social pertanika journal of science a humanities papers/jssh vol. 2 (1) mar. 1994... · issn: 0128...

ISSN: 0128 - 7702

P e r t a n i k a J o u r n a l o f

socialscience AHumanities

VOLUME 2 NO. 1 MARCH 19 9 4

A s c i e n t i f i c j o u r n a l p u b l i s h e d by U n i v e r s i t i P e r t a n i a n M a l a y s i a P r e s s

Pertanika Journal o f Social Science and Humanities■ About the Journal

Pertanika, the pioneer journal of UPM, began publication in 1978. Since then, it has established itself as one of the leading multidisciplinary journals in the tropics. In 1992, a decision was made to streamline Pertanika into three journals to meet the need for specialised journals in areas of study aligned with the strengths of the university. These are (i) Pertanika Journal of Tropical Agricultural Science, (Ii) Pertanika Journal of Science and Technology and (iii) Pertanika Journal of Social Science and Humanities.

■ Aims and ScopePertanika Journal of Social Science and Humanities aims to develop as a flagship journal for the Social Sciences with a focus on emerging issues pertaining to the social and behavioural sciences as well as the humanities, particularly in the Asia Pacific region. It will be published twice a year in March and September.The objective of the journal is to promote advancements in the fields of anthropology, business studies, communications, economics, education, extension studies, psychology, sociology and the humanities. Previously unpublished, original, theoretical or empirical papers, analytical reviews, book re

EDITORIAL BOARD \Prof. Dr. Sulaiman Hj. M ohd Yassin - Chief Editor Centre for Instructional ServicesAssoc. Prof. Dr. Zainal Abidin M ohamed Faculty of Economics and ManagementAssoc. Prof. Dr. M ohd Ghazali Mohayidin Faculty of Economics and ManagementAssoc. Prof. Dr. Roslan bin A GhafFar Faculty of Economics and ManagementAssoc. Prof. Dr. Abdul Halin Hamid Faculty of Human EcologyAssoc. Prof. Dr. Khor Geok Lin Faculty of Human EcologyAssoc. Prof. Dr. Abdul Rahman Md Aroff Faculty of Educational StudiesSumangala Pillai - Secretary Universiti Pertanian Malaysia Press

Published by University Pertanian Malaysia PressISSN No: 0128-7702Printed by: Am pang Press Sdn. Bhd.

views and readers’ critical reactions may be submitted for consideration. Articles may be in English or Bahasa Melayu.

■ Submission of ManuscriptThree complete clear copies of the manuscript are to be directed to

The Chief EditorPertanika Journal of Social Science and HumanitiesUniversiti Pertanian Malaysia43400 UPM Serdang, Selangor Darul EhsanMALAYSIATel: 9486101 Ext: 1325; Fax (603) 9483745

■ Proofs and OffprintsPage proofs, illustration proofs, the copy-edited manuscript and an offprint order form will be sent to the author. Proofs must be checked very carefully within the specified time as they will not be proofread by the Press editors.Authors will receive 20 offprints of each article. Additional copies can be ordered from the Secretary of the Editorial Board by filling out the offprint order form.

^NTERN^IONALPANELM EM B^^Prof. Jean Louis FloriotInternational Graduate Institute of AgribusinessProf. Bina Agarwal Institute of Economic Growth University Enclave, IndiaProf. V.T. King University of HullProf. Royal D. Colie Cornell University, IthacaProf. Dr. L inda J. Nelson Michigan State UniversityProf. Dr. Yoshiro H atano Tokyo Gakugei UniversityProf. Dr. Max Langham University of FloridaDr. M oham ed Ariff National University of SingaporeProf. Fred Luthens University of NebraskaProf. D.H. Ritchie University of ToledoProf. Gavin W. Jones Australian National UniversityProf. Dr. Lehm an B. F letcher Iowa State UniversityProf. Ranee P.L. Lee Chinese University, Hong KongProf. S tephen H.K. Yeh University of Hawaii at ManoaProf. Graham W. T hurgood California State University

ARCHIVE COPY(Please Do Not Remove)

p e r ta n ik a editorial o f f i c eResea, ch Management Centre (RMC)

, .n il5it FI°or, IDEA Tower II UPM-MTDC, Technology Centre

U™ efs't' Putfa Malaysia 43400 Serdang, Sfi!«noo(, Maiai/fti*

! *603 0647 1822, 8S4^1M W 7 Wfl

Pertanika Journal of Social Science & Humanities Volume 2 No. 1 March 1994

Contents

Projection of Sectoral Value-added: Comparative Analysis of Alternative Methods - 1Zakariah Abdul Rashid and Sabaruddin Ahmah Khair

Learning English in a non- supportive Environment among Malay Learners in 1 1Secondary Schools - Jamali Ismail

The Relationship between Quality of Home Environment and Mental Scores of 21Children Attending the UPM Laboratory Preschool - Rohani Abdullah, Siti Nor Yaacob and Rozumah Baharudin

Factors Associated with N on-adoption o f Technology by R ubber Sm allholders - 29Rahim M. Sail and Mazanah Muhamad

The Stability and Predictability of Betas: Evidence from the Kuala Lumpur Stock 4 3Exchange - Shamsher Mohamad and Annuar Md. Nassir

Determinants of Foreign Direct Investment in the Malaysian Manufacturing Sector - 53Zulkamain Yusop and Roslan A. Ghaffar

Kajian Kesesuaian dan Keberkesanan Raven SPM untuk Mengenal Pasti 63Kanak-kanak Melayu yang Pintar Cerdas - Abd Majid Mohd Isa

Malaysian English: Exploring the Possibility o f S tandardization - Rosli Talif and 69TingSu Hie

PertanikaJ. Soc. Sci. & Hum. 2(1): 1-10 (1994) ISSN: 0128-7702 © Universiti Pertanian Malaysia Press

Projection of Sectoral Value-added: Comparative Analysis o f Alternative Methods

ZAKARIAH ABDUL RASHID and SABARUDDIN AHMAH KHAIR Economics D epartm ent,

Universiti Pertanian Malaysia,43400 UPM Serdang, Selangor Darul Ehsan, Malaysia.

Keywords: value-added projections, input-output analysisABSTRAK

Kertas kajian ini membincangkan kesesuaian kaedah penganggaran nilai ditambah ekonomi Malaysia. Tiga kaedah penganggaran, kaedah blow-up KDNK, kaedah blow-up permintaan akhir dan kaedah input-output, dibandingkan di segi ketepatan dan kecekapan relatif. Perbincangan tertumpu kepada tiga perkara. Pertama, perbincangan tentang metodologi, manakala perkara kedua dan ketiga, masing-masing tentang ketepatan dengan menguji kesilapan minnya dan kecekapan dengan membanding variannya. Keputusan kajian menunjukkan bahawa kaedah input-output lebih baik di segi ketepatan (kesilapan min terendah) dan kecekapannya (varian terendah).

ABSTRACT

The paper discusses validation of methods used in projecting Malaysian value-added. Three methods of projection, the GDP blow-up, the final demand blow-up and the input-output methods are compared in terms of their accuracy and relative efficiency. The discussion focuses on three aspects. The first aspect deals with the methodological issues, while the second and third aspects, respectively, determine the accuracy by testing their mean errors and efficiency by comparing their variances. The results show that the input-output method is superior to the other two methods in terms of accuracy (lowest mean error) and efficiency (lowest variance).

INTRODUCTIONAlmost all developing countries in the world rely greatly on economic planning to influence and control the level and growth of their principal economic variables to achieve a predeterm ined set of development objectives. The planning process itself involves preparation of an economic plan which contains a specific set of quantitative economic targets to be reached in a given time period. These targets are forecast or projected values of certain principal economic variables expected in the future, which even though unknown at the time of making the forecast, are nevertheless, im portant in deciding the future path of the economy At the macro level, the forecast values which comprise Gross Domestic Product (GDP), employment, consumption (Private and Public), Investment, exports and imports are derived from a set of forecast values at the sectoral level.

Value-added forecast enables planners to chart the future statistical time path of the economy’s key variables; particularly, if it is done at a sectoral level, it will provide a quantitative indication of rising and declining sectors. Industrialisation will see an increasing share of manufacturing and a declining share of agricultural value- added. An economy undergoing rapid transformation will generally experience a marked shift in the share of sectoral added value. Sectoral value- added forecast has been widely used in both developed countries such as the U nited States (Leontief, 1986), the United Kingdom (Gosh, 1964) and Netherlands (Theil, 1975) as well as in developing countries such as India (Slastry et a l 1975) and Malaysia (Aziz, 1984; and MIDA, 1985)x. The preparation of Malaysia’s Industrial Master Plan, for instance, involves a projection of twelve, key economic variables of the sector and uses an

1 The Central Planning Bureau of Netherlands provides annual Central Economic Planning which is essentially a multiple forecast of the following year’s economic conditions. The Bureau uses the input-output techniques extensively in its forecasting of all the household, business, government as well as the international sectors. Theil (1975), pp. 44-77.

Zakariah Abdul Rashid and Sabaruddin Ahmah Khair

input-output technique; the forecast values become the im portant motivating factor that leads to the excellent performance of the plan as reported by the review of the plan.

The aim of making a forecast is to reduce uncertainty, which is easily verified by comparing the forecast values with the actual values. A forecast value which does not differ significantly from its actual value is described as an efficient forecast. In this paper, a set of projections of Malaysian sectoral value-added is made by using different methods, including that of input-output; and their forecast performance is validated against the actual values. Section 2 discusses the assumptions of input-output analysis and structure of the Malaysian input-output table. Section 3 presents the methodologies of value-added forecasting. The results and discussion are presented in Section 4, followed by a summary and conclusions of the paper in Section 5.

The results of our analysis show that the Malaysian input-output tables are consistent with those of other countries in terms of their predictive ability; and thus the tables can be used confidently as a basis of economic projection and planning.

Input-output tables are an accounting framework for assembling data on industry input and output, which reveal the many interrelated transactions occurring in the economy. The constructed tables closely reflect the national accounts which show the flow of final goods currently produced and the equivalent flow of factor and nonfactor income generated in the production of the output. National product, which equals the sum of the value-added in all individual industries, when broken down into contributions made to it by individual industries, is generally referred to as national product by industry of origin.

In input-output tables, the value of final goods and services produced by the business, personal, government and foreign sectors, as described by the national accounts, is found in the value-added row(s); when data are available the row(s) can be reduced to wages, depreciation, indirect taxes, and

residual profits. From the table, value-added row(s) will show what value each separate industry adds to the final output; the value-added by each industry to the raw materials, or other goods and services that it bought from other industries is accumulative at each stage in the chain of production.

ASSUMPTIONS AND STRUCTURES OF INPUT-OUTPUT TABLES

T here are im portant practical and theoretical objections to the use of input-output coefficients matrices in forecasting the impact of changes in one sector on gross output and thus on added value in the economy as a whole. In the simplest possible case where a single set of input-output coefficients from a previous year are used to forecast added value, limiting assumptions are required. In oudine, these are: (i) each unit of commodity output is perfectly substitutable for every other unit of output produced; (ii) the production function in each sector is linearly hom ogeneous of degree one; in other words, there are constant returns to scale in all types of commodity p ro d u c tio n ; (iii) th e re is p e r fe c t complementarity and zero subsitutability between commodity inputs to the production of each type of commodity, so that a change in the level of output requires changes in fixed proportion of the inputs consumed; (iv) prices are fixed (these are normally expressed in terms o f ‘Leontief units’ such that the price of each commodity is equal to unity); and (v) supply is perfecdy elastic. If these assumptions hold in the short run, the model has sufficient reality to provide an efficient added value forecast.

In the Malaysian 1983 input-output tables published by the Departm ent of Statistics2, value- added is shown as the difference between the total value of primary inputs, (the primary inpu t’s quadrant being not well developed as yet) and total imports and commodity taxes. Value-added will therefore include salary and wages, interest on capital, profits and depreciation charges. This definition is in line with that of another publica-

2 The first input-output table of West Malaysian (Peninsular Malaysia only) economy was compiled by the Department of Statistics for the year 1960. Following this was the publication of input-output tables for the years 1965 and 1970. The first set of input-output tables for the whole of Malaysia were published by the same departm ent for the year 1978, followed by that for 1983 which is the latest published.

2 Pertanika J. Soc. Sci. 8c Hum. Vol. 2 No. 1 1994

Projection of Sectoral Value-Added: Comparative Analysis of Alternative Methods

tion of the Department, the Industrial Survey3, which clearly defines value-added as the difference between the gross value of output and the cost of inputs.

Why are value-added projections im portant in the economic planning of a country ? The sectoral value-added forecast is im portant not only because it enables us to know the expected shifts in the economy’s production structure which undergoes a process of transformation but also to quantify the anticipated sectoral value-added in the target year. The trend and magnitude of the shifts provide invaluable information for planning purposes; to minimise bottle-necks in a particular industry caused by the inability of its up-stream industries to supply the necessary intermediate inputs, and which consequently create another bottle-neck in its down-stream industries. The forecast is also im portant because it can identify bo th growing and shrinking industries, and thereby help prepare the economy for the antici

pated organisational and administrative changes caused bv structural changes.

GDP by industrial origin which is the total sum of sectoral value-added, will show the relative contribution of each sector to the economy’s final output. When time-series data of the sectoral value added are available, it will show structural changes in the econom y’s final output. Table 1 shows Malaysian GDP by industrial origin, typfying a developing country striving for industrialization, where the agriculture sector is on the decline and m anufacturing sectors are on the increase. The input-output tables will not only show value-added originating from each industry but also the share of each industry’s value-added to its gross value of output. This information can be found in the economy’s input structure as shown in Table 2. The table shows that despite the declining trends in the contributions from the agriculture and mining industries to the GDP, their proportions of value-added from total gross value of outputs remain the largest.

TABLE 1 Malaysian GDP by industrial originPercentage share in GDP Percentage growth

1979 1984 1987 1990 1979-84 1985-90Agriculture 24.3 20.1 21.7 18.9 2.9 4.3Mining 11.1 10.5 10.5 9.7 5.8 4.0Manufacturing 19.3 20.3 22.6 26.9 7.9 10.6Others 45.3 49.1 45.2 44.5 9.7 3.4

GDP 100 100 100 oo1-H

Source. Ministry of Finance, Economic Report, various issues.

