan, 1993). e second perspective suggests that service quality is antecedent to satisfaction. e basis of this perspective is that service quality, as a cogni-tive evaluation (according to its attitudinal conceptualization) precedes the more emotive satisfaction evaluation. Simply stated, positive service quality perceptions will lead to a more satisfied consumer. As such, satisfaction is considered superordinate to service quality. A number of studies offers empirical evidence and theoretical support for this conceptual (Cronin and Taylor, 1992; Anderson and Sullivan, 1993; Gotlieb et al., 1994). e foundation of this relationship is the appraisal, emotional response, coping framework suggested by Lazarus (1991) and Bagozzi (1992). is conceptualization posits service quality as antecedent to satisfaction based on its cognitive evaluation and satisfaction, due to its emotive nature (Dabholkar, 1995), as directly influencing behavioral intention (Brady and Robertson, 2001). Empirical research found that satisfaction has a direct effect on behavioral intention (Gotlieb et al., 1994). Service quality has an indirect effect on behavioral intention. Service quality is antecedent to satisfaction (Cronin and Taylor, 1992; Rust and Oliver, 1994). e final perspective suggests that neither satisfaction nor perceived service quality is considered superordinate (Cronin and Taylor, 1994; Dab-holkar, 1995). Given these three perspectives, it is not surprising that Rust and Oliver (1994) consider the causal ordering of the relationship between satisfaction and service quality as “perhaps the most intriguing issue” facing the discipline (Cronin et al., 2000). ere is ample evidence of relationships between both service quality and satisfaction with behavioral intention. Although it is clear that service quality is an important determinant of behavioral intention, the exact nature of this relationship remains unresolved. It is evident that few studies have investigated multiple direct links between service quality, satisfaction, and behavioral intention. Further, there is few reported investigation of whether

CM


180


either or both service quality and satisfaction directly influence behavioral intention when the effects of both two are simultaneously considered. Another belief is that partial examination of the simple links between any of the constructs and behavioral intention may mask or overstate their true relationship due to omitted variable bias.

Conceptual Framework of the Study In order to understand the underlying relationships that exist among these variables, the PSQ-HAART model was tested (Figure 1). e model presented within this study attempts to aggregate the divergent views expressed in the literature addressing a multi-dimensional conceptualiza-tion of service quality. Moreover, the application of this model within the context of pharmacy service of HAART programs represents a first attempt at integrating the diverse models from the service marketing literature with models of quality assurance and health status from health care literature. is study tested whether service quality and patient satisfaction directly lead to a favorable behavioral intention simultaneously.

Figure 1: Proposed Path Model of Perceived PSQ-HAART.

e researchers hypothesized the PSQ-HAART model based on the second perspective, which was mentioned previously. at is, that service quality has a direct effect on satisfaction (Cronin and Taylor, 1992; Gotlieb et al., 1994; Dabholkar, 1995; Lazarus, 1991; Bagozzi, 1992;). Addition-ally, satisfaction has a direct effect on behavioral intention (Goltieb et al., 1994, Brady and Robertson, 2001) and service quality has direct effect on behavioral intention (Cronin et al., 2000). Furthermore, service quality also has an indirect effect to behavioral intention through satisfaction (Cro-nin and Taylor, 1992; Rust and Oliver, 1994). is extension is offered: (1) to further our understanding of how service quality and satisfaction influence behavioral intentions, (2) to add to the growing body of litera-ture that specifies the interrelationships between these variables and (3) because these effects have yet to be considered. Specifically, the question is whether service quality perceptions have a significant indirect influence on behavioral intention through patient satisfaction.

Satisfaction

Service quality Behavioral intention

CM


181


Objectives of the Study To identify the relationships between service quality, patient satisfac-tion and behavioral intention in relation to a provision of pharmacy service of HAART program in community hospitals in Chiang Rai province.

Definitions of the Study Constructs Service quality Service quality in this study was defined based on the Parasuraman, et al. (1985) as a set of key discrepancies or existing gaps regarding execu-tive perceptions of service quality and tasks associated with service delivery to patients.

Patient satisfaction Patient satisfaction in this study was defined based on Hunt, et al. (1977) as an evaluation of an emotion of patients on how much they satisfy with the pharmacy service in the HAART program.

Patient behavioral intention Patient adherence to antiretroviral regimens is an extremely impor-tant behavior to ensure success of a pharmacy service of HAART program. However, studies on actual behavior of HAART patients are difficult to carry out and expensive. As patient behavioral intention was identified the as a predisposition to the patient’s future behavior in a study by John, et al.(1991) the researcher chose to study patient behavioral intention as an indicator of patients’ actual behavior. Patient behavioral intention in this study included continuing antiretroviral drugs adherence, complying with pharmacist’s advices, keeping pharmacist’s appointments and com-municating positive things about the pharmacy service to others.

Pharmacy service Pharmacy service is a service provided by pharmacists for the HAART program.

MATERIALS AND METHODS

Study Design: is study applied a descriptive cross-sectional design to identify the relationships among the study variables.

Population: e population in this study included all patients who enrolled in the HAART program at 16 community hospitals in Chiang Rai province. Selection of the subjects Forty subjects were randomly selected from the HAART program of each community hospital. e ‘Access to Care’ (ATC) number list of

CM


182


Chiang Rai province was used as the sampling frame. e HAART patients who had participated in the development of the questionnaire (Chaitha et al., 2005) were excluded from this study. Sample size calculation: A minimum sample size calculated by Daniel’s (1995) formula was 506 people. By allowing 20% non-respond-ing, incomplete questionnaires, or invalid answer of questionnaire items a total number of 608 people was required. A decision was made that problematic items were: (1) those which had ten percent or more of all response items either missing or had multiple answers, or (2) those which the values of all items response of each questionnaire are the same number as ‘4’ or ‘5’. In this study, 640 questionnaires were distributed to patients in 16 community hospitals in Chiang Rai province.

Study Measures is study simultaneously measured the direct effects of service quality and patient satisfaction on patient behavioral intention. It also measured the direct effects of patient service quality on patient behavioral intention. In addition, it measured the indirect effect of service quality on patient behavioral intention through patient satisfaction.

Instrument for Data Collection e study instrument was a questionnaire contained four parts with a total of 63 questions (Appendix). It took approximately 20 minutes to complete. Part I: Service quality A 33-item in six dimensions of service quality scale (Chaitha et al., 2005) was developed using the Gap Analysis Model (Parasuraman et al., 1985) as a theoretical base. e dimensions are: tangibles-reliability (8 items), assurance-empathy (8 items), empathy (6 items), responsiveness I (4 items), responsiveness-assurance (5 items), and responsiveness II (2 items). Respondents were asked to rate their desired service level (E), and the perception of actual service level (P) for each item regarding the pharmacy service using a 5-point Likert scale. Perceived service quality (Q) was then calculated. A score of Q = P - E indicated the gap discrepancy between desired service expectation and perception of actual service. e construct reliabilities of expectation and perception were 0.95 and 0.94, respectively (Chaitha et al., 2005). Part II: Patient satisfaction is part contained a 15-item scale. Re-spondents were asked to rate their perception using a 5-point Likert scale (1 = lowest to 5 = highest). e mean score of the 15 items was used as an index of patient satisfaction. All items produced a reliability of 0.90 (Chaitha et al., 2005).

CM


183


Part III: Patient behavioral intention is part contained four items. e items adapted from several studies included: continuing antiretroviral drugs adherence (Bartlett et al., 1984; Peck, 2002; Ley, 1988); comply-ing closely with pharmacist’s advices; keeping pharmacist’s appointments, and communicating positive things about the pharmacy service to others (Woodside et al., 1989). Respondents were asked to rate their perception on behavioral intention using a 5-point Likert scale (1 = strongly disagree to 5 = strongly agree). e mean score of the four items were used as an index of patient behavior intention. All items scale produced a reliability of 0.66 (Chaitha et al., 2005). Part IV: Patient characteristics is part included 11 questions about their gender, age, career, education, status, income, length of time they have been receiving pharmacy service, whether or not prescription insurance was covered, health status, numbers of prescription medication taken on a daily basis (excluding antiretroviral regimens), their antiretroviral formula, and the duration enrolled in the HAART program. Duration of the Study is study was conducted during May 2003 to June 2005. Data Collection Data collection was conducted in April, 2005. A holistic care center nurse from each hospital acted as research assistant and two PLWHA volunteers from each hospital were the fieldworkers, and underwent intensive training to administer the survey to the patients. Participants had to sign an informed consent form to be included in the study. e method of responding to the survey depended on the ability of each respondent to answer. In the case that the respondent was a child who could not write or answer the questionnaire by themselves, a parent or a guardian would respond on their behalf. For illiterate respondents and those with poor eyesight, or any other limitations that prevented them from completing the survey alone, the fieldworkers read out the questions and wrote down their answers. Each manner of survey administration was noted. e fieldworkers ensured that patients filled out the survey in private, which protected patient confidentiality. e protocol of this study was approved by the Ethical Review Committee of the Faculty of Pharmacy, Chiang Mai University.

Data Analysis All statistical analyses were carried out using SPSS for Windows Version 11.0 (Nurosis, 1990a; 1990b). Descriptive statistics were used to describe the respondents’ characteristics. To test the path correlation between the variables in the PSQ-HAART model, including service quality, patient satisfaction and behavioral intention (Figure 1), simple regres-

CM


184


sion was analyzed to predict patient satisfaction from their perception of service quality. en, multiple regression was analyzed to predict patient behavioral intention from patient perception of service quality at statistical significant level of 0.05.

RESULTSDescription of the Study Samples

The 640 questionnaires were distributed to patients receiving phar-macy service at 16 community hospitals in Chiang Rai province (40

patients/hospital). Because one hospital had only 35 HAART patients and the other had only 18 HAART patients 613 questionnaires were actually distributed. A total 607 questionnaires were returned (99.0 % response rate). Of the returned questionnaire, twelve were not used for data analy-sis because the responses were all ‘4’ or ‘5’ for every questionnaire item. e average age of the respondents from the study sample was 34.9 ± 8.3 [range 1 - 66] years. eir average income per month was 1,093.4 ± 1351.0 [0-15,000] Baht. e average duration receiving pharmacy service was 10.4 ± 16.4 [1-101] months. e respondents reported taking an average of 1.97 ± 1.10 [0-8] prescription medications per day (exclud-ing antiretroviral regimen). e majority of the respondents was female (59.9%), completed high school (58.0%) and worked as laborers (36%). Most of them were married (47.4 %), had a universal health care coverage insurance card (89.2%), reported their health status as good (47.4%), and received the GPO vir antiretroviral formula (a generic fixed dose combina-tion of stavudine, lamivudine and nevirapine) (82.7%). A total of 77.8% of the questionnaires were self-administered. For the remaining 22.2%, fieldworkers read out the questionnaire to them (13.0 % were illiterate and 9.2% were hindered by poor eyesight) (Table 1). Table 1: Characteristics of Respondents of the Study (n = 595).