TABLE 2Input structure of the Malaysian Economy 1978 and 1983 (%)

Agriculture Mining Manufacturing Others1978 1983 1978 1983 1978 1983 1978 1983

Agriculture 1.03 3.43 0.14 0.03 11.72 10.36 0.70 0.47Mining 0.02 0.03 0.27 0.52 4.97 4.70 1.12 0.85Manufactg. 8.56 10.43 4.93 3.96 15.60 18.35 17.02 16.44Others 3.25 4.91 8.61 8.83 9.71 10.11 15.55 16.05Tot. intmd. 12.83 18.80 13.95 13.34 42.00 43.52 34.39 33.81Comm, taxes 0.62 0.55 0.70 0.27 4.00 3.40 1.55 1.31Imports 6.24 6.16 5.21 6.15 29.80 29.82 12.86 17.24Value added 80.28 74.49 80.15 80.24 24.20 23.26 50.66 47.64Total 100 100 100 100 100 100 100 100Source. Department of Statistics, Input-output tables, 1978 and 1983.

3 Industrial Survey (1987), p. xii

Pertanika J. Soc. Sci. 8c Hum. Vol. 2 No. 1 1994


METHODOLOGYThe projections of 1983 sectoral value added of the Malaysian economy were compiled by using the 1978 input-output tables as the base. Information known at the time of making the projections is derived from the 1978 and 1983 input- output tables published by the Department of Statistics. Three methods of projection are used. The simplest and most straightforward m ethod is by blowing-up the base year’s sectoral value added in proportion to the growth of aggregate GDP of the economy during the 1978-83 period. This method of partial forecasting is similar to fitting a mathematical curve to an individual time series and extrapolating this curve to some future date. If aggregate GDP records a rise of 10 percent during the period, the present m ethod will blow-up estimates of sectoral value-added by adding 10 percent to those of the base year.

The second m ethod assumes that the base year’s sectoral proportion of final dem and to total output remains constant over the period of projection. Applying these proportions on the vector of final demand of the year of projection, the estimates of sectoral output can be obtained. Similarly, the m ethod assumes invariant value- added coefficients of the base year. The estimates of sectoral value-added for the projected year are obtained by multiplying the fixed value-added coefficients by the initial estimates of sectoral total output of the projected year. That is:

The subscripts 0 and 1 refer to the base and projected years respectively.

It is im portant to note that the first m ethod takes no account of final dem and breakdown in the projected year and considers only a change in aggregate GDP. The second method, on the other hand, takes into account the changes in final dem and over the period. In the above methods, the interdependence between the different sectors of the economy has been ignored.

The third m ethod of forecasting, the input- output m ethod which is also considered as consistent forecasting, ensures that the output of each industry is consistent with both final and interm ediate demands. This means that the projection for individual industries will add up to a total projection (of Gross National Product, for example) if the structural relations of the economy remain the same over the projection period, or if allowance can be made for the anticipated changes in the structural relations.4

The balance equation of the static input-output model can be written as:

X. = a nX i a12Xg + a, X + F.In n 1

(l)• • F,

(4) X2 - a2|X ( + a^X^H- ...+ a2nXn + F2 X = a ,X, + a ,X , + ...+ a X + Fn nl 1 n2 2 iin n n

For idl industry, the above relationship can be expressed as:

(5) Xj = X V k + F, i = 1, n.

(2) and V, • X,

where a„ = XVX. is the am ount of output of industry i purchased by industry j per unit of output of industry j. O r equation (5) can simply be written as:

(3)

where X F

and V

sectoral total output sectoral final dem and sectoral value added

(6) X = AX + F

Where X and F are vectors of total output and final dem and respectively and A is the structural matrix. By assuming exogenous final dem and, the general solution of the balance equation can be shown as:

1 One of the major problems involved in consistent forecasting is that of allowing for changes in the structural matrix, particularly when long term projections are being made. For short term forecasts, it is fairly safe to assume that the matrix remains the same or it does not change significantly. For long term projections, however, one normally does not assume the matrix to remain the same.

4 Pertanika J. Soc. Sci. & Hum. Vol. 2 No. 1 1994


(7) X = (I - A)_1F

While such a result would be tautological for the year for which the table is constructed, it would be substantive when the above com putation is done for some other specified period.

By inserting projected year’s final dem and as an exogenous elem ent in equation (7), given the base year’s structural matrix, sectoral output for the projected year can be derived. And by assuming constant value-added ratio, projected sectoral value-added is derivable from the pro jected sectoral output. The estimation of projected value- added may be summarised by the following relationship:

(8) V.1 = k°Xnb.. F l v 7 1 1 j y j

where V, k and b are the am ount of value added, the value-added ratio and the inter-industry coefficient (elem ents of L eontief inverse m atrix) respectively. The subscripts refer to industry while the superscripts refer to the base and projected years.

It is clear from equation (8) that estimates of value-added of sector i in year t is given by a linear combination of final dem and deliveries in year t. The inter-industry coefficient is derived from technological relationship in year 0. In other words, the sectoral estimates of value added depends not only on the value-added to total output ratio but is also influenced by parameters expressing direct and indirect requirem ents of output to satisfy a unit worth of final dem and of different industries.

RESULTS AND DISCUSSIONThis section discusses the results of our analysis and presents the various estimates of value-added in 1983 by using the above three methods and comparing their relative efficiencies in projecting the economy’s value-added. More specifically, the efficacy of each method of projection is examined in terms of the differences between their respective estimates and the actual values, or their errors. The best m ethod would be the one that gives the smallest error. Table 3 compares the various estimates of value-added in 1983 com puted by

using the above three m ethods with the actual value-added published by the D epartm ent of Statistics. The sectoral value-added forecast for 1983 is obtained by multiplying the sectoral value-added for 1978 by the GDP average growth rate for the 1978-83 period, which is calculated to be 3.47 percent. A scrutiny of the table reveals that a particular m ethod may provide the best estimate of value-added for one particular sector bu t may not for other sectors.

Invariably the input-output m ethod of projecting the value added seems to provide the best estimate for other agriculture sector (Sector no.1) in comparison to the final demand blow-up and GDP blow-up methods. The o ther agriculture sector’s actual value added amounted to RM2268.4 million; estimates of value added by the input-output m ethod is the closest to the actual, am ounting to RM2260.4 million (about 0.35 percent error); while those of final dem and blow-up and GDP blow-up methods am ounted to RM2450.2 million (8.01 percent error) and RM4538.8 million (100.08 percent error) respectively.

In the case of the dairy products sector (Sector no. 8), the final dem and blow-up m ethod appears to be the best m ethod followed by input- ou tpu t and GDP blow-up m ethods. The dairy products sector’s actual value added am ounted to RM 180.9 million; estimates of value added by final dem and blow-up m ethod is the closest to the actual, am ounting to RM 187.8 million (about 3.81 percent error); while those of input-output and GDP blow-up methods am ounted to RM 167.3 million (7.51 percent error) and RM766.7 million (323.8 percent error) respectively.

Similarly, the ordering of best method differs for the tobacco sector (Sector no. 16), where the GDP blow-up m ethod ranked first, followed by the input-output and final dem and blow-up methods. The tobacco sector’s actual value-added amounted to RM339.2 million; estimates of value added by GDP blow-up m ethod is the closest to the actual, am ounting to RM358.7 million (about 5.74 percent erro r); while those of input-output and final dem and blow-up methods am ounted to RM220.3 million (35.05 percent error) and RM219.1 million (35.40 percent error) respectively.

Pertanika J. Soc. Sci. & Hum. Vol. 2 No. 1 1994 5


TABLE 3Gross Value-Added, 1983 Actual versus Estimates (RM million)

SectorNo

EstimatesActual GDP blow-up

methodFinal-demand

blow-up methodInput-output

method1 Other Agriculture 2268.4 4538.8 2450.2 2260.32 Rubber Planting 2472.4 5883.4 2223.0 2476.73 Oil Palm 2497.2 3968.2 1719.8 2207.44 Livestock 700.8 909.1 654.4 584.85 Forestry 2169.5 4196.4 2755.3 2585.96 Fishing 1674.1 4130.7 1428.5 1377.07 Mining 7272.8 9368.8 8414.1 6643.98 Dairy Product 180.9 766.7 187.8 167.39 Vegetables and Fruits 87.6 366.2 120.7 123.110 Oil and Fats 558.8 1581.4 1072.8 1061.111 Grain Mills 143.3 851.1 131.9 146.512 Baker Confectionery 140.7 211.1 133.2 129.913 Other Foods 258.7 433.5 230.7 221.214 Animal Feed 99.5 124.6 96.2 81.215 Beverages 271.4 367.0 252.7 232.816 Tobacco 339.2 358.7 219.1 220.317 Textiles 313.4 763.2 367.1 381.418 Wearing Apparel 247.2 495.9 310.7 298.419 Sawmills 788.5 1796.3 812.5 978.820 Furniture and Fixtures 84.5 133.8 85.0 84.921 Paper Printing 589.1 757.9 871.3 373.622 Industry Chemical 742.3 350.6 2336.8 442.623 Paints Etc. 40.6 86.2 -484.1 70.824 Other Chemical Prod. 230.0 387.5 205.4 201.925 Petrol Products 252.4 508.9 566.0 337.926 Rubber Processing 279.3 584.2 268.2 271.127 Rubber Products 409.9 566.2 361.5 331.728 Plastic Products 193.1 267.4 243.9 175.629 Glass Products 211.7 317.0 529.0 267.330 Cement 270.5 224.5 -783.6 209.531 Non Metallic 187.6 209.6 437.4 187.432 Basic Metal 522.1 888.0 235.3 330.533 Other Metal 312.3 451.6 376.8 323.734 Non Elec. Machine 322.4 423.8 344.8 298.635 Electrical Machine 1428.8 1526.4 1521.1 1456.536 Motor Vehicle 658.7 366.4 323.2 3122.637 Other Transport 236.9 321.3 344.4 274.538 Other Mfg. Prod. 170.1 254.0 161.1 149.639 Construction 3845.6 4494.3 4141.0 3770.540 Other Services 27952.1 41204.0 28171.3 4598.3

61426.0 94437.2 63638.1 36649.1Source. Zakariah etal. (1991), p. 6

Pertanika J. Soc. Sci. & Hum. Vol. 2 No. 1 1994


Table 4 ranks the m ethods according to the num ber of sectors that give the smallest error for each m ethod. Among the three methods, the input-output m ethod has fourteen sectors with best estimates of value added while the final dem and blow-up and the GDP blow-up methods have the best estimates in thirteen and three sectors respectively. These figures indicate that the input-output m ethod gives the largest num ber of sectors whose projected value-added are closest to the actual value-added, followed by the final dem and blow-up and GDP blow-up methods.

The absolute errors in the estimates value added of each m ethod are in turn aggregated and presented in the table which show that the input- output m ethod has the lowest aggregate absolute error. The input-output m ethod has an aggregate of absolute errors of RM29.80 million with average errors per sector of RM2.12 million while those of final dem and blow-up and GDP blow-up

methods are RM32.84 million and RM34.47 million with average errors of RM2.52 million and RM 11.49 million respectively. These findings would support the view that the input-output m ethod provides the best m ethod of projecting value added, followed by the final dem and blowup and GDP blow-up methods.

Expressing the absolute errors in terms of their percentages to the actual values, Table 5 distributes the results according to various ‘class errors’. The first column of the table shows how much the estimated value added has deviated from its actual values in percentage terms while columns 2 - 4 show the num ber of sectors for each m ethod of projection corresponding to the percentage error class. The table shows that for the final dem and blow-up and input-output m ethods the lowest classes are also the modal classes whereas for the GDP blow-up m ethod the highest class is its modal class.

TABLE 4Evaluation of various methods of projecting Value Added

Methods No. of Sectors best estimate

Aggregate absolute error (RM million)

Rank

GDP blow-up 3 34.47 (11.49) 3Final demand blow-up 13 32.84 ( 2.52) 2Input-output 14 29.80 ( 2.12) 1Total 40 97.11Source. Computed from the models. Figures in parentheses indicate average errors.

TABLE 5Distribution of the Errors in Projection

Percentage Number of SectorCIIUI5

GDP blow-up Final demand blow-up Input-output0-20 5 21 25

21-40 9 6 1041-60 10 6 161-80 3 0 181 -100 5 0 2101 and above 12 7 1Total 40 40 40Source, computed from the models.

PertanikaJ. Soc. Sci. 8c Hum. Vol. 2 No. 1 1994 7


The table also shows that the input-output method has the lowest weighted mean percentage errors (24 percent), followed by final demand blow-up method (36.5 percent) and GDP blowup method (70 percent). Figure 1 which shows cumulative distribution of the errors in projection for the three methods substantiates this view by showing that the cumulative distribution curve for input-output method is located clearly above those for final demand blow-up and GDP blow-up methods, confirming that the input-output method has the largest number of sectors with the smallest percentage errors (the modal class of the input-output method is in the lowest class error).

Based on the above findings, we may conclude that of the three methods of projecting value- added, the inpu t-output m ethod is the best method of projection. We next tested the hypothesis that the mean errors of the three methods are significantly different from zero to find out which of the methods is the most efficient estimate of value added. A cursory look at the three arrays of projected value-added presented in Table 3 clearly shows that mean errors of the GDP blow-up m eth od are significan tly d iffe ren t from zero as the individual sector’s errors are consistently positive (indication of over-forecast) for all the sectors except for three sectors: industrial chemicals, cement and motor vehicles.

Fig 1: Percentage Errors of various methods compared

The above test on the remaining arrays of value-added projected was conducted by using the final demand blow-up and the input-output m ethods. Table 6 summarizes the results of the test of zero means on the two methods. The means of both methods are not significantly different from zero, leading us to infer that, on the average, the final demand blow-up m ethod and input-output

method would provide zero error in their projection of Malaysian sectoral value-added.

In assessing methods of projection of value- added between final dem and blow-up and input- output methods, even though they are equally good methods in terms of giving zero mean error (on the average), we would be inclined to choose the one which is relatively more efficient. To determine the relative efficiency between the final dem and blow-up m ethod and the input-output m ethod in projecting value-added, we would normally compare their variances.

Table 6 can be used to measure approximate relative efficiency of estimating value-added using the input-output m ethod with respect to the final demand blow-up m ethod by the following expression:

Variance of final dem and method Variance of input-output m ethod

Relative efficiency of estimating value-added using the input-output m ethod with respect to final dem and blow-up m ethod is approximately 175 percent, indicating the form er is 75 percent more efficient than the latter (given the same sample size which is the num ber of sectors in the economy). This measure of relative efficiency lends us to conclude that in estimating sectoral value added of the Malaysian economy, the input-output m ethod would give the best results.

TABLE 6Mean and standard deviation of errors in Projection

Method Mean Standard DeviationFinal demand 346 1650Input Output -21 191No. of sectors 39 39

Note: Not significandy different at one percent significant level. Sector 40 (Services sector) has been omitted to remove its extreme values from both arrays of value- added.

Having known that the best and most efficient method of estimating sectoral value added is the input-output method, we may then ask the question: will adjusting the value of the estimated value added by its mean error improve the estimates ? The answer is yes, if and only if, apart from hav

Pertanika J. Soc. Sci. Sc Hum. Vol. 2 No. 1 1994


ing zero mean, the variance of the estimated errors are independent and serially uncorrelated. But due to the limited availability of input-output tables, which to-date are available only for 1978 and 1983, the test for independent variance and correlation as suggested by Theil (1975) could not be done.

To carry out the test, we need time series of errors in projection (which can be derived from Table 1 for 1983 only) for a reasonable num ber of years. At this juncture, we can only assume that the variances of the errors in projection are indep en d en t and uncorrelated ; necessary adjustments, however, can be m ade to improve the forecast. Of course, violation of this assumption would increase the sampling variance and render the estimates inefficient

SUMMARY AND CONCLUSIONSThe results of this study show that the input-output m ethod of projecting Malaysia’s sectoral value- added is the best and most efficient m ethod of the three m ethods analysed, and are consonant with those of other studies. A priori, while final dem and blow-up m ethod has the advantage of allowing for different structures of final dem and over the GDP blow-up m ethod, the input-output m ethod has the added advantage by taking into account inter-industrial relationship which the other two m ethods do not.

By using the Malaysian input-output tables for 1978 and 1983, our analysis shows that the input- output m ethod has the largest num ber of sectors with estimates of sectoral value added being closest to the actual, rendering it to have the largest percentage m ean errors and the lowest percentage errors class as its modal class. In addition, even though average errors of the input-output and final dem and blow-up methods are not significantly d ifferen t from zero, the form er is found to be a m ore efficient estim ator as its variance is lower than that of the latter.

We can infer from the above results that the input-output m ethod is the best and most efficient m ethod in projecting Malaysia’s sectoral value added am ong the three m ethods considered. If the assumption of uncorrelated and independent variance am ong its errors is correct, adjusting the estimates with the erro r’s factor will certainly improve the estimates.

Reiterating the basic assumptions, for input- output forecasts to be accurate, the productive sectors described must behave approximately according to the assumptions governing the Leontief production function. The assumption of constant returns to scale is often argued to be reasonable when it is the behaviour of the sector as a whole which is under consideration. This is especially so in the manufacturing sector where relatively large concentrations of fixed capital can produce short-run cost curves which may be described as flat-bottomed (Midmore, 1992). This lends support to the argum ent that, also in the short-run, the proportions of intermediate inputs used will be more or less fixed in response to changes in demand. Whilst such a rationalisation may be appropriate in genaral terms, it can be argued that it is not so in the specific case of agriculture which is characterised by the presence of many, relatively small producers with a range of potential alternative outputs. Varying returns to scale, the diversification of output in response to changes in demand, and the structural instability of input-output coefficients, however, should all tend to cause bias in forecasts.

REFERENCESA b d u l A z iz A bd u l Ra h m a n , 1984. Some projection of

the West Malaysian economy to 1985 with emphasis on agriculture. Malay J. Agric. Econs. 1(1): 72-88.

F r e u n d , J.E, 1974. Mathematical Statistics, Second edition, Prentice/Hall International. London:

G o s h , A., 1964. Experiments with input-output model - an application to the economy of the U.K. 1948 - 55, University of Cambridge, Department of Applied Economics. Monograph 9. Cambridge University Press.

L eo n t ie f , W. 1986. Input-Output Economics, Second edition, Oxford: Oxford University Press.

M alaysian Industrial D evelopm ent A uthority (M I D A ) . 1985.

Malaysia, Department of Statistics. Input-output Table 1978.

Malaysia, Department of Statistics. Input-output Table 1983.

Malaysia, Ministry of Finance. Economic Reports, various issues.



Malaysia, Department of Statistics, Industrial Survey. 1987.

M i d m o r e , P. 1993. Input-output forecasting of regional agricultural policy impacts. JAgric. Econs., 44(2): 284-300.

S la s t r y , N.S.R., V.M. D a n d ek a r , S .G .T iw a r i, a n d U m a D a t t a , 1975. Paper on National Income and Allied Topics, Vol. 1-3.

T h e i l , H e n r i, 1975. Applied Economic Forecasting, Studies in Mathematical and Managerial Economics, Vol. 4, Amsterdam: North-Holland Publishing Co.

Zakariah A b d u l R a sh id , M o h d . S h ahw ahid H j . O th m an a n d A b d u l A ziz A b d u l Ra h m a n , 1991. Agricultural Projection and Policy Analysis - some experiments using input-output analysis. P a p e r p r e s e n t e d a t t h e Workshop on Agricultural Forecasting and Policy Models: Some empirical evidence. O r g a n is e d b y C e n t r e f o r A g r i c u l t u r a l P o l i c y S t u d i e s , U n iv e r s i t i P e r ta n ia n M a la y s ia .

(Received 30 June 1992)

10 PertanikaJ. Soc. Sci. 8c Hum. Vol. 2 No. 1 1994

Pertanika J. Soc. Sci. 8c Hum. 2(1): 11-20 (1994) ISSN: 0128-7702 © Universiti Pertanian Malaysia Press

Learning English in a Non-supportive Environment among Malay Learners in Secondary Schools

JAMALI ISMAILD epartm ent of Languages,

Faculty o f Educational Studies Universiti Pertanian Malaysia

43400 UPM Serdang, Selangor Darul Ehsan, Malaysia

Keywords: English competence, non-supportive environment, exposure of Malay learnersABSTRAK

Kertas keija ini dihasilkan berasaskan pada satu kajian yang bertujuan untuk menyelidiki perkara berikut: (i) tahap pendedahan terhadap bahasa Inggeris sebagai bahasa kedua di kalangan pelajar Melayu, and (ii) kekuatan pertalian antara pendedahan terhadap bahasa Inggeris di kalangan pelajar berkenaan dengan pencapaian mereka dalam bahasa tersebut. Andaian yang digunakan ialah semakin tinggi tahap pendedahan terhadap bahasa Inggeris yang diterima oleh pelajar berkenaan, maka semakin cekaplah mereka dalam menguasai bahasa tersebut. Sampel kajian melibatkan 441 orang pelajar Melayu Tingkatan Empat dari beberapa buah sekolah pilihan di Selangor. Untuk mengumpulkan data, soal selidik digunakan bagi mendapatkan maklumat tentang pencapaian pelajar dalam kertas Bahasa Inggeris peringkat SRP dan untuk mengukur tahap pendedahan mereka terhadap bahasa Inggeris. Penganalisisan data dilakukan dengan menggunakan prosedurjadual silang dan korelasi. Hasil kajian ini membuktikan bahawa para pelajar yang pada keseluruhannya lemah dalam bahasa Inggeris didapati menerima kadar pendedahan yang minimum terhadap bahasa tersebut. Terdapat juga bukti tentang wujudnya korelasi yangsignifikan antara pendedahan terhadap bahasa Inggeris dengan kecekapan dalam bahasa tersebut. Kertas kerja ini seterusnya m em bincangkan situasi pem elajaran bahasa Inggeris di Malaysia yang pada keseluruhannya tidak menggalakkan. Kertas keija ini juga mencadangkan strategi pengajaran yang mungkin boleh diambil untuk mengatasi masalah pemelajaran ini.

ABSTRACT

This study investigated the degree of exposure to English as a second language amongst Malay learners and the strength of the relationship between exposure to English amongst the learners and their competence in the language. It was assumed that the higher the degree of exposure to English the learners received, the more com petent they were in the language. The sample of the study consisted of 441 Form Four Malay learners from selected schools in Selangor. For the collection of data, a questionnaire was used to gather information relating to the learners’ achievement in the SRP English Language paper and to measure the degree of their exposure to English. The analysis of data was carried out by using the cross tabulation and correlation procedures. The study showed that the learners, who were generally weak in English, received a minimal am ount of exposure to the language. There was also indication of the presence of a significant relationship between exposure to English and competence in the language. The learning environment for English in Malaysia which is generally non- supportive is also discussed and possible teaching strategies to overcome the learning problem are suggested.

INTRODUCTIONEnglish in the Malaysian Education System Teachers of English today are generally disheartened by the deteriorating standard of competence in English am ong learners. In spite of its status as a second language (L2), English has in reality

moved towards that of a foreign language (FL). The assumption am ong educators and the public is that this is due to the existing education system (first implem ented in 1970) in which Malay is the medium of instruction in all schools while English, formerly the medium of instruction, is taught only as a subject.

Jamali Ismail

The Standard of English Proficiency Although English occupies the status of an L2 within the education system, it is genuinely an L2 only to a handful of English-educated urbanites to whom the deteriorating standard of competence in the language is of concern. Nonetheless, this situation is inevitable because the education system is not producing English-educated individuals as it had during the pre-1970s. At the same time, the majority of Malay-medium learners of today see English as nothing more than a school subject.The English Language Programmes in Schools Following the change in the education system, the exposure to English among learners has reduced considerably. To meet this situation, several measures have been taken by the Ministry of Education, such as setting up committees responsible for the planning of the English programmes for the primary and secondary school levels. Syllabuses have been designed and put into practice (e.g. the Communicational Syllabus for U pper Secondary Schools in 1977, the New Curriculum for Primary Schools in 1982, and the New Curriculum for Secondary Schools in 1988) and supporting materials such as handbooks and textbooks are produced for teachers and learners. Extensive exposure to English for learners is provided by the Educational Radio and Television Programmes of the Ministry through the educational media services. Apart from this, more teachers of English are trained through the pre-service and in-service Teaching of English as a Second Language (TESL) training programme offered in local colleges and universities; some promising candidates are sent overseas for the same purpose.

English in Tertiary Institutions In conformity with the education system, English is no longer part of the requirem ent for admission into local tertiary institutions. It is at the discretion of these institutions to determ ine the level of English required for their students and to design their own English curriculum. Normally, the programmes range from the very basic to the most advanced, depending on student needs. The primary concern of teaching English is not so much to enable students to speak or to write effectively, but rather to enable them to extract information

from reference materials which are mostly in English. In other words, the emphasis is on English for Special Purposes (ESP). Subsequently, student achievement in speaking and writing, as it is in schools, is far from satisfactory.

Factors contributing to Low Achievement in English There are a num ber of factors related to the education system that have contributed to low achievem ent in English, some of which are discussed below.

Contact with EnglishThe design and implementation of the existing education system are such that opportunities for contact with English among learners are greatly reduced. In schools, the allocation of class time for the learning of English, i.e. between 200 to 300 minutes per week as the sum total of exposure to the language, is insufficient. Given the time constraint, nothing much can be achieved by the learners.

Outside school, the situation is far from encouraging. The majority of learners come from deprived homes or from rural areas whose immediate need for English is almost non-existent; whose environment is detached from the use of the language. Unlike the situation in the pre- 1970s, the English learning environment for most learners today has diminished.

The SyllabusThe common content syllabus has to be covered by the teacher irrespective of the learners’ ability. Every item in the syllabus is taught and equal weight is given to all items for fear that an item no t taught may be tested in the exam ination (Rodgers 1979).

Thus teachers are left with no alternative apart from covering the entire syllabus and drilling the learners to answer examination type questions so that the passing rate could be m aintained at an acceptable level. In such a situation, very little learning takes place.

However, the pace of teaching cannot be slowed down for the under-achievers. What matters to the education authorities and the public is that learners should achieve some degree of success in examinations. Therefore, it has become a common practice among teachers that the focal

12 PertanikaJ. Soc. Sci. & Hum. Vol. 2 No. 1 1994



JAMALI ISMAILDepartment of Languages,

Faculty o f Educational Studies Universiti Pertanian Malaysia


Keywords: English competence, non-supportive environment, exposure o f Malay learners

ABSTRAKKertas keija ini dihasilkan berasaskan pada satu kajian yang bertujuan untuk menyelidiki perkara berikut: (i) tahap pendedahan terhadap bahasa Inggeris sebagai bahasa kedua di kalangan pelajar Melayu, and (ii) kekuatan pertalian antara pendedahan terhadap bahasa Inggeris di kalangan pelajar berkenaan dengan pencapaian mereka dalam bahasa tersebut. Andaian yang digunakan ialah semakin tinggi tahap pendedahan terhadap bahasa Inggeris yang diterima oleh pelajar berkenaan, maka semakin cekaplah mereka dalam menguasai bahasa tersebut. Sampel kajian melibatkan 441 orang pelajar Melayu Tingkatan Empat dari beberapa buah sekolah pilihan di Selangor. Untuk mengumpulkan data, soal selidik digunakan bagi mendapatkan maklumat tentang pencapaian pelajar dalam kertas Bahasa Inggeris peringkat SRP dan untuk mengukur tahap pendedahan mereka terhadap bahasa Inggeris. Penganalisisan data dilakukan dengan menggunakan prosedurjadual silang dan korelasi. Hasil kajian ini membuktikan bahawa para pelajar yang pada keseluruhannya lemah dalam bahasa Inggeris didapati menerima kadar pendedahan yang minimum terhadap bahasa tersebut. Terdapat juga bukti tentang wujudnya korelasi yangsignifikan antara pendedahan terhadap bahasa Inggeris dengan kecekapan dalam bahasa tersebut. Kertas kerja ini seterusnya membincangkan situasi pemelajaran bahasa Inggeris di Malaysia yang pada keseluruhannya tidak menggalakkan. Kertas keija ini juga mencadangkan strategi pengajaran yang mungkin boleh diambil untuk mengatasi masalah pemelajaran ini.