Characteristics N Mean SD

Age (year) 590 34.9 8.31Income per month 587 1093.4 1351.02Time receiving the pharmacist service (month)

581 10.4 16.41

Number of prescription medication per day (excluding antiretroviral drug)

589 1.97 1.10

Frequency PercentGender Female Male

594358236

60.339.7

CM


185


Education High school Primary school Never study College Bachelor degree

589345161393311

58.627.36.65.61.9

Occupation Laborer Private business worker No career Agriculture Laborer Trade Government officer

56321414011270104234

38.024.919.912.434.84.10.7

Marital Married Widow Single Divorce

5932821858343

47.631.014.07.2

Health insurance Universal health care coverage insurance Pay from own pocket Charity fund hospital Social security insurance Government insurance

5875311817174

90.43.12.92.90.7

Health status Good Fair Very good Poor Very poor

591282162134112

47.727.422.71.90.3

Antiretroviral formula GPO vir* ree single drugs regimen

587486101

78.417.2

Level of Service Quality, Patient Satisfactionand Behavioral Intention

Service quality scores ranged between -3.12 to 1.79, with a mean of -0.026 ± 0.649 and the score skewed to the left (skewness = -1.120). is meant that the sample had high perception of service quality in pharmacy service. Patient satisfaction scores ranged between 1 to 5, with a mean of 4.115 ± 0.650 and the score distributed to the left (skewness = -0.901). is meant that the sample had high satisfaction with pharmacy service. Patient behavioral intention score ranged between 1 to 5, mean 4.623 ±


CM


186


0.517 and the score distributed to the left (skewness = -1.858). is meant that the sample had high behavioral intention to pharmacy service.

Correlation of the Study Variables As shown in Table 2, the study variables had low to moderate cor-relation, while patient satisfaction and behavioral intention yielded the highest significant correlation (r = 0.523, p < 0.001). Service quality and patient satisfaction had significant correlation (r = 0.430, p < 0.001). However, service quality had marginally significant correlation with patient behavioral intention (r = 0.047, p < 0.001).

Table 2: Correlation Coefficient Between Variables in the Study Model (n = 595).

Variable Service quality Patient satisfaction

Patient behavioral intention

Service quality 1.00Patient satisfaction 0.430* 1.00Patient behavioral intention 0.047* 0.523* 1.00

*p < 0.001

Relationship among Service Quality,Patient Satisfaction and Behavioral Intention

As shown in Table 3, service quality had a direct effect on patient satisfaction as service quality could predict patient satisfaction by 18.5 %. β = 0.431, t = 11.635, p < 0.001. Table 4 showed service quality predicted patient behavioral intention by 68.8 %, F (2, 592) = 134.46. When all variables were entered into the structural equation, service quality could predict patient behavioral intention by 31.2 %, and β = 0.218, t = 5.787, p < 0.001; patient satisfaction could predict patient behavioral intention by 27.3%. β = 0.617, t = 16.342, p < 0.001. Table 3: Simple Regression Equation and Path coefficient of the Study (n = 595).

Variable R-square Unstandardized coefficient, B

(SE)

Standardized coefficient,

Beta

t p-value

Patient satisfaction Service quality 0.185 0.431 (0.037) 0.431 11.635 0.000

CM


187


Table 4: Multiple Regression Equation and Path Coefficient of Hypothesis 2 (n = 595).Variable R-square Unstandardized

coefficient, B (SE)

Standardized coefficient,

Beta

t p-value

Patient behavioral intention Service quality Patient satisfaction

0.3120.273

0.174 (0.030)0.490 (0.030)

0.2180.617

5.78716.342

0.0000.000

e path correlations between service quality and patient satisfaction, between patient satisfaction and patient behavioral intention, and between service quality and patient behavioral intention were positive (β = 0.431, 0.617 and 0.218 respectively, P < 0.001) (Figure 2).

Figure 2. Path Model of PSQ-HAART *p < 0.001

e indirect effects of service quality on patient behavioral intention were examined on an exploratory basis. Table 5 showed a decomposition of the standardized effects into direct, indirect, spurious and total effects categories, the latter of which was the sum of the former three. While service quality had a significant direct effect on the patient behavioral intention score (β = 0.218, p < 0.001) it had a significant indirect effect on patient behavioral intention scores via patient satisfaction (β = 0.266, p < 0.001).

Indirect PathService quality Satisfaction Behavioral intention = 0.266*

Satisfaction

Service quality Behavioral intention

0.431* 0.317*

0.218*

CM


188


Table 5: Decomposition of Standardized Effects for Path Model Predicting Patient Satisfaction and Behavioral Intention Scores (n = 595).

Predictor VariableVariable

Patient satisfaction

Patient behavioral intention

Service quality Direct effectIndirect effect via patient satisfactionSpurious effect

0.431*-

0.218*0.266*0.09

Total effect 0.431* 0.574*Patient satisfaction Direct effect - 0.617*

*p < 0.001

DISCUSSIONPrediction of Patient satisfaction

To test the pharmacist model of PSQ-HAART, prediction of patient satisfaction by the service quality was examined. e service quality

component of the model was found to have a significant positive, direct effect on patient satisfaction. A higher level of patient satisfaction with the service was more likely when patients indicated they had a higher service quality score (β= 0.431, P < 0.001). is was in accordance with Cronin and Taylor’s (1994) conclusion that health service customers might not distinguish service quality from satisfaction when they responded to patient satisfaction or service quality surveys. In addition, the result from this study correspond to the others that perceptions of service quality were positively associated with overall patient satisfaction (Lazarus, 1991; Bagozzi, 1992; Cronin and Taylor, 1992; Gotlieb et al., 1994; Sudibyo, 2002).

Prediction of Behavioral Intention is study used the PSQ-HAART model to confirm whether patient satisfaction would predict behavioral intention. Patients who reported higher levels of satisfaction with pharmacy service tended to report higher levels of behavioral intention (β = 0.617, P < 0.001). is finding suggests the strong positive association of patient satisfaction and patient behavioral intention, corresponding with previous studies. (Goltieb et al., 1994; Brady and Robertson, 2001; Sudibyo, 2002) In addition to the prediction of patient behavioral intention by patient satisfaction, patients who indicated higher perception of service quality tended to report higher levels of behavioral intention (? = 0.218, P < 0.001). Our result suggests that service quality positively affected pa-tient behavioral intention through the association with patient satisfaction

CM


189


(β = 0.266, P < 0.001). is indirect association between service quality and behavioral intention via patient satisfaction as a mediator variable correspond with the studies by Cronin and Taylor (1992) and Rust and Oliver (1994). e influence of perceptions of service quality on behav-ioral intentions was considerably more integrated than was reported in the literature. e result suggests that patients’ perception of service quality was the important predictor of patient satisfaction. Patient satisfaction was important in establishing patient behavioral intention as a mediator between service quality and behavioral intention. is implied that from the patients’ perception, a high level of service quality must be developed before satisfying the patient. If patients are satisfied with pharmacy service of the HAART program, they would behave in a way that reflects posi-tively towards the program, that is, continuing antiretroviral drugs adher-ence, complying closely with pharmacist’s advices, keeping pharmacist’s appointments, and communicating positive things about the pharmacy service to others. Patients’ decision-making, relative to their behavioral intention of service quality, is a complex and comprehensive process. at is, the use indirect path analysis suggests that consider only individual variables or direct effects are likely to yield incomplete assessments of the relation-ships among them. Establishing plans to improve only service quality or patient satisfaction might not be sufficient without improving the other. Pharmacy administrators may consider strategies in proving both service quality and patient satisfaction of the HAART program to achieve the desired behavioral intention of the patients. e study had some limitations that should be considered. First, this study was the analysis of causal model based on cross-sectional data; strong evidence of causal effects cannot be inferred. Second, the study model was designed to include only two variables related to the behavioral intention other factors that might also be influenced to the behavior intention were not studied and could not be explained. ird, the model chose behavioral intention to represent behavior. Although, it has been proved a good indicator of actual behavior the outcomes in from the HAART program on actual behaviors were not studied. Lastly, since the model was built from HAART program for HIV/AIDS patients and carried out in only community hospitals in Chiang Rai province in ailand generalizability may be limited to other disease models or types of hospitals. There is the need for further consideration of similar models. Although our results were rather consistent across the pharmacy service of the HAART program provided by community hospitals in Chiang Rai

CM


190


province, there were some variations worth attention. For future research, the study design should include the various decision-making variables which might effect the patients’ perception on service quality, their satisfaction and behavioral intention. ese included the quality of service environ-ment, individual patients and service characteristics such as experience or expertise. Finally, this study also shows the need for additional research that considers the influence of service quality and patient satisfaction on value and trust, and the influence of these variables on patient behavioral intention and health care outcome. Further research is needed to develop better measures of the con-structs and the revision should be repeatedly tested with a sample group from hospitals in other provinces in order to guarantee its clarity and the appropriateness of this model in community hospitals throughout the country.

CONCLUSION

The objective for this study was to clarify the relationships between service quality, satisfaction, and behavioral intention in pharmacy

service of HAART program. It suggested that the patient decision-making process for service quality was a complex model that incorporates both direct and indirect effects of service quality on behavioral intention. Service quality has direct effect on patient satisfaction and patient satisfaction has direct effect on patient behavioral intention. Specifically, there is evidence that service quality and satisfaction directly influenced behavioral intentions, even when the effects of all two constructs were considered simultaneously. Service quality does not only have indirect effect on patient behavioral intention by increasing the satisfaction associated with pharmacy services, it also influences patient behavioral intention directly.

ACKNOWLEDGEMENTS

Thank you for the supporting of the University Income of 2005, Faculty of Pharmacy, Chiang Mai University; the Graduate School and

the Norwegian Church AID. anks for collaboration go to the Chiang Rai regional hospital, all community hospitals in Chiang Rai province and all participants in the study. Special thanks go to Suparet Wongputh, statistic consultant.