ABSTRACTThis study investigated the degree of exposure to English as a second language amongst Malay learners and the strength of the relationship between exposure to English amongst the learners and their competence in the language. It was assumed that the higher the degree of exposure to English the learners received, the more competent they were in the language. The sample of the study consisted of 441 Form Four Malay learners from selected schools in Selangor. For the collection of data, a questionnaire was used to gather information relating to the learners’ achievement in the SRP English Language paper and to measure the degree of their exposure to English. The analysis of data was carried out by using the cross tabulation and correlation procedures. The study showed that the learners, who were generally weak in English, received a minimal amount of exposure to the language. There was also indication of the presence of a significant relationship between exposure to English and competence in the language. The learning environment for English in Malaysia which is generally non- supportive is also discussed and possible teaching strategies to overcome the learning problem are suggested.

INTRODUCTION

English in the Malaysian Education System Teachers of English today are generally disheartened by the deteriorating standard of competence in English among learners. In spite of its status as a second language (L2), English has in reality

moved towards that of a foreign language (FL). The assumption among educators and the public is that this is due to the existing education system (first implemented in 1970) in which Malay is the medium of instruction in all schools while English, formerly the medium of instruction, is taught only as a subject.

Jamali Ismail

The Standard of English Proficiency Although English occupies the status of an L2 within the education system, it is genuinely an L2 only to a handful of English-educated urbanites to whom the deteriorating standard of competence in the language is of concern. Nonetheless, this situation is inevitable because the education system is not producing English-educated individuals as it had during the pre-1970s. At the same time, the majority of Malay-medium learners of today see English as nothing more than a school subject.

The English Language Programmes in Schools Following the change in the education system, the exposure to English among learners has reduced considerably. To meet this situation, several measures have been taken by the Ministry of Education, such as setting up committees responsible for the planning of the English programmes for the primary and secondary school levels. Syllabuses have been designed and put into practice (e.g. the Communicational Syllabus for Upper Secondary Schools in 1977, the New Curriculum for Primary Schools in 1982, and the New Curriculum for Secondary Schools in 1988) and supporting materials such as handbooks and textbooks are produced for teachers and learners. Extensive exposure to English for learners is provided by the Educational Radio and Television Programmes of the Ministry through the educational media services. Apart from this, more teachers of English are trained through the pre-service and in-service Teaching of English as a Second Language (TESL) training programme offered in local colleges and universities; some promising candidates are sent overseas for the same purpose.

English in Tertiary Institutions In conformity with the education system, English is no longer part of the requirement for admission into local tertiary institutions. It is at the discretion of these institutions to determine the level of English required for their students and to design their own English curriculum. Normally, the programmes range from the very basic to the most advanced, depending on student needs. The primary concern of teaching English is not so much to enable students to speak or to write effectively, but rather to enable them to extract information

from reference materials which are mostly in English. In other words, the emphasis is on English for Special Purposes (ESP). Subsequently, student achievement in speaking and writing, as it is in schools, is far from satisfactory.

Factors contributing to Low Achievement in English

There are a number of factors related to the education system that have contributed to low achievement in English, some of which are discussed below.

Contact with EnglishThe design and implementation of the existing education system are such that opportunities for contact with English among learners are greatly reduced. In schools, the allocation of class time for the learning of English, i.e. between 200 to 300 minutes per week as the sum total of exposure to the language, is insufficient. Given the time constraint, nothing much can be achieved by the learners.

Outside school, the situation is far from encouraging. The majority of learners come from deprived homes or from rural areas whose immediate need for English is almost non-existent; whose environment is detached from the use of the language. Unlike the situation in the pre- 1970s, the English learning environment for most learners today has diminished.

The SyllabusThe common content syllabus has to be covered by the teacher irrespective of the learners’ ability. Every item in the syllabus is taught and equal weight is given to all items for fear that an item not taught may be tested in the examination (Rodgers 1979).

Thus teachers are left with no alternative apart from covering the entire syllabus and drilling the learners to answer examination type questions so that the passing rate could be maintained at an acceptable level. In such a situation, very little learning takes place.

However, the pace of teaching cannot be slowed down for the under-achievers. What matters to the education authorities and the public is that learners should achieve some degree of success in examinations. Therefore, it has become a common practice among teachers that the focal



point of classroom teaching is the syllabus and not the learners. The syllabus, which should have been treated as a guide for teaching, is not adapted to meet the level of a particular class.

Public ExaminationThere are two public examinations conducted by the Ministry of Education in which English is a compulsory subject for all candidates, i.e. the Sijil Rendah Pelajaran or SRP (the junior certificate) and the Sijil Pelajaran Malaysia or SPM (the school certificate) examinations. But English is not a prerequisite to certification in the sense that even if a candidate fails the subject he could still pass the examinations. This has affected the learners’ attitudes and motivation to learn the language which are detrimental to achievement.

Teachers of EnglishThe shortage of English teachers continues to be acute especially in the rural schools. Some do not have any TESL training and are specialists in other disciplines (see Table 6). The shortage is aggravated by the annual increase in the number of the school-going population1. At the same time, it is difficult to get candidates from among the Malay- medium school leavers to be recruited as teachers of English. Therefore, with a class of between 30 to 40 learners and with several classes of English to teach, the workload for teachers is very heavy. This has serious repercussions on the quality of teaching.

The LearnersAmong learners, at one extreme, there is a handful who are proficient in English. At the other extreme, there are the poorly proficient ones who form the majority. The latter are generally not exposed to English except, of course, during the English lessons. Their attitudes are largely determined by the language learning situation and by examination priorities. They would rather spend more time on other subjects which are compulsory for certification and neglect English which has no impact on examination results.

The negating factors discussed above illustrate the fact that the learning environment for most

learners is generally non-supportive. This is chiefly due to the lack of exposure to English as observed in a study reported below.

THE STUDYPurpose of the StudyThe study examined the issue of the failure of learners to achieve an acceptable level of competence in English despite having learned the language for many years in school. Teachers often remarked that the learners’ failure was due to, among other things, insufficient exposure to the language. To confirm this, quantitative data relating to competence as the dependent variable and exposure as the independent variable were gathered for analysis.

The questions that guided the study were as follows:(1) Does the level of English competence among

M alay-medium learners ind icate u n d erachievement?

(2) Are the learners sufficiently exposed to English?

(3) Is there a significant relationship between competence in English and exposure to the language amongst the learners?The assumptions were as follows:

(1) The learners were generally weak in English.(2) They were not highly exposed to the language.(3) There was a significant relationship between

competence and exposure, i.e. the higher the degree of exposure to English the more competent the learners were in the language.

METHODOLOGYThe study involved 441 Form Four Malay learners of English as an L2 from nine selected schools in the Kelang and Kuala Langat-Sepang districts of Selangor who had been studying the language for nine years. The sampling technique adopted was that of cluster sampling, i.e. a number of schools were identified and selected, and from these schools all learners from the required ethnic group (i.e. Malay) and the required educational level (i.e. Form Four) were selected.

Although the study was conducted within the confines of an insufficiently large sample of

'Based on figures from the Ministry of Education, 117,242 government school candidates sat for the SPM examination in 1980 and 168,344 in 1990; therefore the annual increase of the school-going population is about 2.76%.

PertanikaJ. Soc. Sci. & Hum. Vol. 2 No. 1 1994 13


point of classroom teaching is the syllabus and not the learners. The syllabus, which should have been treated as a guide for teaching, is not adapted to m eet the level of a particular class.

Public ExaminationThere are two public examinations conducted by the Ministry of Education in which English is a compulsory subject for all candidates, i.e. the Sijil Rendah Pelajaran or SRP (the ju n io r certificate) and the Sijil Pelajaran Malaysia or SPM (the school certificate) examinations. But English is not a prerequisite to certification in the sense that even if a candidate fails the subject he could still pass the examinations. This has affected the learners’ attitudes and motivation to learn the language which are detrim ental to achievement.

Teachers of EnglishThe shortage of English teachers continues to be acute especially in the rural schools. Some do not have any TESL training and are specialists in other disciplines (see Table 6). The shortage is aggravated by the annual increase in the num ber of the school-going population1. At the same time, it is difficult to get candidates from among the Malay- medium school leavers to be recruited as teachers of English. Therefore, with a class of between 30 to 40 learners and with several classes of English to teach, the workload for teachers is very heavy. This has serious repercussions on the quality of teaching.

The LearnersAmong learners, at one extreme, there is a handful who are proficient in English. At the other extreme, there are the poorly proficient ones who form the majority. The latter are generally not exposed to English except, of course, during the English lessons. Their attitudes are largely determ ined by the language learning situation and by examination priorities. They would rather spend more time on other subjects which are compulsory for certification and neglect English which has no impact on examination results.

The negating factors discussed above illustrate the fact that the learning environment for most

learners is generally non-supportive. This is chiefly due to the lack of exposure to English as observed in a study reported below.

THE STUDYPurpose of the StudyThe study examined the issue of the failure of learners to achieve an acceptable level of competence in English despite having learned the language for many years in school. Teachers often remarked that the learners’ failure was due to, among other things, insufficient exposure to the language. To confirm this, quantitative data relating to com petence as the dependent variable and exposure as the independent variable were gathered for analysis.

The questions that guided the study were as follows:(1) Does the level of English competence among

M alay-m edium lea rn e rs in d ica te u n d e rachievement?

(2) Are the learners sufficiently exposed to English?

(3) Is there a significant relationship between competence in English and exposure to the language amongst the learners?The assumptions were as follows:

(1) The learners were generally weak in English.(2) They were not highly exposed to the language.(3) There was a significant relationship between

competence and exposure, i.e. the higher the degree of exposure to English the more competent the learners were in the language.

METHODOLOGYThe study involved 441 Form Four Malay learners of English as an L2 from nine selected schools in the Kelang and Kuala Langat-Sepang districts of Selangor who had been studying the language for nine years. The sampling technique adopted was that of cluster sampling, i.e. a num ber of schools were identified and selected, and from these schools all learners from the required ethnic group (i.e. Malay) and the required educational level (i.e. Form Four) were selected.

Although the study was conducted within the confines of an insufficiently large sam ple of

'Based on figures from the Ministry of Education, 117,242 government school candidates sat for the SPM examination in 1980 and 168,344 in 1990; therefore the annual increase of the school-going population is about 2.76%.


Jamali Ismail

Malay learners of English as an L2, being Malay- educated and with a generally rural background, they characterized other Malay learners of English as an L2 in the country.

A questionnaire was used, modelled on the Likert-type scale, to measure the degree of exposure to: (1) written English (through reading materials such as books, magazines/periodicals and newspapers), (ii) scripted spoken English (as heard through radio and television); and (iii) unscripted spoken English (through verbal interaction with family members and friends). The questionnaire also provided information relating to the learners’ competence in English as indicated by their achievement in the SRP English paper.

A total score of a cluster of items for each of these three variables was computed - the top two- fifths of the total score representing high exposure, the middle one-fifth moderate exposure, and the bottom two-fifth low exposure. Therefore, it was the overall score on exposure and not the individual score for each item that formed the basis of categorization.

RESULTS AND DISCUSSIONCompetence in English:The learners were found to be generally weak in English as indicated by their achievement in the SRP English paper (Table 1). Only 14 (3.2%) obtained grades 1-2 (distinction) and 116 (26.3%) grades 3-6 (cred it), as many as 212 (48.1%) grades 7-8 (weak pass) and 99 (22%) grade 9 (fail).

TABLE IDistribution of sample on SRP English paper

Grade Frequency1-2 (Distinction) 14 (3.2%)3-6 (Credit) 116 (26.3%)7-8 (Pass) 212 (48.1%)9 (Fail) 99 (22.4%)Total 441 (100%)

Exposure to English:Overall, the students were generally found to be insufficiently exposed to: (i) written English

(Table 2), (ii) sc rip ted spo ken E nglish (Table 3) and (iii) unscripted spoken English (Table 4).

TABLE 2Distribution of sample on exposure to

written EnglishExposure Frequency

LowModerateHigh

367 (83.2%) 64 (14.5%) 10 (2.3%)

Total 441 (100%)


scripted spoken EnglishExposure Frequency

LowModerateHigh

173 (39.2%) 218 (49.4%)

50 (11.3%)Total 441 (100%)


unscripted-spoken EnglishExposure Frequency

Low 355 (80.5%)Moderate 74 (16.8%)High 12 (2.7%)Total 441 (100%)

The learners were not highly exposed to written English in the sense that very few of them read enough English books, magazines/periodicals and newspapers. Their contact with the scripted form of the language, through radio and television, was slightly higher than their contact with the written and unscripted spoken forms of the language. This could possibly be due to the entertaining nature of the media. Exposure to the unscripted spoken form of the language was also minimal. Very few used English when interacting with family members and friends.



Relationship between English Competence and Exposure:The correlation (Pearson product-moment) matrix (Table 5) shows significant relationships between com petence in English (SRP) and exposure (i.e. exposure to written English [EWE]; exposure and scripted spoken English [ESSE]; and exposure and unscripted spoken English [EUSE] as a cluster of variables).

TABLE 5Inter-variable correlation coefficients

(Competence-exposure)SRP EWE ESSE EUSE

SRP 1.000 .183 .156 .264*** *** ***EWE 1.000 .190 .397*** ***ESSE 1.000 .285***EUSE 1.000

Total Sample: 441 *** Significant at <0.001 ** Significant at <0.01 * Significant at <0.05Abbreviations:SRP Sijil Rendah Pelayaran English Paper EWE Exposure to Written English ESSE Exposure Scripted Spoken English EUSE Exposure to Unscripted Spoken English

The correlations between the variables yield values of:(1) r-0.183 p<0.001 on EWE and SRP;(2) r=0.156 p<0.001 on ESSE and SRP; and(3) r=0.264 p<0.001 on EUSE and SRP.

The low but positively significant correlations offer slight support for the assumption that learners who had a higher degree of exposure to English (in this case written English, scripted spoken English and unscripted spoken English) are more competent in the language than those who did not.

The link between the two variables as indicated in the table is com patible with previous studies (Briere 1978, Chandrasegaran 1979, Chesterfield etal. 1983, Hamayan etal. 1977 and Seliger 1977). This implies that exposure is im portant in attaining proficiency in the Target Language (TL). Exposure provides a learner with opportunities to communicate in the TL which is considered

the most effective means of learning the language (Spolsky 1968). Hence, the lower the degree of exposure the lower will be the expected level of proficiency.

The Learning Environment for English as an L2 in MalaysiaStudies byjam ali Ismail (1990a, 1990b) indicate that the Malay-medium learners’ weakness in English is neither a question of negative attitudes nor the lack of motivation. No doubt attitudinal-motivational variables are im portant but that favourable attitudes and strong motivation are not necessarily a condition for success in L2 learning. Their weakness is apparently a question of insufficient exposure to the language and probably the outcom e of some instructional as well as family socio-economic variables.

Outside the classroom, at home especially, exposure to English among learners is at its minimum. In other words, the environment is not at all supportive for the learning of English. In schools every measure is taken to facilitate learning - i.e. the introduction of new curricula, the use of effective methodology, the training and retraining of teachers. However, there is no denying that learning needs to be supplemented by a supportive environment outside the classroom to enable the learners to improve their performance in English through active use of the language. Upshur (1968) has confirmed that the most efficient language learning occurred in informal situations outside the classroom when the learner must make communicative use of the language variety to be learned.

Studies conducted by Briere (1978), Chesterfield et al. (1983), Hamayan et al. (1977) and Seliger (1977), for exam ple, provide am ple evidence indicating the importance of exposure in determ ining success in L2 learning. This being the case, learners of the L2 in the L2 community (i.e in the ‘natural’ L2 setting) are at the advantage of substantially being exposed to the language whereas the great majority of L2 learners learning the L2 outside the L2 community (as in the case of learning English in Malaysia) are not.

It is now evident that learning an L2 in a formal setting must be supplemented by exposure and practice in a supportive informal setting. But the setting in Malaysia is not always supportive in

Pertanika J. Soc. Sci. 8c Hum. Vol. 2 No. 1 1994 15

Jamali Ismail

terms of exposure to the language and in terms of the attitutes of the teachers, the learners and the community.

First, in schools some teachers fail to understand the learning situation and the attitutes of the learners. They also have a low opinion and low expectation of their learners. But teachers are rarely aware of their own attitudes. In addition, the teachers might come from different socio-economic and educational backgrounds, or they might be teachers of other subjects and therefore lack the insight of psychological, sociological, linguistic and methodological aspects of L2 teaching. Worse still, the teachers might not be proficient enough in the language (probably due to the environment and not that they refuse to improve themselves - The Star; May 6, 1991: 2) and teach the subject due to the shortage of staff. Consequently, there is a great range of difference in the m anner teachers deal with learners.

In the placement of teachers, the tendency is to send the newly trained young teachers to serve in rural schools (where the bulk of the school-going population in the country is) since in such schools the shortage of English teachers is critical. Senior teachers generally refuse to serve in the schools for too long, their preference being the schools in the bigger towns. This also applies to women teachers (the majority of whom happen to be teachers of English) who prefer to be with their husbands who serve as senior executives in town ( The Star; March 6, 1991: 1). Consequently, some urban schools have excess teachers of English who end up teaching o ther subjects while the rural schools continue to be staffed by new and inexperienced teachers (Table 6).

Second, learners, whether from the villages, small towns or even the state capitals, generally share a common ethos, i.e. respect for authority and the elderly. Therefore in class, in front of their teacher - the authority - they appear passive. In reality they are shy but, given the proper encouragement, they could be very responsive. Unfortunately, their passiveness has often been misinterpreted as laziness, which some teachers fail to understand especially those who are ill- equipped for the job but have been compelled by circumstances to teach English.

Shyness among the learners leads to reluctance to practise speaking English. They are generally reticent and lack the confidence to speak the language for fear of making mistakes and being laughed at. It has been proven that shyness affects performance in the language; the less shy the learners are the better they perform (Hamayan et al. 1977).

Finally, the com m unity within which the learners live generally disfavour the use of English. Those who try to speak English might be ridiculed or regarded as showing off. At the same time the general feeling is that the ability to speak English is regarded as a sign of belonging to a higher social class and a symbol of urbanization although not all urban people are English-speaking (Asmah Haji O m ar 1975). Actually, since colonization, the prestige of English am ong Malaysians has always been unquestionably high. Their expressed lack of interest in English is merely an excuse for their being weak in the language. Malay learners are therefore ‘sort of in a dilem m a’ (Jamali Ismail 1990a).

TABLE 6Distribution of English teachers (%) based on years of teaching experience by schools

Schools Years of teaching experience Total:-5 6-10 11-15 16-20 21-25 +26

Rural 73.0 18.9 5.4 2.7 - - 100.0%(N=37)

Urban 30.6 32.7 6.1 18.4 10.2 2.0 100.0%(N=49)

Note: The total N (i.e. 86) were sampled from eleven schools (four rural and five urban) in Selangor. Among them, 59.5% of those in rural schools and 32.7% of those in urban schools were teachers of other disciplines not trained to teach English.



SUGGESTIONS FOR TEACHINGIn spite of the non-supportive environment, some measures could still be taken to encourage the learning of English. First, the possibility of streaming or regrouping learners needs to be taken into account since the wide range of learner ability in class is rarely given a tten tio n . T he effect of regrouping on the weaker learners can be psychologically detrimental when the feeling of being inferior creeps in. However, regrouping can narrow the range of language ability of a particular class since the advanced and the weaker learners are placed in separate classes. At the same time, this will ease the teacher in his task as he no longer has to face a class with a wide range of language ability, leaving the brighter learners unchallenged and the weaker learners neglected. Besides, whenever necessary, remedial measures can be taken to help the weaker students.

Second, class size has to be reduced to enable the teacher to pay individual attention to learners. It is not uncom m on to see, at all school levels, classes consisting of about 30 to 40 (or even more!) learners. This is due to the shortage of teachers and classrooms both as a result of a growing num ber of school children from year to year (as discussed earlier). In such a situation, a teacher will find it difficult to attend to the learners’ individual needs. For the English teacher, with several classes of English to teach and with so many learners in a class, a thorough assessment of learners’ work is a heavy burden. The only way out is to reduce the am ount of assignments; this means that there will be lack of practice on the part of the learners, the outcome of which is detrimental to achievement. Further, in the context of Malay- medium schools, since it is the English teacher who is the main conversational partner for the learners as a resource of increasing English proficiency, the smaller the class size the greater will be the opportunity for teacher-leamer interaction in the language.

Third, teacher qualifications need to be improved by sending more teachers for in-service courses or to provide wider opportunities for them to go for further studies. A teacher should have a sound knowledge relating to his subject and to his job. He should

... have acquired an understanding of current theories of language acquisition and

to be familiar with and attuned to the sociocultural traditions of the students.

(Tucker cited in Alptekin 1981: 281). Unfortunately, the training programmes them selves do no t often help develop cross-cultural understanding as part of teachers’ pedagogic task. In the Malaysian context, this is of prime importance owing to the multi-ethnic background of the country. Teachers, who might come from different sociocultural backgrounds, should be sensitive to the rural sociocultural traditions of his students.

Teacher attitudes and teacher expectation of learner perform ance need to be suitably adjusted to m eet the present English learning situation. He must understand his learners and, in addition, he must be interested in them more than anything else. It is the teacher who determines how well and how much the learners should learn. An enterprising and creative teacher will always be able to devise and adapt his teaching materials to his learner’s ability. In the hands of a skillful teacher, even unsatisfactory teaching material can be used successfully in class. However, not all teachers are ‘b o m ’ - creativity and skill are developed through experience.

The shortage of teachers has to be overcome by training more teachers, i.e. pre-service training. The possibility of getting assistance from Eng- lish-speaking countries in the supply of teachers should be given due consideration. Above all, well- tra ined and experienced teachers should be posted to rural schools ( The Star, May 5, 1991).

Finally, ways and means have to be sought to compensate for the insufficient exposure to English, at least in class. This can be achieved by providing as much opportunity as possible for the learner to practise speaking (and later writing) in the language. Speaking is the best way to learn to speak. The speaking task given should be within the learner’s ability to perform . Success in his perform ance will create interest and confidence.

Creating situational contexts and role play can provide opportunity for the learner to practise speaking in the language (Haycraft 1978). Opportunities to be in the real situation for most learners are very rare or even non-existent. So, the teacher has to depend on simulated situations in the classroom.

The situation will be controlled carefully to teach the new language material...in


Jamali Ismail

such a way that there can be no doubt in the learner’s mind of the meaning of what he hears’ (Pittman 1967).

The primary value of foreign language classes....may be, therefore, the creation of a range of situations in which the student may learn the language varieties appropriate to those situations....

(Upshur 1968: 121).

Language laboratories can undoubtedly facilitate aural-oral practice (Dakin 1973; Haycraft 1978: Howattand Dakin 1974). But, the high cost of establishing one is beyond the means of the average school. However, a cassette-recorder can be used as an alternative and, being cheap, almost all schools should be able to afford one. The English teacher should certainly have one as part of his teaching equipment. The cassette-recorder can be a vital tool in listening comprehension exercises. It also provides opportunities for the pupils to listen to other speakers - perhaps even native speakers of English - besides the class teacher. Varieties of teaching m aterials for aural-oral practice are available in the market. A teacher could also prepare his own materials to suit the needs of his class.

The possibility of using a video recorder can also be explored as it is easily available in the market. Video is apparently superior to audio in helping a learner to understand what is being said because ‘...gestures, the physical con tex t and behavioral clues are all p resen t’ (Gower and Walters 1983: 163).

Equipment like computers can also be used in teaching. In fact, this equipment is now becoming increasingly popular among language teachers. Many teacher training institutions have introduced computer-assisted language learning course (or CALL) as a com ponent in their training programme.

There are also other activities, i.e. singing and listening to selected songs, language games, etc., that can be carried out to increase the volume of communicative exercises and at the same time to make English learning more meaningful and enjoyable. Materials from magazines and newspapers can be adapted and used for teaching. All available teaching aids should be used extensively

during the teaching session. Their use should be entertaining in nature so as to create interest among learners. If necessary, the session may be followed by a non-taxing simple discussion to provide the learners opportunities to speak in the language.

Given the time constraint for English teaching, the use of situational contexts in class, audiovisual and other teaching aids would increase the am ount of contact with the TL among learners. The use of such aids is actually nothing new in language teaching, teachers being advised to make full use of them even while under training.

Last, but not least, the school library should be provided with carefully selected entertaining materials such as comic books and pop music, movie or sports magazines, apart from usual materials for heavy reading found in school libraries. It has been the tradition among some school administrators to ban such entertaining materials from school premises. But, considering their popularity among youngsters, undoubtedly due to the pictures rather than what is written, their availability might initiate exposure to written English eventually. School administrators are well aware that most school libraries, stocked with nothing but materials for heavy reading (some in English which, for the majority of students, are incomprehensible,) attract only a handful of book-worms. Surely, with the availability of more entertaining materials, the library might be frequented.

It should be noted that the suggestions above are not really new to teachers. But the question is: To what extent have these suggestions been taken into account in teaching?

It should also be noted that no attem pt is made to discuss teaching methods. This is not to say that m ethods are of no im portance in accounting for learner achievement. Yet, effective L2 learning depends, amongst other factors, on the am ount of exposure to and practice in the language (Spolsky 1968).

CONCLUSIONThe deterioration in the standard of English competence among Malay-medium students is a fact that has to be accepted. Unfortunately, certain quarters of the general public are unable to come to terms with the reality of the situation so much so that even the drop in the passing rate for Eng



lish in the 1990 SPM examination which accounts to only 1.5% (the passing rate for 1989 being 58.6% and for 1990 57.1%) has been given too much coverage in the mass media. But, the drop in Bahasa Malaysia and some of the sciences and arts subjects ( The Star, March 6,1991:1) is ignored. Being proficient in English is undeniably advantageous; bu t only a small p roporation of the Malaysian population really needs English. Among them are those in the diplomatic and commercial professions where English is essential for international communication and those in the academic field where English is needed for the acquisition of knowledge.

Comparing the num ber of candidates taking the English paper in public examinations against the num ber of those who fail to make the grade, the deterioration in the standard of competence appears alarming. However, the public should be aware that no t every Malaysian needs English and that every year there are still quite a num ber of candidates taking and passing the 1119 Cambridge GCE English Language paper.

From the discussion above, it is noted that the direct result of the existing education system is the recession and spread of English in the country. Before the im plem entation of the system, the standard of English was high but the learning of the language was confined only to those attending English schools in towns. Therefore, the language did not really reach the Malaysian masses. Now the standard has declined but everbody in schools throughout the country has the opportunity to learn the language. Despite the decline, there is still room for improvement in order to maintain English as a strong second language without necessarily changing the existing education system.

REFERENCESAsmah Haji Omar. 1975. The Malaysian mosaic of lan

guages. In Essays on Malaysian linguistics, ed. Asmah Haji Omar. Kuala Lumpur: Dewan Bahasa and Pustaka.

A lptek in , C em . 1981. Sociopsychological and pedagogic considerations in L2 acquisition. TESOL Quarterly 15: 275-284.

B r ie re , E u g e n e J. 1978. Variables affecting native Mexican children’s learning Spanish as a second language. Language Learning 28: 159-174.

C han drasegaran , A n t o n ia . 1979. Problems of learning English in national schools in Johor, Malaysia: an investigation of attitudinal-motivational variables, learning strategies and exposure to English. MA Dissertation. University of Singapore.

C hesterfield , R ay, B. C hesterfield , H ayes-L atimer and C hav ez . 1983. The influence of teachers and peers on second language acquisition in bilingual preschool programs. TESOL Quarterly. 17: 401-419.

D ak in , J u l ia n . 1973. The language laboratory and language learning. London: Longman.

G ow er , R. and S. W alters. 1983. Teaching practice handbook. London: Heinemann Educational Books.