CM


191


REFERENCESAnderson, E. W., and M. Sullivan. 1993. e antecedents and conse-

quences of customer satisfaction for firms. Marketing Science. 12: 125-143.

Bagozzi, R. P. 1992. The self regulation of attitudes, intentions and behavior. Social Psychology Quarterly. 55: 178-204.

Bartlett, E., M. Grayson., and R. Barker. 1984. e effects of physician communication skills on patient satisfaction, recall and adherence. Journal of Chronic Disease. 37: 755-64.

Bitner, M. J., and A.R. Hubbert. 1994. Encounter satisfaction versus overall satisfaction versus Quality: e customer’s voice. P. 72-94. In RT. Rust and R.L., Oliver. (eds.). Service Quality: New Directions in eory and Practice. CA: Sage, ousand Oaks.

Bolton, R.N., and J.H. Drew. 1991. A longitudinal analysis of the impact of service changes on customer attitudes. Journal of Marketing. 55: 1-9.

Brady, M.K., and C.J. Robertson. 2001. Searching for a consensus on the antecedent role of service quality and satisfaction: An exploratory cross-national study. Journal of Business Research. 41: 53-60.

Bureau of Epidemiology Department of Disease Control, Ministry of Public Health. 2005. e number of HIV/AIDS patients classified by the region and province during September, 1984 to October 31, 2005. [Online] Available http://epid.moph.go.th. [2005, November 2]

Chaitha, W., C. Viboonsunti, T. Kittisopee, and R. Awipan. 2005. e development of patients’ perceptions of pharmacy service quality of highly-active antiretroviral therapy program questionnaire in Chiang Rai community hospitals, ailand. Unpublished paper.

Cronin, J.J., and S.A.Taylor. 1992. Measuring service quality: A reexami-nation and extension. Journal of Marketing. 56: 55-68.

Cronin, J.J., and S.A. Taylor. 1994. SERVPERF vs SERVQUAL: Recon-ciling performances based perception - minus - expectations measure-ment of service quality. Journal of Marketing. 58: 125-131.

Cronin, J.J., M.K. Brady, and G.T.H. Hult. 2000. Assessing the effects of quality, value, and customer satisfaction on customer behavioral intention in service environments. Journal of Retailing. 76(2) : 193-218.

Crosby, L., K. Evan, and D. Cowles. 1990. Relationship quality in services selling: An interpersonal influence perspective. Journal of Mar-keting. 54: 68-81.

Dabholkar, P. 1995. e convergence of customer satisfaction and service quality evaluations with increasing customer patronage. Journal of Consumer Satisfaction, Dissatisfaction and Complaining Behavior. 8: 31-43.

CM


192


Daniel, W.W. (1995). Biostatistics : A foundation for analysis in the health sciences. (6th ed.). New York : John - Wiley & Sons.

Gotlieb, J.B., D. Grewal, and S.W. Brown. 1994. Consumer satisfaction and perceived quality: Complementary or divergent constructs? Journal of Applied Psychology. 79 (6): 875-885.

Horizon Program. 2004. (Community Medicine Department, Chiang Mai University; Office of Disease Prevention and Control Region 10, Chiang Mai, AIDS Section, Ministry of Public Health). A rapid situation analysis of the access to care project in Northern ailand. e Population Council Inc., Bangkok, ailand.

Hunt, H.K. 1977. Conceptualization and measurement of consumer satisfaction and dissatisfaction, (Report No. 77-103). Cambridge, MA: MSI.

John, J. 1991. Research in brief, Patient satisfaction: e impact of past experience. Journal of Health Care Marketing. 12(3): 56-64.

Lazarus, R.S. 1991. Emotion and adaptation. Oxford University Press, New York.

Ley, P. 1988. Communication with patients: Improving communication, satisfaction and compliance. Croom Helm, New York.

Norusis, M.J. 1990a. SPSS advance statistics student guide. SPSS Inc., Chicago. ILL.

Norusis, M.J. 1990b. SPSS base system user’s guide. SPSS Inc., Chicago. ILL.

Parasuraman, A., V.A. Zeithaml, and L.L Berry. 1985. A conceptual model of service quality and its implications for future research. Journal of Marketing. 49(Fall): 41-50.

Parasuraman, A., V.A. Zeithaml, and L.L Berry. 1988. SERVQUAL: A multiple-item scale for measuring consumer perceptions of service quality. Journal of Retailing. 64: 12-40.

Peck, B.M. 2002. Doctors and patients : Expectations, control and health care satisfaction. [Online] Available http://wwwlib.umi.com/dissertations /preview_all/306870 [2003, November 23].

Rust, R.T., and R. Oliver. 1994. Service quality: Insights and managerial implications. In R.T. Rust and R.L. Oliver (eds.). Service Quality: New Directions in eory and Practice. CA: Sage Publications. Tierney, ousand Oaks.

Scott-Lennox, J., J.F. Braun, and J.E. Morrow. 1998. Development of the HIV treatment

satisfaction survey (HTSS) to improve adherence and quality of out patient care. Int. Conf AIDS. 12: 102. [Online] Available http// www.aegis.com/pubs/aissline/1998/dec/m98c5916.htm [2003, September24].

CM


193


Sudibyo, Y.A. 2002. e relationship of business process capabilities and internal service quality to health service quality patient satisfaction, and business performance. Ph.D. Dissertation. University of Nova Southeastern, Wayne Huzenga of Business and Enterpreneurship.

e AIDS Section, Chiang Rai Provincial Health Officer. 2005. Chiang Rai, ailand.

Woodside, A.G., L.L. Frey., and R.T. Daly. 1989. Linking quality service, customer satisfaction and behavioral intention. Journal of Health Care Marketing. 9 (4): 5-17.

CM


194


AP

PEN

DIX

e

PSQ

-HA

AR

T Q

uest

ionn

aire

Serv

ice

Qua

lity

Whe

n I

rece

ived

the

phar

mac

y se

rvic

e....

......

Expe

ctat

ion

Perc

eptio

nLo

w

Hig

hLo

wH

igh

1. M

ater

ials

asso

ciat

ed w

ith p

harm

acy

serv

ice

are

suita

ble-

appe

alin

g; fo

r ex

ampl

e, a

ntire

trovi

ral o

r ant

i- op

port

unist

ic

infe

ctio

n pa

mph

lets.

12

34

51

23

45

2. I

rece

ive

antir

etro

vira

l inf

orm

atio

n co

mpl

etel

y be

fore

star

t th

e dr

ugs.

12

34

51

23

45

3.

e ph

arm

acist

insis

ts on

my

heal

th c

are

reco

rd.

12

34

51

23

45

4.

e ph

arm

acist

hel

ps m

e m

anag

e m

y an

tiret

rovi

ral s

ched

ule

that

acc

omm

odat

es fo

r my

life

style

.1

23

45

12

34

5

5.

e ph

arm

acist

pro

vide

s mem

ory

tabl

e fo

r my

antir

etro

vira

l se

lf-ad

min

istra

tion.

12

34

51

23

45

6.

e ph

arm

acist

usu

ally

mon

itors

my

antir

etro

vira

l the

rapy

ou

tcom

e.1

23

45

12

34

5

7.

e ph

arm

acist

hel

ps m

e ca

re m

y he

alth

.1

23

45

12

34

58.

Whe

n I

have

a p

robl

em, t

he p

harm

acist

show

s a si

ncer

e in

tere

st in

solv

ing

it.1

23

45

12

34

5

9.

e ph

arm

acist

doe

s ser

vice

by

a ce

rtai

n tim

e of

AT

C c

linic

.1

23

45

12

34

510

. e

phar

mac

ist h

as th

e kn

owle

dge

to e

nsur

e m

y qu

estio

ns.

12

34

51

23

45

CM


195


Whe

n I

rece

ived

the

phar

mac

y se

rvic

e....

......

Expe

ctat

ion

Perc

eptio

nLo

w

Hig

hLo

wH

igh

11.

e ph

arm

acist

has

app

ropr

iate

inte

rper

sona

l co

mm

unic

atio

n an

d co

unse

ling.

12

34

51

23

45

12.

e ph

arm

acist

serv

ices

me

suita

bly

acco

rdin

g to

pr

ofes

siona

l rol

e.1

23

45

12

34

5

13. I

feel

safe

in th

e tr

ansa

ctio

n of

the

phar

mac

ist.

12

34

51

23

45

14. I

rece

ive

serv

ice

from

the

phar

mac

ist e

quity

to o

ther

pa

tient

s.1

23

45

12

34

5

15.

e ph

arm

acist

giv

es m

e in

divi

dual

atte

ntio

n.1

23

45

12

34

516

. e

phar

mac

ist h

as m

e be

st in

tere

st at

hea

rt.

12

34

51

23

45

17.

e ph

arm

acist

und

ersta

nds m

y sp

ecifi

c ne

eds.

12

34

51

23

45

18.

e ph

arm

acist

resp

onse

s to

my

need

s.1

23

45

12

34

519

. e

phar

mac

ist re

spon

ses t

o m

y fe

elin

g ap

prop

riate

ly.1

23

45

12

34

520

. e

phar

mac

ist g

ives

me

pers

onal

atte

ntio

n.1

23

45

12

34

521

. e

phar

mac

ist re

spec

ts m

y rig

ht to

mak

e m

y ow

n ch

oice

.1

23

45

12

34

522

. e

phar

mac

ist h

olds

my

info

rmat

ion

prov

ided

secu

re.

12

34

51

23

45

23.

e ph

arm

acist

tells

me

exac

tly w

hen

serv

ice

will

be

perfo

rmed

.1

23

45

12

34

5

24.

e ph

arm

acist

hel

ps p

hysic

ians

in c

arin

g m

e.1

23

45

12

34

525

. e

phar

mac

ist p

rovi

des m

e co

ntin

uous

car

e w

ith

phys

icia

ns a

nd o

ther

hea

lth c

are

prov

ider

s.1

23

45

12

34

5

CM


196


Whe

n I

rece

ived

the

phar

mac

y se

rvic

e....

......

Expe

ctat

ion

Perc

eptio

nLo

w

Hig

hLo

wH

igh

26.

e ph

arm

acist

sets

syste

m to

con

tact

me

if I

miss

ed th

e ap

poin

tmen

t.1

23

45

12

34

5

27.

e ph

arm

acist

giv

es p

rom

pt se

rvic

e fo

r me.