H amayan, G enesee and G . T ucker. 1977. Affective factors and language exposure in second language learning. Language Learning 27: 255-241.

H aycraft, J o h n . 1978. An introduction to English language teaching. London: Longman.

HowATrandJ. D ak in . 1974. Language laboratory materials. In The Edinburgh course in applied linguistics volume 3: techniques in applied linguistics, eds J.P.B. Allen and Pit Corder. Oxford: Oxford University Press.

J amali Ism ail . 1990. Perception of learners’ motivation in relation to competence in English as a second language. Paper presented at the National Seminar on Teaching of English as a Second Language, Aminuddin Baki Insitute, Genting Highlands, January.

------------ 1990b. C om petence, attitudes andmotivation in language learning with special reference to learners of English as a second language. Paper presented at The Workshop on Pragmatic Communicative Language Teaching Methodology. MARA Institute of Technology, Shah Alam, Selangor, July.

P it tm a n , G.A. 1967. Teaching structural English. London: Ginn and Co. Ltd.

R o d g er s , J. 1979. The English language programme in Malaysian schools. Paper presented at the National Seminar on The Teaching of English in Institutions of Higher Learning in the 1980% Institut Teknologi MARA, Shah Alam, Malaysia.

S eliger, W. 1977. Does practice make perfect? A study of interaction patterns and L2 competence. Language Learning 27: 263-278.


Jamali Ismail

S polsky, B. 1968. Recent research in TESOL. TESOL U psh u r , A. 1968. Four experiments on the relationQuarterly, 2, 304-307, 1968.

T hk S tar, March 6, 1991.----------- March 5, 1991._______ May 6, 1991

between foreign language teaching and learning. Language Learning 18: 111-124.

(Received 11 November 1992)


PertanikaJ. Soc. Sci. & Hum. 2(1): 21-28 (1994) ISSN: 0128-7702© Universiti Pertanian Malaysia Press

The Relationship between Quality of Home Environment and Mental Scores o f Children attending the UPM Laboratory Preschool*

ROHANI ABDULLAH, SITI NOR YAACOB,ROZUMAH BAHARUDIN

D epartm ent o f Family Developm ent Studies,Universiti Pertanian Malaysia,

43400 UPM Serdang, Selangor Darul Ehsan, Malaysia.

Keywords: preschool children, mental performance, home environment, family backgroundABSTRAK

Kajian ini bertujuan untuk menentukan perkaitan di antara kualiti persekitaran rumah yang disediakan oleh ibu bapa untuk anak-anak mereka dan prestasi mental anak-anak. Kajian ditumpukan kepada 64 orang kanak-kanak berum ur di antara 5 hingga 6 tahun, yang menghadiri Makmal Prasekolah Universiti Pertanian Malaysia. Hasil kajian menunjukkan secara keseluruhannya kanak-kanak yang terlibat di dalam kajian ini mempunyai keluarga yang menyediakan persekitaran rumah yang menggalakkan. Selaras dengan kajian-kajian yang lalu, kualiti persekitaran rumah responden mempunyai perkaitan dengan latarbelakang keluarga. Kedua-dua pendidikan dan pendapatan ibu bapa mempunyai perkaitan dengan kualiti persekitaran rumah. Kualiti persekitaran rumah juga didapati mempunyai perkaitan yang signifikan dengan prestasi mental kanak-kanak. Kanak-kanak daripada persekitaran rumah yang meransang kognitif memperolehi skor yang lebih tinggi di dalam semua sub-skala Weschler Preschool and Primary Scales of Intelligence daripada kanak-kanak lain dalam kajian.

ABSTRACT

The purpose of this study was to determine the reladonship between the quality of home environment that parents provide for their children and the children’s mental performance. The study focused on 64 children, aged 5 to 6 years, from the Universiti Pertanian Malaysia Laboratory Preschool. Results of the study show that, generally, most of the children came from a relatively supportive home environment. Consistent with past research, the quality of home environment correlated with family background. Both parents’ education and income correlated with the quality of the home environment. The quality of the home environment correlated significantly with the children’s mental performance. Children from a cognitively stimulating home environm ent achieved higher scores on all the subscales of the Weschler Preschool and Primary Scales of Intelligence.

INTRODUCTIONO f the various environmental factors that influence developm ent during early childhood, the family and home environment represents one of the most significant factors in a young child’s life. Previous studies have provided evidence that a secure child who has had a harmonious and positive relationship with his parents tends to become a com petent, responsible and productive adult (Belsky et al. 1984; Maccoby and Martin, 1983). Findings from recent studies also support the association betw een hom e env ironm ent and a child’s level of development during early childhood (Beckwith and Cohen, 1988; Bradley and

Caldwell, 1989; Bradley and Casey, 1984; Bradley etal.f 1989; Bullock and Pennington, 1988; Ninio and Rinott, 1988; Rozumah and Luster, 1992). These studies revealed that a home with appropriate tools for stimulation and a warm, facilitative relationship between parents and children made up the best combination that would encourage optimal learning in children.

Bee (1985) in her review of past studies summarized that the home environment of children who achieved higher IQ scores had appropriate play materials and parents who were democratic, emotionally and verbally responsive and had realistic goals for their children. A study by Rohani

* An earlier version of this paper was presented at the ‘Seminar Toward Enhancing the Quality of Life\ Universiti Pertanian Malaysia, Serdang, Selangor, 18th December 1989.

Rohani Abdullah, Siti NorYaacob and Rozumah Baharudin

et al. (1989) reported similar findings. The mental abilities of the children were associated with the availability of appropriate play materials, variety of stimulation, organization of the physical environment and availability of language and academic stimulation within the home environment. Bradley^ al (1989) reported a collaborative study which also provided similar evidence. The study found that measures of specific aspects of the ch ild ’s hom e env ironm en t such as paren ta l responsitivity and availability of stimulating play materials were more strongly related to children’s cognitive development than global measures of environmental quality such as family socio-economic status. The importance of certain aspects of the home environment on children’s achievem ent was also investigated by Rozumah (1992). The study found that children who did well on achievem ent tests had m others who provided more cognitively stimulating home environments. Children whose parents had a higher level of education and a higher family income were also found to perform better on the tests.

The effect of the early home environment is not transient. Longitudinal studies have indicated that children who had experienced a stimulating home environment during their childhood years tended to thrive better at schools (Bradley and Caldwell, 1984a; Bradley et a l , 1988a; Bradley et al., 1988b). The early home environment is therefore a valuable indicator of the learning environm ent available to the child early in life and the potential perform ance of the child when he enters formal schooling.

The present study was designed to determine the relationship between the quality of home environment and the children’s level of mental tests perform ance with three specific objectives in mind: 1) To determ ine the quality of the home environment available to the preschool child; 2) To determ ine the relationship between family background and the quality of the ch ild ren’s home environment; and 3) To determ ine the relationship between the quality of the home environm ent and the children’s mental performance.

METHODOLOGYSampleSubjects for the study were 64 children aged 5 to 6 years from the Universiti Pertanian Malaysia

Laboratory Preschool. Most (90%) of the children were Malays, and the rest were Indians and Chinese. Fifty-two percent were males, and 48% were females.Instrumenta. Home Inventory

The quality of the children’s home environm ent was assessed using the Preschool Version of the HOME (Home Observation for Measurement of the Environment) Inventory by a trained research assistant. The HOME has been shown to be a highly reliable and valid measure, and a strong predictor of child outcomes (Bradley and Caldwell, 1984a; Gottfried, 1984; Suriani, 1988). Suriani found an alpha coefficient of .89 for the HOME in her study on Malaysian children. The preschool HOME Inventory is an observation/interview technique that assesses the quality of physical, cognitive, social and emotional stimulation available to the child within the home environment.

The Preschool Version of the HOME Inventory (henceforth HOME) is composed of eight subscales. These subscales include: 1) toys, games and reading materials; 2) language stimulation; 3) physical environment; 4) pride, affection and warmth; 5) stimulation of academic behavior, 6) modeling and encouragem ent of social maturity;7) variety of experience; and 8) acceptance. Each item of the subscales is scored by giving 1 point for the presence of stimulation or 0 for the absence of stimulation for the child within the home. The total HOME is determ ined by adding scores of the 8 subscales of HOME. High scores indicate high quality home environment. Caldwell and Bradley (1984) reported an internal consistency of .93 for HOME. HOME was translated into Bahasa Melayu (the national Malaysian Language) and retranslated into English. A ninety-five percent agreem ent was calculated between the original and the retranslated versions.

b. Weschler Preschool and Primary Scales ofIntelligence (WPPSI).The WPPSI, designed to assess the intellec

tual abilities of preschool children, are divided into 10 major components: 1) information; 2) vocabulary; 3) arithmetic; 4) similarities; 5) comprehension; 6) animal house; 7) picture completion;


PertanikaJ. Soc. Sci. & Hum. 2(1): 21-28 (1994) ISSN: 0128-7702© Universiti Pertanian Malaysia Press

The Relationship between Quality of Home Environment and Mental Scores of Children attending the UPM Laboratory Preschool*

ROHANI ABDULLAH, SITI NOR YAACOB,ROZUMAH BAHARUDIN

Department of Family Development Studies,Universiti Pertanian Malaysia,


Keywords: preschool children, mental performance, home environment, family background

ABSTRAKKajian ini bertujuan untuk menentukan perkaitan di antara kualiti persekitaran rumah yang disediakan oleh ibu bapa untuk anak-anak mereka dan prestasi mental anak-anak. Kajian ditumpukan kepada 64 orang kanak-kanak berumur di antara 5 hingga 6 tahun, yang menghadiri Makmal Prasekolah Universiti Pertanian Malaysia. Hasil kajian menunjukkan secara keseluruhannya kanak-kanak yang terlibat di dalam kajian ini mempunyai keluarga yang menyediakan persekitaran rumah yang menggalakkan. Selaras dengan kajian-kajian yang lalu, kualiti persekitaran rumah responden mempunyai perkaitan dengan latarbelakang keluarga. Kedua-dua pendidikan dan pendapatan ibu bapa mempunyai perkaitan dengan kualiti persekitaran rumah. Kualiti persekitaran rumah juga didapati mempunyai perkaitan yang signifikan dengan prestasi mental kanak-kanak. Kanak-kanak daripada persekitaran rumah yang meransang kognitif memperolehi skor yang lebih tinggi di dalam semua sub-skala Weschler Preschool and Primary Scales of Intelligence daripada kanak-kanak lain dalam kajian.

ABSTRACTThe purpose of this study was to determine the reladonship between the quality of home environment that parents provide for their children and the children’s mental performance. The study focused on 64 children, aged 5 to 6 years, from the Universiti Pertanian Malaysia Laboratory Preschool. Results of the study show that, generally, most of the children came from a relatively supportive home environment. Consistent with past research, the quality of home environment correlated with family background. Both parents’ education and income correlated with the quality of the home environment. The quality of the home environment correlated significantly with the children’s mental performance. Children from a cognitively stimulating home environment achieved higher scores on all the subscales of the Weschler Preschool and Primary Scales of Intelligence.

INTRODUCTIONOf the various environmental factors that influence development during early childhood, the family and home environment represents one of the most significant factors in a young child’s life. Previous studies have provided evidence that a secure child who has had a harmonious and positive relationship with his parents tends to become a competent, responsible and productive adult (Belsky et al. 1984; Maccoby and Martin, 1983). Findings from recent studies also support the association between hom e environm ent and a child’s level of development during early childhood (Beckwith and Cohen, 1988; Bradley and

Caldwell, 1989; Bradley and Casey, 1984; Bradley etal.f 1989; Bullock and Pennington, 1988; Ninio and Rinott, 1988; Rozumah and Luster, 1992). These studies revealed that a home with appropriate tools for stimulation and a warm, facilitative relationship between parents and children made up the best combination that would encourage optimal learning in children.

Bee (1985) in her review of past studies summarized that the home environment of children who achieved higher IQ scores had appropriate play materials and parents who were democratic, emotionally and verbally responsive and had realistic goals for their children. A study by Rohani

* An earlier version of this paper was presented at the ‘Seminar Toward Enhancing the Quality of Life\ Universiti Pertanian Malaysia, Serdang, Selangor, 18th December 1989.


et al. (1989) reported similar findings. The mental abilities of the children were associated with the availability of appropriate play materials, variety of stimulation, organization of the physical environment and availability of language and academic stimulation within the home environment. Bradley^ a l (1989) reported a collaborative study which also provided similar evidence. The study found that measures of specific aspects of the ch ild ’s hom e environm ent such as parental responsitivity and availability of stimulating play materials were more strongly related to children’s cognitive development than global measures of environmental quality such as family socio-economic status. The importance of certain aspects of the home environment on children’s achievement was also investigated by Rozumah (1992). The study found that children who did well on achievement tests had mothers who provided more cognitively stimulating home environments. Children whose parents had a higher level of education and a higher family income were also found to perform better on the tests.

The effect of the early home environment is not transient. Longitudinal studies have indicated that children who had experienced a stimulating home environment during their childhood years tended to thrive better at schools (Bradley and Caldwell, 1984a; Bradley et a l , 1988a; Bradley et al., 1988b). The early home environment is therefore a valuable indicator of the learning environment available to the child early in life and the potential performance of the child when he enters formal schooling.

The present study was designed to determine the relationship between the quality of home environment and the children’s level of mental tests performance with three specific objectives in mind: 1) To determine the quality of the home environment available to the preschool child; 2) To determ ine the relationship between family background and the quality of the children’s home environment; and 3) To determine the relationship between the quality of the home environment and the children’s mental performance.

METHODOLOGYSample

Subjects for the study were 64 children aged 5 to 6 years from the Universiti Pertanian Malaysia

Laboratory Preschool. Most (90%) of the children were Malays, and the rest were Indians and Chinese. Fifty-two percent were males, and 48% were females.

Instrumenta. Home Inventory

The quality of the children’s home environment was assessed using the Preschool Version of the HOME (Home Observation for Measurement of the Environment) Inventory by a trained research assistant. The HOME has been shown to be a highly reliable and valid measure, and a strong predictor of child outcomes (Bradley and Caldwell, 1984a; Gottfried, 1984; Suriani, 1988). Suriani found an alpha coefficient of .89 for the HOME in her study on Malaysian children. The preschool HOME Inventory is an observation/interview technique that assesses the quality of physical, cognitive, social and emotional stimulation available to the child within the home environment.