12

34

51

23

45

28.

e ph

arm

acist

is a

lway

s will

ing

to h

elp

me.

12

34

51

23

45

29.

e ph

arm

acist

esta

blish

es a

nd m

aint

ains

rapp

ort w

ith m

e to

ens

ure

my

adhe

renc

e to

ant

iretro

vira

l the

rapy

.1

23

45

12

34

5

30.

e be

havi

or o

f the

pha

rmac

ist in

still

confi

denc

e in

me.

12

34

51

23

45

31.

e ph

arm

acist

em

pow

ers m

e.1

23

45

12

34

532

. I c

an c

onsu

lt th

e ph

arm

acist

abo

ut so

cioe

cono

mic

issu

es.

12

34

51

23

45

33. I

can

con

sult

the

phar

mac

ist b

y ph

one

if I

need

.1

23

45

12

34

5R

emar

k: ta

ngib

les-

relia

bilit

y (it

ems 1

thro

ugh

8) a

ssur

ance

-em

path

y (it

ems 9

thro

ugh

16);

empa

thy

(item

s 17

thro

ugh

22);

resp

onsiv

enes

s I (i

tem

s 23

thro

ugh

26);

resp

onsiv

enes

s-as

sura

nce

(item

s 27

thro

ugh

31);

resp

onsiv

enes

s II

(item

s 32

thro

ugh

33).

CM


197


Wilawan Senaratana1, Wanchai Moongtui1*, Wichit Srisuphan2, Nunta Leksawasdi1, Poonsap Soparat1, Susanha Yimyam1,

Jittaporn Chittreecheur1, Wilawan Picheansathian1 and Wipada Kunaviktikul1

Development of Quality Indicators on Important Aspects of Nosocomial Infection Prevention and Control

1Faculty of Nursing, Chiang Mai University, Chiang Mai 50200, ailand 2Professor Emeritus, Faculty of Nursing, Chiang Mai University, Chiang Mai 50200, ailand *Corresponding author. E-mail:[email protected]

ABSTRACT

Nosocomial infection (NI) stills remain a problem in ailand. An infection prevention and control program plays an impor-

tant role in reducing morbidity and mortality due to NI. e purpose of this study was to develop quality indicators (QIs) on the following aspects of NI: surgical site infection, urinary tract infection, ventilator-associated pneumonia, intravenous catheter-associated infection, standard precau-tions-related infection, dialysis-related infection, hospital waste manage-ment-related infection and employee health-related infection. is study was divided into 4 stages, including current situation analysis, development of QIs, feasibility and applicability test of the QIs and the forum for the QIs refinement. Major findings are that the majority of experts agreed with the QIs items and the majority of infection control nurses rated those items as fea-sible for the use. e QIs were categorized into two groups: general QIs and specific QIs. e specific QIs were divided into 8 aspects and each aspect consisted of three components: structure, process, and outcome. erefore, most of the QIs developed from this study could be used to measure the quality of the nosocomial infection prevention and control in the context of ai health care delivery system. Health care workers, especially infection control nurses, should try out in their settings to test for the actual use of the indicators. For further research, it is necessary to test the use of quality indicators in actual practice.

Key words: quality indicators; infection control; nosocomial infection

CM


198


BACKGROUND AND SIGNIFICANCE

In many countries, NI is still considered as a problem for almost all levels of settings. For example, Sax and co-investigators from the University of

Geneva Hospital, Geneva, Switzerland studied the 7-day-period prevalence of NI in large primary and tertiary health care centers in Geneva and found that overall prevalence of NI was 11.3%. e investigators concluded that sub-acute and chronic care settings are associated with high infec-tion prevalence (Sax et al., 2001). In ailand, few studies on NI have been published. One study examined surgical site infection rates among provincial, regional and university hospitals and found NI rates ranging from 6.8 to 12.7 infections per 100 surgical operations (Danchaivijitra, 2001). e rate of infection is higher than that of approximately 8.69% in other countries (Mayhall, 1993). ese high prevalence rates may be a reflection of heightened awareness of the surveillance team or a break-down of cross-infection protective measures. In addition, inappropriate timing of prophylactic antibiotic administration, length of hospital stay and duration of operation may be contributory factors to post-operative wound infection. In another study, Srisuphan et al. (1989) found 6.3% of patients developed NI in a three-month period during the study. e investigators (1989) also found that NI had caused patients to remain in the hospital 17 days longer and the mortality rate was 3.8 times higher than that of patients who did not have NI. Based on the results of these studies, the number of infection may actually be considerably higher since the information on NI is not yet available from outpatient departments or discharged patients. In ailand, overcrowding and understaffing of wards occur in many levels of hospitals; it is unreasonable to assume that all NI are identified and documented and thus NI rates may be higher than that found in the reviewed studies. In addition, recent data showed that patients are now more immunocompromised, more severely ill and receiving complex treatment and thus more susceptible to infection (Calfee and Farr, 2002). Many aspects of infection control challenges for ailand are posed by these changes, including surgical site infection, urinary tract infection, ventilator-associated pneumonia, intravenous catheter-associated infection, standard precautions-related infection, dialysis-related infection, hospital waste-related infection and hospital employee health-related infection. Sur-gical site infection (SSI) rate in ailand ranges from 6.8-12.7% which is higher than that of developed countries (Danchaivijitra, 2001). SSI prolonged hospital stay up to 8 to 12 days and cost the patients and the hospital up to 500 million ai Baht (38 Baht = 1 US$) (Danchaivijitra, 2001). Patients admitted to the hospital are at risk of having urinary tract infection. Urinary tract infection (UTI) rates found among hospitalized

CM


199


patients in ailand ranged from 19.7-32.3% ( Danchaivijitra, 2001) which was rather high, compared with the rate of 6-12.6% found in many developed countries (Maki and Tambyah, 2001). UTIs caused approxi-mately 431 deaths of hospitalized patients annually and cost ai hospital more than 400 million baht annually. e closed-system drainage of the urinary catheter and the aseptic technique in the care of the catheter are important aspects of preventing UTIs. However, the prevalence of UTI still ranks high in ailand. With the advancement of technologies in the care of the respiratory system, a ventilator is commonly used to maintain air ventilation among the patients. As a result, ventilator-associated pneumonia (VAP) is very common, especially in the intensive care unit (ICU) due to high use of a ventilator among susceptible patients. In a study of the rate of VAP in the ICU of Ramathibodi Hospital in 1999, Piamsuwam and colleagues (2001) found that the rate of VAP was 17 infections per 1,000 ventilator-days. is rate is high, compared with that of 13.2 infections per 1,000 ventilator-days found in the United States (National Center for Infectious Diseases, 2001). In ailand, VAP prolongs a hospital stay up to 4 to 21 days (Piamsuwan et al., 2001). e benefits of intravenous catheter lines for the patients are enor-mous. Unfortunately, many of the patients develop intravenous catheter-associated infection and consequently cause serous illness such as bacte-remia. Wildmer (1997) found that 50,000 patients developed bacteremia annually in the United States associated with intravenous catheter uses. In ailand, Danchaivijitra (2001) found that 3.2 primary septicemia cases per 1,000 patients were found to be associated with the use of intravenous catheter line. is incidence caused the patients to stay 12 to 29 days longer in the hospital. Bringing together a large group of sick people in a hospital environ-ment has some advantages but there are also disadvantages such as transmis-sion of infection from one person to another, especially in a hospital where the wards are overcrowded and understaffed in many health care facilities in ailand. Consistent practices of standard precautions are crucial such as handwashing, wearing protective equipment, appropriate management of patients’ belongings. ese practices should be implemented in all levels of health care settings and to all patients regardless of their health status. However, there is a plethora of evidence that ai health care workers (HCWs) still do not adhere the recommended practices (Moongtui et al., 2000; Danchaivijitra, 2001). e use of hemodialysis for the renal failure patients brings them better quality of life. e process of hemodialysis, however, can potentially

CM


200


induce microorganisms into the body of the patient, causing septicemia as evidenced by a study done by Neilson, et al. (1998) in which 33% of dialysis patients had septicemia. Also, many patients may acquire blood-borne pathogens such as Hepatitis B and C viruses from the hemodialysis process. e consequences of the DRI are similar to that of the mentioned intravenous catheter-related infection such as prolonged hospital stay, dis-ability and even death. us, the DRI prevention is crucial. Hospital infectious wastes should be managed properly to prevent the spread of infection to the patients and health care workers. In ai-land, Wongkrea (1997) found that 76% of university hospital garbage bags contained used sharp needles. In another study done by Boparai (1997), examining needle stick injuries among hospital workers who disposed hospital waste, found that 8% of workers had a history of needle stick injuries from the waste management process. e inappropriate practices of collecting and disposing wastes may cause health workers and the pa-tients to acquire blood borne diseases such as Hepatitis B and C viruses and AIDS. Transmission of blood-borne pathogens, such as Hepatitis B virus (HBV), Hepatitis C virus (HCV) and human immunodeficiency virus (HIV) in health care settings are well-documented. Nosocomial transmission of HBV is a serious risk for health care personnel. Approximately 1,000 health care personnel were estimated to have become infected with HBV in 1994 (Bolyard, 1998). During the past decade, the annual number of newly acquired HCV infections has raised from an estimated 180,000 in 1984 to an estimated 280,000 in 1995. Of these, an estimated 2% to 4% occurred among health care personnel who were occupationally exposed to blood. According to prospective studies of health care person-nel, percutaneously exposed to HIV-infected blood, the average risk for HIV infection has been estimated to be 0.3%. e risk of transmission of infection among personnel can be reduced by handwashing before and after all patient care and by the use of appropriate barriers, such as a mask and glove to prevent the contagious agent. However, the compliance practices among personnel are still low. Quality indicator (QI) is a methodology, developed from the principle of total quality improvement, used widely in hospitals in North America to improve the effectiveness and efficiency of care (Huskins et al., 1998). It provides description of structure, process and outcome of the interested aspect of quality. In recent years, quality indicator has gained increasing attention as a tool to evaluate and monitor the delivering of health care services such as nursing services (Katz and Green, 1992). In the area of NI prevention and control in ailand, there has been no study that develops

CM


201


QIs specifically for the prevention and control of NI yet. e benefits of QIs developed for NI prevention and control is enormous. For health care workers (HCWs) especially infection control nurses (ICNs), QIs would provide guidelines, developed based on the con-sideration in the context of financial, cultural and behavioral aspects, for them to do the infection control practices. is will gain the acceptance from them to do the infection control and prevention activities. e QIs will also benefit the hospital administrators and infection control practitio-ners; so they can analyze, plan, monitor and evaluate the infection control program in the institution and consequently plan the intervention activities related to infection control. e hospital administrations would benefit from the QIs to monitor the infection control program and manage the program effectively. It is, therefore, our purpose to develop QIs related to important aspects of nosocomial infection prevention and control.