The Preschool Version of the HOME Inventory (henceforth HOME) is composed of eight subscales. These subscales include: 1) toys, games and reading materials; 2) language stimulation; 3) physical environment; 4) pride, affection and warmth; 5) stimulation of academic behavior, 6) modeling and encouragement of social maturity;7) variety of experience; and 8) acceptance. Each item of the subscales is scored by giving 1 point for the presence of stimulation or 0 for the absence of stimulation for the child within the home. The total HOME is determined by adding scores of the 8 subscales of HOME. High scores indicate high quality home environment. Caldwell and Bradley (1984) reported an internal consistency of .93 for HOME. HOME was translated into Bahasa Melayu (the national Malaysian Language) and retranslated into English. A ninety-five percent agreement was calculated between the original and the retranslated versions.

b. Weschler Preschool and Primary Scales ofIntelligence (WPPSI).The WPPSI, designed to assess the intellec

tual abilities of preschool children, are divided into 10 major components: 1) information; 2) vocabulary; 3) arithmetic; 4) similarities; 5) comprehension; 6) animal house; 7) picture completion;


The Relationship between Quality of Home Environment and Mental Scores of Preschool Children

8) mazes; 9) geometric design; and 10) block design. Each component of the WPPSI is scored according to specific criterion as described in the manual by Weschsler (1963). Weschsler found a strong correlation between the WPPSI and the Stanford-Binet Intelligence Scale indicating that the WPPSI is a reliable measure of children’s mental performance. The first five components (1-5) are calculated to make up the Verbal IQ while the second group of five components (6-10) make up the Performance IQ. The scores of all ten components are combined to calculate the Full IQ. The WPPSI was administered in Bahasa Melayu.

Procedure

The WPPSI test was individually administered to the children in their preschool setting. A home visit was then arranged. The parents were allowed to choose the time at their own convenience. During the home visit, the research assistant conversed with the parents on topics related to the family background, the target child’s abilities and how parents usually spend their time with the target child. At the same time the research assistant would mark her checklist on the HOME Inventory, questioning the parents on the items which she could not observe herself. The visit lasted, on an average, about an hour. To determine the relationships between the independent and dependent variables the Pearson Correlation analysis was used.

RESULTSFamily Background

The education level of the respondents’ parents ranged from primary education to doctoral degrees in specific fields. Table 1 indicates that 22% of the children’s fathers had masters degrees and 24% had the Malaysian Certificate of Education. Twenty-three percent of the mothers had the Malaysian Certificate of Education. In addition, 81% of the mothers involved in the study were employed outside the home.

The parents’ monthly income ranged from below RM500 to above RM3000 (Table 2). Most (34%) of the fathers earned between RM500 to RM999 while most of the mothers (37%) earned less than RM500 per month. Almost one quarter of the fathers earned more than RM2000.

TABLE 1Percentage distribution of parents’ educational

background

Level of education Father MotherN % N %

Primary education 4 6.0 6 9.4Lower Certificate of Education/SRP

6 10.8 5 7.8

Malaysian Certificate of Education/SPM

16 24.1 15 23.4

Higher School Certificate/STPM

4 4.8 15 23.1

Diploma 2 3.1 4 6.3Degree 3 4.7 9 14.0Masters 14 21.8 10 15.6Ph.D 5 7.8 5 7.8Not available 10 15.6 8 12.5

64 100.0 64 100.0

TABLE 2 Percentage distribution of parent’s

monthly income

Father MotherIncome (RM) N % N %

Less than RM500 10 15.6 24 37.4500 - 999 22 34.4 16 25.01000-1499 10 15.6 5 7.81500-1999 4 6.3 5 7.82000 - 2499 7 10.9 8 12.52500 - 2999 3 4.7 1 1.6Above 3000 5 7.8 4 6.3Not available 3 4.7 1 1.6TOTAL 64 100.0 64 100.0

In terms of family size, a majority of the respondents came from a family of four to six members (Table 3). This is a relatively average family size, that is, a family of two to fo\ir children. The largest family size among the re*spond- ents was 10 persons.

TABLE 3Percentage distribution of family size

Family size N %

Less than 4 1 1.64 -6 56 87.5Above 7 7 / 10.9TOTAL 64 100.0



8) mazes; 9) geometric design; and 10) block design. Each com ponent of the WPPSI is scored according to specific criterion as described in the manual by Weschsler (1963). Weschsler found a strong correlation between the WPPSI and the Stanford-Binet Intelligence Scale indicating that the WPPSI is a reliable measure of children’s mental performance. The first five components (1-5) are calculated to make up the Verbal IQ while the second group of five components (6-10) make up the Performance IQ. The scores of all ten components are combined to calculate the Full IQ. The WPPSI was administered in Bahasa Melayu.

ProcedureThe WPPSI test was individually administered to the children in their preschool setting. A home visit was then arranged. The parents were allowed to choose the time at their own convenience. During the home visit, the research assistant conversed with the parents on topics related to the family background, the target child’s abilities and how parents usually spend their time with the target child. At the same time the research assistant would mark her checklist on the HOME Inventory, questioning the parents on the items which she could not observe herself. The visit lasted, on an average, about an hour. To determ ine the relationships between the independent and dependent variables the Pearson Correlation analysis was used.

RESULTSFamily BackgroundThe education level of the respondents’ parents ranged from primary education to doctoral degrees in specific fields. Table 1 indicates that 22% of the children’s fathers had masters degrees and 24% had the Malaysian Certificate of Education. Twenty-three percen t of the m others had the Malaysian Certificate of Education. In addition, 81% of the mothers involved in the study were employed outside the home.

The parents’ monthly income ranged from below RM500 to above RM3000 (Table 2). Most (34%) of the fathers earned between RM500 to RM999 while most of the mothers (37%) earned less than RM500 per month. Almost one quarter of the fathers earned more than RM2000.

TABLE 1Percentage distribution of parents’ educational

backgroundLevel of education Father Mother

N % N %Primary education 4 6.0 6 9.4Lower Certificate of Education/SRP

6 10.8 5 7.8Malaysian Certificate of Education/SPM

16 24.1 15 23.4Higher School Certificate/STPM

4 4.8 15 23.1Diploma 2 3.1 4 6.3Degree 3 4.7 9 14.0Masters 14 21.8 10 15.6Ph.D 5 7.8 5 7.8Not available 10 15.6 8 12.5

64 100.0 64 100.0

TABLE 2 Percentage distribution of parent’s

monthly incomeFather Mother

Income (RM) N % N %Less than RM500 10 15.6 24 37.4500 - 999 22 34.4 16 25.01000-1499 10 15.6 5 7.81500-1999 4 6.3 5 7.82000 - 2499 7 10.9 8 12.52500 - 2999 3 4.7 1 1.6Above 3000 5 7.8 4 6.3Not available 3 4.7 1 1.6TOTAL 64 100.0 64 100.0

In terms of family size, a majority of the respondents came from a family of four to six members (Table 3). This is a relatively average family size, that is, a family of two to fo\ir children. The largest family size among the re*spond- ents was 10 persons.

TABLE 3Percentage distribution of family size

Family size N %Less than 4 1 1.64 -6 56 87.5Above 7 7 / 10.9TOTAL 64 100.0



HOME and Mental ScoresTable 4 presents the results of the assessment of the quality of the children’s home environment. In comparison to Caldwell and Bradley’s (1984) report on the preschool version of the HOME Inventory, the respondents scored relatively high on three subscales of the HOME: verbal stimulation, academic stimulation and modeling of social m aturity. The m ean for the rest of the subscales (toys, game and materials; physical environm ent; warmth and affection; and acceptance) was about the same. However, the respondents scored much lower in the variety of experience subscale when com pared to the mean reported by Caldwell and Bradley (1984).

TABLE 4Mean scores and standard deviations

for pre-school home inventoryHOME Mean SDToys, games and reading 6.02 0.60Verbal stimulation 7.60 0.55Physical environment 5.97 0.23Warmth and affection 6.12 0.80Academic stimulation 4.66 0.67Modeling 3.92 0.86Variety of experiences 6.48 1.12Acceptance 3.18 0.59Total HOME (P) 43.95 3.69

The results of the children’s scores on the WPSSI test are shown in Table 5. As noted earlier, the WPPSI was used to assess the preschool child ren ’s intellectual abilities. The WPPSI results showed that the respondents had lower means for five of the subscales than those repo rted by Weschler (1967). With reference to Table 5, the respondents in this study had lower means for information, vocabulary, comprehension, animal house and picture completion. Mean scores for arithmetic, similarities, mazes, geometric design and block design did not differ much from those reported by Weschler (1967).

T he WPPSI test gen era ted th ree scores, namely, Verbal IQ Performance IQ, and Full IQ. A majority of the respondents (72%), obtained above average scores (110-129) on Performance IQ and a large proportion (48%) obtained above average scores on Full IQ (Table 6).

TABLE 5Mean scores and standard deviations of

subscales of WPPSISubscales Mean SDInformation 10.50 2.44Vocabulary 9.42 1.86Arithmetic 11.09 2.47Similarities 12.95 2.70Comprehension 9.05 2.25Animal house 10.64 2.65Picture completion 12.92 2.31Mazes 12.58 2.16Geometric design 12.75 2.63Block design 14.05 1.78

TABLE 6Percentage distribution of WPPSI IQ scores

IQ scores Verbal Performance Overall% % %

Below 90 (learning 6.3 4.7 3.1disability)

90-109 (average) 68.7 10.9 45.3110-129 (above 25.0 71.9 48.4

average)Above 130 (superior) 0.0 12.5 3.1

Relationship between HOME and Family BackgroundThe results of the study indicated that the total HOME and one subscale of the HOME ( toys, games, and materials) were significantly related to four of the family variables: father’s education, m other’s education, father’s income and m other’s income. Table 7 shows that toys, games and reading materials were significantly related to all the four variabes, i.e. father’s education (r =.75, p < .01), m other’s education (r = .64, p < .01), father’s income (r = .64, p < .01), m other’s income (r = .50, p < .01). The total HOME was related significantly to all the four family variables: Father’s education (r = .68, p < .01), m other’s education (r = .59, p < .01), father’s income (r = .57, p < .01) and m other’s income (r = .46, p < .01). Variety of experience was significantly related to father’s education (r = .40, p < .01) and acceptance was significantly related to m other’s income (r = .33, p < .01). O ther subscales of the HOME were also related to the family background variables;



TABLE 7Correlations between selected family background and preschool HOME

family background

HOMEFather’seducation

Mother’seducation

Father’sincome

Mother’sincome

Toys, games andmaterials .75*Verbalstimulation .14Physicalenvironment .09Warmth andaffection .12Academicstimulation .16Modeling .28Variety ofexperiences .40*Acceptance .27Total HOME (P) .68*

.64*.11

.09

.20

.27

.24

.25

.29.59*

.64*-.03.07.21

.17

.26

.32

.15.57*

.50*-.20

.12

.28

.16

.30

.11.33*.46*

* p < .01however, the relationships were not significant at the .05 level.

Relationship between M ental Scores and Family BackgroundC orrelational analyses com puted between the children’s WPPSI scores and family background p ro d u ced several sign ifican t findings. T he subscales of WPPSI tended to correlate significantly with m other’s education, father’s income and m other’s income (Table 8). M other’s education correlated significandy with vocabulary (r = .35, p < .01) and verbal (r =.33, p < .01). Mothe r’s income was found to correlate significandy with inform ation (r =.37, p < .01), vocabulary (r=.47, p < .01), animal house (r =.35, p < .01) and Verbal IQ (r =.40, p < .01). Father’s income correlated significandy with information (r = .34, p < .01), block design (r = .36, p < .01), and Verbal IQ (r = .38, p < .01). These findings suggest that the m ore educated the parents are, the more likely they will structure their home environment in ways that encourage children’s intellectual development. In addition, parents with higher income may be m ore able financially to provide educational resources that prom ote positive development in children.

TABLE 8Correlation between selected family background

factors and selected subscales of WPPSI

WPPSI

family backgroundMother’seducation

Father’sincome

Mother’sincome

Information .28 .34* .37*Vocabulary .35* .36* .47*Animal house .16 .22 .35Block design .31 .36* .28Verbal IQ .33* .38* .40** p < .01Relationship between HOME and Children's Mental PerformanceCorrelational analyses to determ ine the relationships between the quality of the children’s home env ironm en t and th e ir perfo rm ance on the WPPSI showed that three HOME subscales correlated significantly with WPPSI subscales. Table 9 shows that toys, games and materials correlated significandy with information (r =.45, p < .01), similarities (r =.36, p < .01),/block design (r =.33, p < .01), Verbal IQ (r =.42, p < .01) and Full IQ (r =.43, p < .01). Warmth and affection correlated



TABLE 9Correlations between selected subscales of WPPSI and Preschool HOME

WPPSI

HOMEtoys, games

and materialswarmth

and affectionacademic

stimulationTotal

HOMEInformation .45* .04 .21 .32Vocabulary .20 .47* .22 .28Similarities .36* .18 .04 .29Comprehension .31 .36* .34* .35*Picture completion .15 .34* -.01 .08Block design .33* .01 .11 .25Verbal IQ .42* .39* .23 .38*Full IQ .43* .34* .17 .30*p < . 01

significantly with vocabulary (r =.47, p < .01), comprehension (r=.36, p< .01), picture completion (r =.34, p < .01), Verbal IQ (r =.39, p < .01) and Full IQ (r =.34, p < .01). Significant correlations were found between stimulation of academic behaviour and comprehension (r =.34, p < .01), while the Total HOME correlated significantly with com prehension (r =.35, p < .01) and Verbal IQ (r =.38, p < .01).

DISCUSSIONThe findings of this study show that on the whole, the respondents do well on both measures of the HOME inventory and the WPPSI m ental tests. These results indicate that the children in this study are from families that provide a cognitively stimulating home environment. Consistent with the findings from other studies (Gottfried, 1984; Rozumah, 1992), children who experience a qualitatively supportive home environment tend to perform well on the WPSSI scales.