OBJECTIVES

To develop quality indicators on the following aspects of nosocomial infection prevention and control; surgical site infection, urinary tract

infection, ventilator-associated pneumonia, intravenous catheter-associated infection, standard precautions-related infection, dialysis-related infection, hospital waste management-related infection and employee health-related infection.

MATERIALS AND METHODS

In this study, there were 4 Stages of the development of QIs: Stage 1, Current Situation Analysis; Stage 2, Development of the QIs; Stage 3,

Feasibility and Applicability test of the QIs; and Stage 4, Forum for the QIs refinement

Setting and Sample Sample consists of (a) group of experts in the area of nosocomial infection control throughout the country, recruiting by name nominations by the investigators, five experts were recruited for each aspect of NI (b) group of health care workers (HCWs) responsible for the NI prevention and control from all levels of health care settings including physicians, nurses and health care workers working in dialysis unit, numbers of subjects varied from one aspect to another using stratified random sampling method and (c) infection control nurses (ICNs) and infection control ward nurses (ICWNs) in all regions throughout the country from university, regional,

CM


202


general and community hospitals. Numbers of subjects also varied from one aspect of NI to another, using stratified random sampling technique.

Instruments Instruments used to collect data were; 1. A set of questionnaires to test for availability and use of QIs. A set of open ended and semi-structure questionnaire comprised of information for the availability and use of QIs among health care workers for each aspects of NI which was constructed by the investigators. 2. A set of questionnaires to test for the validity of the proposed QIs. A set of questionnaire for each aspect of NI, developed by the investigators, requesting experts to check whether they agree or disagree with each of the proposed QIs item. 3. A set of questionnaires to test for the feasibility and applicability of the proposed QIs. A set of questionnaires, developed by the investigators, was used requesting infection control nurses’ opinions regarding the feasibility and applicability for the use in the practices. All instruments were validated by experts for face validity. eir suggestions were incorporated into the instruments.

Data Collection Procedures Data were collected as follow; Stage 1 Current Situation Analysis e purpose of this stage was to analyze the current situation of the availability and use of QIs in the practice in a health care setting. A set of questionnaires for availability and use of QIs was sent to HCW and requesting them to return a set of questionnaires by mail. e investigators also reviewed literature regarding the use of QIs. Stage 2 Development of quality indicators e purpose was to develop the QIs items and QIs description. Literature related to QIs and its use were reviewed in addition to the results from Stage 1. en, the QIs were developed for each aspect of NI. After that, the QIs were incorporated into a set of questionnaires to test for the validity of the proposed QIs before sending to the experts to rate whether they agree or disagree with the QIs items. If disagree, they were requested to make comments and recommendations on them. e content validity indices were then analyzed. e QIs were revised based on the comments and recommendations from experts. Stage 3 Feasibility and applicability test of the QIs e purpose was to test for feasibility and applicability of the QIs. e QIs developed from Stage 2 were incorporated into a set of question-

CM


203


naires to test for the feasibility of the proposed QIs before sending to infection control nurses. Infection control nurses were requested to rate whether the items were applicable to their settings or not. If they answer ìnot applicableî, they were requested to provide the reason for not being applicable such as not necessary, not applicable to the current time or no data available yet. en, the QIs were revised and modified based on the results from this stage. Stage 4 Forum for the QIs refinement e purpose was to refine the QIs developed in Stage 3. Group of experts, ICNs, and HCWs were invited to attend a one-day forum. e QIs from Stage 3 were presented, followed by the discussion on the issue of validity and applicability of the QIs items, using panel discussion. en, the QIs were refined and proposed as the agreed QIs.

Data Analysis In Stage 1, the QIs items were analyzed by calculating the percent-age of the agreement by items as rated by participants and compared with the items from reviewed literature. In Stage 2, the content validity index (CVI) was calculated, based on the responses from experts. e QIs were categorized into structure, process and outcome components. In Stage 3, the percentage and number of items were analyzed for the feasibility, based on the responses from the ICNs. In Stage 4, the refined QIs were categorized into general and specific QIs. As for the specific QIs, the items were grouped for each aspect and categorized into structure, process and outcome components.

RESULTS e details of findings are as follow: Stage 1 e situation analysis and the reviewed literature of the quality indicators. In this stage, a set of questionnaire were sent to HCWs. Response rates ranged from 70.96%-87.88%. e response rate is highest among subjects in the aspect of ventilator associated-pneumonia (n = 99, 87.88%), followed by the aspect of central venous catheter-related infection (n = 99, 86.80%). Response rates of other aspects were not much varied (standard precautions related infection [n=491, 78.95%], urinary tract infection [n=491, 78.95%], surgical site infection [n=396, 78.28%], employee health- related infection [n=817, 71.60%], infectious waste-related infec-tion [n=817, 71.36%] and dialysis-related infection [n=88, 70.96%].).

CM


204


e Availability and Use of the QIs Based on the results from the subjects who responded to the ques-tionnaire, it was found that the availability and the use of QIs among subjects varied for each aspect of the NI as shown in Table 1. A number of items was comparable in each category of structure, process and outcome. It was also evident that the number of items found in the literature was also comparable to that obtained from the subjects.

Table 1: Number and percentage of availability and use of QIs items reported by subjects and reviewed from literature.

NI ASPECTS Items available and use number (percentage)

Items reviewed number (percentage)

Infectious waste-related infection Structure 15 (50.00) 17 (53.13) Process 5 (16.67) 5 (15.63) Outcome 10 (33.33) 10 (31.25)Employee health-related infection Structure 17 (36.17) 16 (21.33) Process 23 (48.94) 46 (61.33) Outcome 17 (36.17) 13 (17.34)Central venous catheter-related infection Structure 10 (47.62) 10 (31.25) Process 6 (28.57) 17 (53.13) Outcome 5 (21.81) 5 (15.62)Surgical site infection Structure 17 (34.00) 7 (29.17) Process 19 (38.00) 12 (50.00) Outcome 14 (28.00) 5 (20.83)Standard precautions Structure 8 (38.09) 8 (38.09) Process 8 (38.09) 8 (38.09) Outcome 5 (23.80) 5 (23.80)Urinary tract infection Structure 10 (62.50) 10 (50.00) Process 4 (25.00) 6 (30.00) Outcome 2 (12.50) 4 (20.00)Dialysis-related infection Structure 8 (26.67) 10 (27.78) Process 13 (43.33) 17 (47.22) Outcome 9 (30.00) 9 (25.00)

CM


205


Ventilator-associated Pneumonia Structure 12 (50.00) 9 (18.36) Process 9 (37.50) 35 (71.42) Outcome 3 (12.50) 5 (10.20)

Stage 2 e development of QIs Based on the result from Stage 1, the investigators developed and proposed the QIs. e QIs of each aspect of NI were categorized into three components; structure, process and outcome. en, the proposed QIs were sent to experts. Each QIs item was reviewed by five experts for each aspect of NI. Experts were requested to rate and make comments on those items and returned the QIs to the investigators. en, the items were modified based on the comments from experts. e content validity index (CVI) for all aspects ranged from 0.82 to 0.95. Stage 3 Feasibility test of the QIs e QIs from Stage 2 were sent to ICNs and ICWNs who had experience in each aspect of NI to test for feasibility for the use of QIs. e number of subjects who responded to the feasibility test varied from 70.97% to 87.87%. e items were analyzed for the feasibility in the practice. As shown in Table 2, most items of the QIs were rated as feasible for the use except for the aspect of infectious waste-related infection (2 items), employee health-related infection (8 items), central venous catheter-related infection (1 items), surgical site infection (3 items), urinary tract infection (1 item) and dialysis-related infection (1 item).

Table 2: e feasibility of QIs components.Aspect of nosocomial

infectionNumber of QIs items

feasible (more than 50% agreed) Number (percentage)

Infectious waste-related infection Structure 27 27 (100%)

Process 18 18 (100%)

Outcome 6 4 (66.66%)

Employee health-related infection Structure 16 15 (93.75%)

Process 33 29 (87.87%)

Outcome 3 6 (66.66%)

Central venous catheter-related infection Structure 12 12 (100%)

Process 21 20 (95.23%)

Outcome 5 5 (100%)


CM


206


Surgical site infection Structure 23 22 (95.65%)

Process 45 44 (97.77%)

Outcome 2 1 (50.00%)

Standard precautions Structure 7 7 (100%)

Process 30 30 (100%)

Outcome 4 4 (100%)

Urinary tract infection Structure 10 10 (100%)

Process 13 13 (100%)

Outcome 5 4 (80.00%)

Dialysis related infection Structure 26 26 (100%)

Process 32 31 (96.87%)

Outcome 7 7 (100%)

Ventilator-Associated Pneumonia Structure 8 8 (100%)

Process 34 34 (100%)

Outcome 4 4 (100%)

Stage 4 Forum for the QI refinement In this stage, all subjects were invited to participate in the one-day forum to refine the QIs. Twenty-one experts and 17 ICNs and health care workers attended the forum. During the session, the result from each stage of the study was presented to the participants. Also, the result documents were distributed to all participants. en, the participants were assigned into 8 groups according to the aspect of NI using a small group technique to discuss and refine the QIs. en, all participants were invited to join together. e investigators then presented the results from the small group discussion. e participants were encouraged to discuss and find consen-sus for the QI items. e results showed that each aspect of the QIs had comparable number and percentage of items. In addition, the participants proposed that some QI items could be categorized into general and specific indicators. Based on the results, the quality indicators composed of two categories; general quality indicators and specific quality indicators. As for the specific indicators, they were grouped into 8 aspects. Each aspect had three components including structure, process and outcome.


CM


207


1. General Quality Indicators 1.1 Structure • e infection control committee in the hospital must be ap-pointed. • e hospital has the written policy and strategies covering im-portant aspects of nosocomial infection prevention and control (NIPC). • e hospital appoints individual responsible for the NIPC both at organizational and unit levels. • e hospital has the surveillance system covering important aspects of NIPC. • e hospital has standards, guideline and manual of the practice covering important aspects of NIPC. • e hospital has the plan of a training of the personnel cover-ing important aspects of NIPC. • e hospital must identify and establish the definition of the nosocomial infection diagnosis using agreed definition covering important aspects of NI. • e hospital provides sufficient number of human resources for the NIPC. • e hospital has the supervision and evaluation system for the practice of NIPC. e hospital has the policy regarding the communication, re-porting, and the feedback system for the NIPC. 1.2 Process • e hospital infection control committee conducts the meeting regularly. • e hospital implements the NIPC policies and strategies covering important aspects of NIPC. • e hospital implements the surveillance system and reports the results to the administrators. • e hospital conducts the supervision and evaluation of the NIPC program regularly. • Health care workers practise handwashing regularly and ap-propriately as follow: 1. Practise normal handwashing regularly before and after the patient contact. 2. Practise hygienic handwashing regularly before and after the surgical techniques, i.e., venous catheterization, blood transfusion by using either antiseptic solutions (such as 4% chlorhexidine gluconate, 5-7.5% povidone iodine or 3% hexachlorophene) or waterless antiseptic products.

CM


208


• Hospital employees use protective barriers appropriately and regularly as follow: 1.Wearing sterile glove for surgical techniques such as intra-venous catheterization. 2.Wearing gown, goggles, and mask whenever contact with blood and body fluid. 3.Wearing mask whenever acquiring respiratory infection and wearing high-performance mask, i.e., N95 for the prevention of tubercu-losis. 4.Wearing glove, mask, goggles, and apron whenever cleaning of equipment contaminated with blood and body fluid. 1.3 Outcome • Incidence of NI covering important aspects of NI. • Incidence of occupational acquired NI among hospital em-ployees. • Patient satisfaction with the important aspect of NI.

2. Specific quality indicators for NI 2.1 Urinary Tract Infection 2.1.1 Process • e person responsible for the urethral catheterization must be either physician, professional nurses or trained health care workers • ere must be a clear indication for the urethral cath-eterization. • ere must be an educational preparation for the patient undergoing the urethral catheterization in regard to the indication for the urethral catheterization and the prevention of complications. • The standards and guidelines must be followed for selecting the appropriate size of the catheter, the cleaning process prior to the catheterization and the use of aseptic technique strictly in all steps of the catheterization • e patients undergoing the urethral catheterization get appropriate care, including: 1. e catheter must be strapped on the lower part of the abdomen in men and on the area inner part of the upper thigh in women. 2. ere is no reflux of the urine and the closed drainage system must be maintained 3. e urine bag must hang freely and at the level of the bladder 4. e urine bag must be emptied correctly and the urine must be in the bag no more than 3⁄4 of the content of the bag

CM


209


5. e hands of health care workers must be washed before and after emptying the urine bag for each patient 6. e characteristics of the urine must be observed and evaluated at least every 8 hours for the signs of infection. 7. e perineal area must be flushed at least twice a day and every time after bowel movement. • Aseptic technique must be used for collecting urine specimen • e catheter must be replaced when there is a sign of clot or leakage. 2.1.2 Outcome • Rate of patients who compliance with the educational preparation • Compliance rates of the standard or guideline practices among health care workers 2.2 Ventilator-Associated Pneumonia 2.2.1 Process quality indicators • e appropriate practices of suctioning 1. ere is an indication for suctioning 2. e connecting parts and respirators must be cleaned with 70% alcohol and aseptic technique must be maintained 3. Sterile glove must be worn 4. e pressure for the suction must be 80 -120 mm.Hg. and the duration for the suction is between 10-15 seconds in adults • e position for the patient is in semi-Fowler position if no contra-indication • e appropriate practices of gavage feeding must be maintained 1. e head position of the patient is 30-45 degree 2. e secretion must be cleared prior to the feeding 3. e gastric content must be tested, if the content is more than 50 milliliter, the feeding on that meal must be omitted. 4. e gavage feeding tube must be closed after the feeding. • The appropriate care for respirators must be main-tained: 1. e strap plaster must be changed every 24 hours 2. e tracheostomy wound must be cleaned at least three times a day and the gauze must be used to support the area. 3. e respirators must be changed every three days but no less than 48 hours

CM


210


4. The self-inflating bag must be changed every 3 days 5. e distilled water must be changed for the humidi-fier every 8 hours for the open system and the humidifier must be cleaned with 70% alcohol. If the closed system is used, the humidifier must also be changed 6. e spirometer must be cleaned with 70% alcohol before and after use 7. e respirators must be cleaned after use • e respirators must be prepared for the use sufficiently for the patient and must be replaced according to the sterilization cycle. Also, the following equipment and tools must be prepared sufficiently for the use; suctioning tube, sterile glove and selfñinflating bag. 2.2.2 Outcomes • Rate of ventilator-associated infection 2.3 Central venous catheter-related infection 2.3.1 Process • e guideline for the prevention and control of CVC-related infection must be as follow: 1. e solution must be prepared using aseptic technique and cleaned with 70%alcohol 2. e patient must be prepared as follows; 2.1 e patient must be educated regarding the indica-tion and the care for central venous catheterization 2.2 e subclavian vein or the jugular vein must be first choice for the insertion site and the femoral vein must be avoided 2.3 e site for the insertion must be cleaned with an-tiseptic such as 4% Iodophore , 0.5% chlorhexidine or 70% alcohol and the insertion site must be intact with no sign of infection 3. e aseptic technique must be maintained during catheterization 4. e cover gauze must be changed when wet or after being cleaned or opened 5. e catheter must be replaced when there is a sign of clot , inflammation, infiltration and infection 6. If infected, the catheter must be cultured for the causative agent and the surveillance report must be recorded • e changing of solution 1. e solution set must be replaced every 72-96 hours and the solution must not be used more than 24 hours 2. e solution set for fat content must be replaced every 24 hours and the solution must be given within 12 hours

CM


211


3. e solution set for blood transfusion must be replaced after the use for each unit 2.3.2 Outcome Rate of septicemia related to the central venous catheter-ization. 2.4 Dialysis-related infection 2.4.1 Structure • e provision of guidelines or manual of patient screen-ing, immunization, nursing care, the use of protective barrier and the management of sterile water. • e use of guideline from medical council for the require-ments and recruitment of the personnel such as physician and professional nurse. • e management of acquiring appropriate facilities and adequate equipment 2.4.2 Process quality indicators • e patient must be screened for blood-borne diseases and isolated if indicated • Immunization for Hepatitis B virus must be provided if indicated. • e patient must be educated covering the aspect of personal hygiene, the care of the insertion site, the handwashing and observation of signs of infection. • e standards and the guideline of the care must be followed. • Aseptic technique must be used for the care • e guidelines for the re-use of the blood filter must be followed; 1. e tube and the filter must be cleaned with 4% formaline or other appropriate antiseptic and then washed with sterile saline solution. 2. e tube and the filter must be tested for the leakage. • e guideline for the cleaning of dialysis unit must be followed; 1. e unit must be cleaned and sterilized after use for each patient. 2.e unit must be cleaned using clean water, If con-taminated with blood and body fluid, the unit must be cleaned with 70% alcohol or 0.5-1% sodium hypochlorite. • e sterile water for the dialysis must be quality con-trolled using the agreed standards.

CM


212


2.4.3 Outcome quality indicators • rate of AVF/AVG infection • rate of double lumen catheter infection • rate of septicemia related to the dialysis 2.5 Standard precaution-related infection 2.5.1 Structure quality indicators • e establishment of standards for the aspect of personal hygiene, the use of protective barriers, the accident prevention • e availability of isolation room or space for the infec-tion isolation. 2.5.2 Process quality indicators • Health care workers select and use appropriate gloving for each activity. • e glove must be changed when leaked or damaged, replaced when caring for the next patient and removed after finishing the activities. • Health care workers wear gown and apron appropriately for each activity. • Health care workers must change gown immediately when leaked, soaked or damaged and remove immediately after finishing activity. • Health care workers use appropriate goggles and mask for the activities. • Health care workers wear mask covering both mouth and nose and not re-use of mask after being hung on the neck and must remove immediately after the activity. • Health care workers wear appropriate protective barriers before cleaning equipment contaminated with blood and body fluid. • Health care workers manage appropriately for the environment sanitation. • Health care workers practise appropriately to avoid accidents and needle stick by using one-hand technique, no-touch technique, or no-recap. • Health care workers isolate infected or susceptible patients appropriately by separating bed for at least 90 centimeters apart or isola-ting the patients in the isolated room or in the provided area. 2.5.3 Outcome quality indicators • Outbreak of nosocomial infections related to non-com-pliance with standard precautions. 2.6 Employee health-related infection 2.6.1 structure quality indicators

CM


213


• Having employee health program according to CDC recommendation. • Sufficient equipment and tool provision and environ-ment modification for safety workplace. • All units have manual of employee health program 2.6.2 process quality indicators • Hospitals employees get employee health program ac-cording to CDC recommendation. • Hospital employees have sufficient and effective equip-ment and tools for safety practice. • e hospitals have project and activities related to im-munization and provide benefits for employees who are infected, disabled, or dead from occupationally-acquired infection. • e ratio of health care workers who receive pre-employ-ment health screening. • e ratio of high risk of infection among health care workers who receive immunization. • Rate of health care workers who receive annual health examination. 2.6.3 Outcome • Health care workers satisfaction related to vaccine ac-quisition, working system the safety environment. • Rate of health care workers who have knowledge of employee health program. • Rate of occupational acquired infection among health care workers. • Rate of needle stick injury and rate of blood and body fluid exposure. • Rate of infected health care workers who receive con-sultation, get benefit. 2.7 Surgical Site Infection 2.7.1 Structure • Have the policy regarding the room cleaning and the management of operating room air ventilation and the traffic. 2.7.2 process • e appropriate use of protective barrier 1. Pre-operative • wear clean glove when cleaning and preparing the skin 2. Intra-operative • wear sterile gown, mask, goggles as necessary

CM


214


3. Post-operative • wear appropriate protective barrier. • e appropriate preparation of the patient. 1. Pre-operative 1) hospital stay no more than 3 days prior to the opera-tion 2) the skin is appropriately cleaned and prepared 3) the patients get SSE if they are to receive the general anesthesia and check for effectiveness. 4) the patient must quit smoking at least 30 days prior to the operation. 5) the provision of education for the patients for the practice of NIPC covering the area of pre-, intra- and post-operative periods. 2. Intra- operative 1) e skin is prepared and clean appropriately. 2) e antibiotic prophylaxis are given 1 hour prior to the incision and repeat if the operation period is longer than 4 hours. 3) e antibiotic prophylaxis are given 1-2 times only during the operation and not given after 24 hrs for general surgical operation. 4) e antibiotic prophylaxis are given within 48 hrs for the operation of cardiovascular system , joint replacement and emer-gency case. 5) Health care workers practise surgical hand washing with no brush except for the first hand washing of the day. 6) Health care workers strictly follow aseptic technique. 7) Health care workers use one-way traffic. 3. Post-operative 1) e surgical wounds are cared appropriately using either wet or dry dressing. 2) e clean wound are opened and cleaned at mini-mal or once every 2-3 day but the other type of wound may be cleaned as necessary. 3) e wound are monitored, evaluated and reported for any sign of infection and the drainage system must be maintained. 4) If transferred, the information regarding the infection must be provided to the health care workers or setting for the continuity of care. 5) ere must be the provision of the education system related to the prevention and control of infection for the patient. 2.7.3 Outcome • Incidence of infection for clean wound.

CM


215


• Rate of the patients who understand the practice of the surgical site infection prevention and control. 2.8 Waste management-related infection 2.8.1 Structure • Sufficient and appropriate waste bag and collecting equipment. 1. Using thick bag in red color and have signs indica-ting infectious waste. 2. Using thick can/box which not letting the penetration of the sharps. • Assignment of workers responsible for collecting, trans-porting and destroying infectious waste. • Availability of cart for transporting infectious wastes which is strong, rustproof, no leakage, easy to clean and having the cover. • Availability of the path specifically for infectious waste transportation. • Availability of the place specifically for collecting the infectious waste which is separated from the other buildings. • Availability of the incinerator in which the operating temperature must be at least 760o Celsius. • Availability of bathroom for health care workers after finishing their work. 2.8.2 Process • Using appropriate method of collecting infectious waste by tightening the bag when the bag is 2/3 full. • Using appropriate method of transporting infectious waste by carrying the bag which is not contacted with the body and carefully drop the bag without dragging, throwing the bag or delay of transportation. • Transport of the bag must use the assigned path and then clean the cart and the space with detergent. • When finished, health care workers must take a bath and clean their protective barrier with detergent. • Discard infectious waste appropriately: 1. Using the waste-water treatment plant for blood and body fluid. 2. Using the incinerator for needles, sharps and human bodies. 3. Using autoclave or the incinerator for cultured media 2.8.3 Outcome • Rate of number of found infectious waste items in the

CM


216


general waste bag.

DISCUSSION

The QIs were developed based on the results from Stage 1 that focused on the availability and the use of the QIs among ICNs and

the reviewed literature. e QIs for each aspect of the NI were tested for content validity rated by experts which ranged from 0.82-0.95, showing a high degree of agreement among the experts. e high agreement among same experts may be that the proposed indicators are all important and during this time, the quality assurance system for the hospital accredita-tion is emphasized and being implemented. All individuals working for the health care facilities are informed and realize that these indicators are important for the quality of the nosocomial infection prevention and control and therefore should be included as QIs items. Based on the results, all items could be used as QIs for NI due to the high agreement among experts and were tested for the feasibility by requesting ICNs and ICWNs to rate whether the items could possibility be used. e results showed that most of them agreed that most items in all aspect of the NI were possible for use. Based on the results from the consensus, the majority of experts and subjects agreed with most of items. us, these items could be applied directly and immediately to measure the quality of NI prevention and control.

CONCLUSIONS

The quality indicators developed from this study for nosocomial infec-tion prevention and control were divided into two groups; general

quality indicators and specific indicators. e specific indicators were grouped into 8 aspects of NI and each aspect was categorized into struc-ture, process and outcome. Most of items were rated by the majority of experts as highly valid. In terms of feasibility of use, the majority of ICNs and ICWNs rated most of these items as possible to be used as indicators. erefore, most of the QIs could be used to measure the quality of the nosocomial infection prevention and control in the context of ai health care delivery system.

RECOMMENDATIONS

Based upon the findings, conclusions and implication, the following recommendations are made:

CM


217


Health care worker especially infection control nurses should try out in their settings to test for the actual use of the indicators and develop data collection process using a systematic approach. For further research, it is necessary to test the use of quality indica-tors in actual practice.

ACKNOWLEDGEMENTS

We would like to express our sincere gratitude to the World Health Organization, ailand for a grant to support this project. e

researchers appreciate all experts, stakeholders and infection control nurses for giving us their valuable time and for participating in this study. anks are also extended to master’s degree students who partially help in literature review and data collection and to all staff for their assistance.

REFERENCESBolyard, E.A. 1998. Guideline for infection control in health care person-

nel. American Journal of Infection Control, 26, 289-354.Boparai, C. 1997. Hospital waste management among Srisangwan Hos-

pital workers. Master’s thesis: Faculty of Nursing, Chiang Mai University.

Calfee, D.P., and B.M. Farr. 2002. Infection control and cost control in the era of managed care. Infection Control and Hospital Epidemiol-ogy, 23, 407-410.

Danchaivichitra, S. 2001. Nosocomial infection(3rd ed). Siriraj Hospital: Bangkok.

Fonseca, S.S., M. Kunzle, S.B. Silva, J.G. Schmidt, and R.R. Mele. 1999. Cost reduction with successful implementation of an antibiotic pro-phylaxis program in a private hospital in Ribeirao, Preto, Brazil. Infection Control and Hospital Epidemiology, 20,77-79.

Huskins, W. C., B.M. Soule, C. O’Boye, l. Gulasci, E.J. Rourke, and D.A. Goldman. 1998. Hospital infection prevention and control: A model for improving the quality hospital care in low-and middle-income countries. Infection Control and Hospital Epidemiology, 19(2), 125-135.

Kamdee, P., W. Moongtui, and J. Chitreecheur. 2002. Employee health program for health care workers under ministry of public health. Oral presentation at the Annual Conference of Nosocomial Infec-tion Control Group of ailand, Krungsri river hotel, Ayuthaya, July 17-21, 2002.

CM


218


Katz, J., and E. Green. 1992. Managing quality: A guide to monitoring and evaluating nursing services. St. Louis: Mosby.

Kaye, J., V. Ashlina, D. Erickson, K. Zeiler, D. Garison, L. Gannon, P. Wyne, J. Cooper, W. Kittle, K. Sharma, and J. Morton. 2000. Criti-cal care bug team; A multidisciplinary team approach to reducing ventilator associated pneumonia. American Journal of Infection Control, 28(2), 197-201.

Maki, D.J., and P.A. Tambyah. 2001. Engineering out the risk of infec-tion with urinary catheters. Available via www: http://www.cdc.gov/ncidod/eid/Vol7no2/maki.htm

Mayhall, C.G. 1993. Surgical infections including burns. In Wenzel, R.P, ed. Prevention and control of nosocomial infections (2nd ed.). Baltimore, William & Wilkins, 281-286.

Moongtui, W., D. K. Gauthier, and J.G. Turner. 2000. Using peer feedback to improve handwashing and glove usage among ai health care workers. American Journal of Infection Control, 28, 365-9.

National Center for Infectious Diseases. 2001. National nosocomial infec-tious surveillance system report, data summary from January 1992- June 2001, American Journal of Infection Control, 29, 404-421.

NNIS. 1998. National Nosocomial Infections Surveillance (NNIS) System report, data summary from October 1986-April 1998. American Journal of Infection Control; 26:522-533.

Neilson, J., H.J. Kolmos, and V.T. Rosdahl. 1998. Poor value of surveil-lance cultures for prediction of septicemia caused by coagulase-nega-tive staphylococci in patients undergoing hemodialysis with central venous catheters. Scandinavian Journal of Infectious Disease, 30, 569-572.

Nilavok, N. 1997. Effect of multifacet ventilator associated pneumonia control program on the rate of ventilator associated pneumonia among patient depending on the ventilator. Master thesis; Faculty of Nursing, Chiang Mai university.

Orrett, F.A., P.J. Brooks, and E.G. Richardson. 1998. Nosocomial infec-tions in a rural regional hospital in a developing country: Infection rates by sites, services, cost, and infection control practices. Infection Control and Hospital Epidemiology, 19; 136-140.

Piamsuwan, W., S. Kongprasit, L. Pangnukraw, and S.Trakulsomboon. 2001. Efficacy of the use of 2% glutaraldehyde in the ventilator circuit. Bulletin of Nosocomial Control Group of ailand, 11(1), 43-54.

Sax, H., S. Hugonnet, S. Harbarth, P. Herrauth, and D. Pittet. 2001. Variation in nosocomial infection prevalence according to patient care setting: a hospital wide survey. Journal of Hospital Infection,

CM


219


48(1), 27-32.Srisuphan, W., W. Senaratana, W. Pichiansathien, and T.Tongsawas, 1989.

Nosocomial infections in Maharaj Nakorn Chiang Mai Hospital. Journal of Medical Association of ailand. Suppl 2: 7-11.

Wenzel, R.P. 1997. Prevention and control of nosocomial infection. (3rd ed.). Philadelphia: Williams & Wilkins.

Wildmer, A.F. 1997. Management of catheter-related bacteremia and fungemia in patients on total parenteral nutrition. Nutrition, 13(4), sppl 1, 18s-25s.

Wongkrea, P. 1997. Effect of education on the knowledge and prac-tices of waste management of hospital workers in Maharaj Nakorn Chiang Mai Hospital. Master’s thesis: Faculty of Nursing, Chiang Mai University.

CM


221


Chiang Mai University Journal of Social Sciences and Humanities

Instructions for Preparation of Manuscripts

Types of Articles and Formats Manuscripts submitted to Chiang Mai University Journal of Social Sciences and Humanities are critically reviewed before they can be published. e reviewer has the major responsibility for critical evaluation to determine whether or not a manuscript meets standards of social sciences and literary quality. Reviews must be based on manuscripts of satisfactory literary quality. It is not the task of a reviewer to edit poorly prepared papers or to correct readily detectable errors. In order to avoid any embarrassment to authors, any undue burden for reviewers or editors, or any loss of time and effort, every paper should have a thorough preliminary review by competent colleagues of the author before it is sent to the journal.

Submission of Manuscripts Manuscript should be submitted to: e Editor-in-Chief Chiang Mai University Journal Office of the President, Chiang Mai University Chiang Mai 50200, ailand E-mail: [email protected] E-mail: [email protected]

e typed original and three copies of the paper and artworks should be submitted. For the initial submission of manuscript for consideration, hardcopies are sufficient. For the processing of accepted papers, electronic versions are preferred. Once a manuscript has been accepted, a final version on a computer diskette should be submitted. e files should be prepared preferably using the latest version of either Microsoft Word or Word Perfect. e Word processing software should be specified.

Preparation of Manuscripts Manuscript must be written in English and should be typewritten, typed on one side of the paper, with wide margins and double spacing throughout. Every page of the manuscript, including the title page, tables, figures, references etc. should be numbered.

CM


222


Manuscript should be organized in the following order: 1. Title e author should design his title to supply enough infor-

mation for the potential reader to make a reliable decision as to whether the paper is of probable interest.

2. Name(s) of author(s), complete postal address(es) of affiliations and E-mail of the corresponding author.

3. Abstract and Key words e abstract should be a suitable literary adjunct to the

social sciences report and it should meet the needs of the literary researcher or information specialist. e length of the abstract should not exceed 250 words for full-length papers and 75 words for notes and brief articles. Key words for indexing should be listed at the end of the abstract.

4. Introduction Introductions should be kept short. Good introductions

should include: (i) brief statement of the problem that justifies doing the work or of the hypothesis on which it is based; (ii) the findings of others that will be challenged or developed; and (iii) an explanation of the general approach and objectives. e aim of the introduction should be to excite and interest the reader.

5. Materials and Methods is section contains details about materials, techniques,

experimental design, and environment. Sufficient detail should be provided to permit the reader to repeat the experiments. e methods section may be arranged in a chronological pattern, suc-cession of techniques, or other manners which will most effectively assist the reader in studying the paper.

6. Results Use tables, graphs, diagrams, and photographs to provide a

clear understanding of the results. Data included in illustrations and tables should not be discussed extensively in the text, but significant findings should be pointed out. Show how the objec-tives have been achieved. e results should be connected to one another. Some times this causes the results section to be combined with the discussion section.

7. Discussion and Conclusion In the discussion section the author assesses the meaning of

the results. Show how the results provide a solution to the problem stated in the introduction or given as the objective. Connect the work of this study with previous works showing how and why they differ or agree. Point out the significance and implications

CM


223


of the work and indicate possible future developments. Do not give excuses for unexpected results and failures of experiment. Controversial issues should be discussed clearly and fairly. Where results differ from previous results, they should be explained.

Some papers have a conclusion section. is includes any significant conclusions that have been drawn from the work. ese should be carefully worded so there is no misunderstanding on the part of the reader. It is often desirable to present conclusions as part of the discussion section; however, in a paper that is long and complex, it may be helpful to summarize conclusions in a separate section.

8. Reference Citations e form used for giving the reference in the text will vary

according to the construction of the sentence in which it occurs, e.g., Bell (1999) or (Bell, 1999). when there are two authors, name both of them, e.g., Heimann and Willmann (1998) or (Heimann and Willmann, 1998). when there are three or more authors, cite their paper in the form Hildebrandt et al., (1999) or (Hildebrandt et al., 1999). If two or more articles by the same author or authors in the same year are cited, they should be designated as follow: Pandey et al., (1984a, 1984b, 1984c).

9. Reference lists All citations, whether to published literature or to unpublished

work are to be listed alphabetically by surname of senior author at the end of the manuscript. Each reference to a periodical pub-lication must include, in order, the name(s) of the author(s), the year of publication, the full title of the article, the publication in which it appears, and the volume and inclusive page numbers. e reference lists are based on the CBE Style Manual published by the American Institute of Biological Sciences for the Council of Biology Editors(CBE). References must be arranged as follow:

Journal article Halmilton, M.B., E.L. Pincus, A.D. Fion, and R.C. Fleischer.

1999. Universal linker and ligation procedures for con-struction of genomic DNA libraries enriched for micro-satellites. Biotechniques 27: 500-507.

CM


224


Book Sokal, R.R., and F.J. Rohlf. 1995. Biometry: e principles

and practice of statistics in biological research. W.H. Freeman and Co, New York.

Chapter in book Jackson, M.B. 1982. Ethylene as a growth promoting

hormone under flooded conditions. p.291-301. In P.F. Wareing (ed) Plant growth substance. Academic Press, London.

Edited proceeding, symposia etc. Pratt, A., R.J. Gilkes, S.C. Ward, and D.A. Jasper. 2000.

Variations in the properties of regolith materials affect the performance of tree growth in rehabilitated baux-ite mine-pits in the Darling Range, SW-Australia. p.87-88. In A. Brion and R.W. Bell (eds) Proceed-ing of Remade Land 2000, the International Confer-ence on Remediation and Management of Degraded Lands. Fremantle, 30 Nov-2 Dec 2000. Promaco Conventions, Canning Bridge.

Dissertation Senthong, C. 1979. Growth analysis in several peanut

cultivars and the effect of peanut root-knot nematode (Meloidogyne arenaria) on peanut yields. Ph.D. Dis-sertation. University of Florida, Gainesville, Florida, USA.

10. Acknowledgements Any acknowledgements should be typed as text and placed

before the references. e word ACKNOWLEDGEMENTS should be capitalized and centered above any citation.

INDEX TO VOLUME 1 NUMBER 1 (2007)CHIANG MAI UNIVERSITY JOURNAL

OF SOCIAL SCIENCES AND HUMANITIES

AUTHOR INDEX

Ahmed, K.S. 1(1) : 35Awipan, R. 1(1) : 177Bhuiyan, M.N.A. 1(1) : 35Chaitha, W. 1(1) : 177Chiaravutthi, Y. 1(1) : 49Chittreecheur, J. 1(1) : 197Cowie, B. 1(1) : 119Forret, M. 1(1) : 137Hofe, R.V. 1(1) : 69Jahan, R. 1(1) : 35Karnjanarat, P. 1(1) : 177Kittisopee, T. 1(1) : 177Kunaviktikul, W. 1(1) : 197Leksawasdi, N. 1(1) : 197Lissitsa, A. 1(1) : 1Mokhtar, S. 1(1) : 165Moongtui, W. 1(1) : 197Moreland, J. 1(1) : 137Ow, S.H. 1(1) : 165Pattamarakha, K. 1(1) : 21Permsuwan, U. 1(1) : 177Picheansathian, W. 1(1) : 197Promkatkeaw, T. 1(1) : 137Puengpang, N. 1(1) : 119Roadrangka, V. 1(1) : 119Rungsuriyawiboon, S. 1(1) : 1Saithanoo, S. 1(1) : 21Sawatdirakpong, S. 1(1) : 109Senaratana, W. 1(1) : 197Soparat, P. 1(1) : 197Srinoy, B. 1(1) : 21Srisuphan, W. 1(1) : 197Viboonsunti, C. 1(1) : 177Xianming, X. 1(1) : 93Yimyam, S. 1(1) : 197Zainuddin, R. 1(1) : 165

ISSN 1905-906XISSN 1905-906X

CMU. Journal of Social Sciences and Humanities website: http://cmuj.chiangmai.ac.th

Ch

ian

g M

ai U

niv

er

sit

y J

ou

rn

al o

f So

cia

l Sc

ien

ce

s a

nd

Hu

ma

nit

ies

Chiang Mai University JournalResearch Administration Center239 Huay Kaew Road, T. Suthep, A. Muang, Chiang Mai 50200 THAILAND.Tel: 66-53-942676-7 Facsimile: 66-53-943600E-mail : [email protected] [email protected]

CONTENTS





Vol.1, June 2007, Num

ber 1

Publ

ishe

d by

Chi

ang

Mai

Uni

vers

ity, C

hian

g M

ai,

Thai

land

CONTENTS Page

TRANSITION ECONOMY How Far Are the Transition Countries from the Economic Standards in the European Union? 1 Supawat Rungsuriyawiboon and Alexei Lissitsa

Leadership in Dairy-Cattle Production in Southern Thailand : Dimensions in Socio-Economic Status, Adoption of Recommended Practices, and Communications 21 Kriengsak Pattamarakha, Boonlert Srinoy and Somkiat Saithanoo

Projection of GDP of Manufacturing Industries in Bangladesh: An Approach to Growth Model 35 M. N. A. Bhuiyan, Kazi Saleh Ahmed and Roushan Jahan

The Existence and Rationale of Predatory Pricing Strategy in the Network Industry: An Experimental Approach 49 Yingyot Chiaravutthi

Understanding Housing as an Economic Stimulus: A Regional Computable General Equilibrium (CGE) Model for New York State Housing 69 Rainer vom Hofe


A Study of First-Year College Students’ Metacognitive Awareness of Reading Strategies 93 Xu Xianming

Analysis of Dimensionality Underlying EFL Proficiency of Thai Students 109 Sumalee Sawatdirakpong

Development of an Instructional Set for a Laboratory in Biology Course for First-Year Thai Science Student Teachers 119 Nantarat Puengpang, Vantipa Roadrangka and Bronwen Cowie

Page


A Case Study of a Thai In-service Primary Teacher’s Understanding of the Nature of Science 137 Tepkanya Promkatkeaw, Michael Forret and Judy Moreland

A Review on the Teaching and Learning Resources for the Deaf Community in Malaysia 165 Siew Hock Ow, Salimah Mokhtar and Roziati Zainuddin

COMMUNITY HEALTH

Relationships of Service Quality, Patient Satisfaction and Patient Behavioral Intention of Pharmacy Service of Highly-Active Antiretroviral Therapy in Community Hospitals in Chiang Rai Province, Thailand 177 Wacharin Chaitha, Chupen Viboonsunti, Tanattha Kittisopee,

Unchalee Permsuwan, Ratanaporn Awipan and Penkarn Karnjanarat

Development of Quality Indicators on Important Aspects of Nosocomial Infection Prevention and Control 197 Wilawan Senaratana, Wanchai Moongtui, Wichit Srisuphan,

Nunta Leksawasdi, Poonsap Soparat, Susanha Yimyam,

Jittaporn Chittreecheur, Wilawan Picheansathian

and Wipada Kunaviktikul

chiang mai university journal of social sciences...

Documents