As was seen earlier, the respondents in this study came from a varied family background. Certain aspects of the HOME Inventory and mental test tended to associate significantly with the child ren’s family background. Toys, games and reading materials were related to both paren t’s education and parents’ income. The same family variables also related closely to the total hom e environm ent of a preschool child. Items observed in subscale of toys, games and reading materials included toys or games that taught colours, size, shape and numbers, puzzle, music, toys that per

mitted free experience and creativity, toys that necessitated refined m ovem ents, ch ild ren ’s books and family reading materials. The results suggest that parents with a higher education and a higher income have a greater awareness of the importance of stimulating materials; and having the means to purchase these materials, they were more likely to invest in these items.

T he results o f this study indicated tha t vocabulary and Verbal IQ of WPPSI were associated with m other’s education. This results supported findings of Appleton et al. (1975), Clarke- Stewart and Friedman (1977) and Yarrow et al. (1975) which showed th a t b e tte r educated m others were m ore likely to provide verbal stimulation and were more responsive to child ren’s verbal expressions than less educated parents. Their children, therefore, had more opportunities to develop verbal skills.

Verbal and Full IQofWPPSI correlated with toys, games and reading materials, warmth and acceptance and the total HOME environment of the preschoolers. O ur findings support results reported by Yarrow et al. (1975) which indicated that the variety of objects available to the children at hom e were closely associated with their scores on intelligence tests. O ther studies have also suggested that young children’s mental abilities are stimulated when they play with a variety of interesting things; when they are allowed to explore their surroundings; and when parents participate in their children’s games and activities (Bradley et al. 1977; Bradley and



Caldwell, 1984b; Price et a l 1981). Toys and play objects are only a part of the stimulation provided by parents. These objects become more m eaningful and have a greater potential to facilitate optimal learning if parents participate in the games; for example, by showing the child how to build a skyscaper out of the blocks, how to complete puzzles or to build a tractor out of the lasy parts.

Any home environment could be rich with “toys” that children can use to explore, experiment and learn, even if parents cannot buy specific toys for their children. For example, three to six year olds may enjoy helping parents sweep the floor, fold clothes, or even play with pots and pans in the kitchen. Parents can teach their children colours, shapes, sizes, numbers and words by using practically anything and everything that is available within the hom e environment.

It can be concluded that family background, the quality of the hom e environment that the parents provide for their children, and the children’s m ental perform ance are highly related to one another. Future studies could be directed to further exploration of the relationships between the HOME and WPPSI subscales, specifically those that were found to be non-significant in this study. Given that the family is a major setting for child ren ’s development, future researchers could also examine the influence of other family background variables such as num ber of children in the family, child birth order, and family structure on the quality of the hom e environment and child ren ’s mental performance.

REFERENCESA pp l e t o n , T., R. C lifto n and S. G o l d b e r g . 1975. The

development of behavioral competence in infancy. In Review of Child Development Research. ed. F.D. Horowitz. Volume 4. Chicago: University of Chicago Press.

B eckw ith , L. and S.E. C o h e n . 1984. Home environment and cognitive competence in preterm children during the first tive years. In Home Environment and Early Cognitive Development: Longitudinal Research. ed. A.W. Gottfried. New York: Academic Press.

B ee , H. 1985. The Developing Child. New York: Harper and Row.

B elsky, J., R. L erner, and G. S panier . 1984. The child in the family. Reading, MA: Addison-Wesley.

B radley, R.H. and B .M . C aldw ell. 1984a. The relation of infant’s home environments to achievement test performance in first grade: A follow- up study. Child Dev. 9(55): 803-809.

B radley, R.H. and B .M . C aldw ell . 1984b. 174 children: A study of the relationship between home environment and cognitive development during the first five years. In Home Environment, Social Status, and Early Cognitive Development, ed. A.W. Gottfried, San Francisco: Academic Press.

B radley. B.M. and B.M. C aldw ell and R . E la r d o . 1977. Home environment, social status, and mental test performance. J. Educ. Psyc. 69:697- 701.

B radley, R.H., B .M . C aldw ell, and S.L. R o ck . 1988a. Home environment and school performance: A ten year follow-up and examination of three models of environment action. Child Dev. 59:852- 867.

B r a d le y , R.H., B.M. C a l d w e l l , S.L. R o c k , H .M . H amrick , and P. H ar ris . 1988b. Home observation for measurement of the environment: Development of a Home Inventory for use with families having children 6 to 10 years old. Cont. Ed. Psy. 13:58-71.

B radley, R .H ., B .M . C aldwell, S.L. R ock, K.E. B ernard, C . G ary, M .A . H a m m o n d , S. M itchell , L. S iegel, C .T . Ramey, A.W. G o t tfr ied , and D.L. J o h n s o n . 1989. Home environment and cognitive development in the first three years of life: A collaborative study involving six sites and three ethnic groups in North America. Deuel. Psych.: 25:217- 235.

B radley, R.H. and P.M. C asey. 1984. Home environment of low SES non-organic failure-to-thrive infants. Child Dev.: 59:852-867.

B ullock , J.R. and D. R e n n in g t o n . 1988. The relationship between parental perception?'of the family environment and children’s perceived competence. Child St. J. 18:17-31.

C aldw ell, B .M . and R.H. B radley. 1984. Administration Manual (Revised Edition). Home Observation for Measurement of the Environment. Little Rock, Arkansas: University of Arkansas.

C larke-S tewart, A. and S. F riedm an . 1987. Child Development: Infancy Through Adolescence. New York: Wiley.


Rohani Abdullah, Siti Nor Yaacob and Rozumah Baharudin

G o t tfr ied , A. W. (Ed.). 1984. Home Environw£nt and Early Cognitive Development: longitudinal Research. London: Academic Press.

M accoby, E. E., and J. A. M ar tin . 1983. Socialization in the context of the family: Parent-Child interaction. In Handbook of Child Psychology. (4th ed.) Vol. 4. ed. P. Mussen. New York: Wiley.

N in io , A, and N . R in o t t . 1988. Fathers involvement in the care of their infants and their attributions of cognitive competence to infants. Child Dev.: 59:652-663.

P ed ersen , F.A., R. C a in , M. Za slo w , and B. A n d e r s o n . 1983. Variation in infant experience associated with alternative family role organization. In Families as Learning Environments for Children. Eds. L. Larse and I. Siegel. New York: Plenum.

P rice, G.G., R .D . H ess, and W .P. D ick so n , 1981. Processes by which verbal-educational abilities are affected when mothers encourage preschool children to verbalize. Devi. Psych. 17:554-564.

R o h a n i A., B. R ozum ah and S. B. Yew . 1989. Relationship between Home Environment and Children’s Mental Development. Paper presented at The Symposium of Children in Distress. Bangkok, Feb. 2-10.

R ozum ah B. 1992. Predictors of Maternal Behavior and their Effects on the Achievement of Children: Data from the Longitudinal Study ofYouth. Ph.D. Dissertation. Michigan State University, East Lansing, Michigan.

R ozum ah B. and T om L u st e r . 1992. Factors Related to the Quality of the Home Environment and Children’s Achievement. Paper presented at the National Council on Family Relations 54th Conference. Orlando, Florida, 4-11 November.

S uriani M. H. 1989. Penglibatan Ibubapa, Persekitaran Rumah dan Perkem bangan Kognitif Kanak-Kanak. Laporan Projek Penyelidikan Pelajar Tahun Akhir; Jabatan Pembangunan Manusia, Universiti Pertanian Malaysia, Serdang.

W esch ler , D. 1967. Manual for The Weschler Preschool and Primary Scale of Intelligence. (2nded.). Cleveland, Ohio: The Psychological Corporation.

Yarrow , L.J., J.L. R ubenstein and F.A. P edersen . 1975. Infant and Environment: Early Cognitive and Motivation Development. W ashington D.C.: Hemisphere.

(Received 22 January 1993)


P ertan ikaJ. Soc. Sci. 8c Hum. 2(1): 29-41 (1994) ISSN: 0128-7702 © Universiti Pertanian Malaysia Press

Factors Associated with Non-adoption of Technology by Rubber Smallholders

RAHIM M. SAIL and MAZANAH MU HAMAD D epartm ent o f Education, Faculty o f Educational Studies,


Keywords: non-adoption o f technology, rubber smallholders, extension workers, constraintsABSTRAK

Penerimaanguna ataupun penolakan teknologi oleh pekebun kecil adalah satu proses membuat keputusan dan perubahan yang kompleks. Kajian ini mengandaikan bahawa maklumat adalah punca bagi memudahkan proses membuat keputusan dan perubahan dan seterusnya boleh membawa kepada penerimaanguna teknologi di kalangan pekebun-pekebun kecil. Untuk menentukan tahap penerimaanguna teknologi di kalangan pekebun kecil dan faktor-faktor yang berkaitan dengannya, kajian ini telah memilih sejumlah 787 pekebun kecil, ketua dan pemimpin masyarakat pekebun kecil dan pegawai-pegawai RISDA barisan hadapan di Semenanjung Malaysia. Kajian ini m enggunakan teknik perbincangan kum pulan terfokus (focus group discussion) untuk mengumpulkan data. Secara keseluruhannya, tahap penerimaanguna teknologi di kalangan pekebun kecil getah adalah rendah. Ada beberapa faktor yang menyokong penemuan ini, antaranya adalah: sikap pekebun kecil terhadap teknologi baru, masalah kewangan, limitasi fizikal (kebun yang kecil dan kebun yang terpencil yang susah untuk dikunjungi), ketidaksuaian setengah teknologi bagi sektor pekebun kecil, pemilikan tanah dan perkhidmatan pengembangan dan sistem sokongan komunikasi yang lemah. Tidak terdapat satu faktor yang dominan, malahan semua faktor ini saling mempengaruhi penyaluran maklumat untuk memudahkan proses membuat keputusan dan perubahan di kalangan pekebun kecil bagi menerimaguna teknologi. Kekangan- kekangan yang dihadapi oleh pekebun-pekebun kecil merupakan satu rangkaian kuasa yang mempengaruhi secara negatif terhadap penerim aanguna amalan-amalan yang disyorkan. Maklumat dari penyelidikan pengem bangan seharusnya meliputi aspek-aspek ‘principle’, ‘bagaimana’ di samping maklumat-maklumat ‘lingkungan’ dan pembangunan sumberdaya manusia. Maklumat dari pengembangan kepada pekebun kecil seharusnya memfokuskan kepada ‘bagaimana’ dan faedah-faedah yang boleh didapati dari menerimaguna sesuatu teknologi.

ABSTRACT

This study argues that information holds the key to decision-making and change processes, and hence technology adoption among smallholders. To determine the levels of technology adoption among smallholders and factors associated with it, a total of 787 smallholders, key informant smallholders and field officers from selected areas in Peninsular Malaysia were interviewed in groups using the focus group discussion technique. Generally, technology adoption level among rubber smallholders was relatively low. Several factors accounted for this finding, among them were: smallholders attitudes towards new technology, financial constraints, physical limitations (uneconomic holding size and accessibility of holdings), labour shortage, inappropriate technology, land ownership and weakness in extension service and communication support system. These factors were found not to act singly but in various combinations which compounded the problem of information flow decision making and change among smallholders in technology adoption. The constraints faced by the smallholders form negative forces that act against the adoption of recommended practices. Some of these barriers can be overcome or minimized with the provision and implementation of a systematic flow of information from research to extension and from extension to smallholders. Information from research to extension should cover ‘principle’, ‘how-to’ information, besides relevant circumferential and human resource development information. InformaKon from extension to smallholders should focus mainly on ‘how-to’ and benefits to be accrued from adopting a technology.

INTRODUCTIONAdoption or rejection of a technology by clients is a complex process. It involves a dynamic decision-making process often influenced by deliberate plans or strategies made by change agents as well as the perceived positive or negative attributes of the technology. Adoption of a technology requires some form of structural, functional as well as behavioural changes on the part of the clients. These changes must not be perceived by clients

to be disruptive but must harmonize with existing values and practices. More importandy, these changes must translate into practical ways of solving clients’ problems and needs. In other words, adop tion of a technology requires clients to change - a complex process but not an impossible goal to achieve, as long there is a firm commitment from all parties concerned to ensure that the complex process of change happens smoothly and systematically.

Rahim M. Sail and Mazanah Muhammad

On the other hand, rejection of a technology by clients involves a similar process of decisionmaking except that it may be suspended temporarily until a final decision is made; in such cases, adequate information on the technology is unavailable or more time is required to fully understand the technology and the implications of its adoption. In cases where clients fully understand the technology but decide not to adopt it, the technology may be perceived to have negative attributes or clients are not convinced that the technology will solve their problems and fulfil their needs; or they may perceive the technology to have a minimal effect on existing practice in terms of benefits to be accrued. Whatever their reasons for rejecting a technology, they follow a process of decision-making which Rogers and Shoemaker (1971) term “The Innovation-Deci- sion Process”. It is a mental process through which an individual passes from first knowledge of an innovation to a decision to adopt or reject and to a confirmation of this decision (Rogers and Shoemaker, 1971).

Assuming a technology reaches the client, adoption or rejection of the technology passes through two basic processes: (1) decision-making and (ii) change. Both processes are critical in technology adoption or rejection as they motivate individuals to act when they acquire knowledge about the need for change. The provision of knowledge and awareness and the stimulation of interest to change are usually done by change agents. Clients are persuaded and encouraged to evaluate the technology in the socio-cultural and economic context of their situation to ensure that the technology is more advantageous than the existing practice (relative advantage), compatible with the norms and values of society, easy to understand (not complex), easy to be tried out on a small scale ( triability), and its perform ance is easy to be observed for comparison with existing technologies.1

Natural rubber production is a major industry in Malaysia, contributing 2.3% of Malaysia’s total export value of primary commodities (Economic Report 1993/94). Rubber is produced both by the estate and smallholder sector. Although occupying 82% of the total planted area, production by the smallholder sector was very much less than that of the estate. For example, in 1992 the

average yield per hectare for the sm allholder sector was alm ost 30% less (Rubber Statistic Handbook, 1992). One of the factors associated with higher production by the estate was good agronom ic and processing practices. A few studies had been carried out to determ ine the adoption rate of rubber technology by the smallholders (Sulaiman Yassin et al, 1985, Raja Badrul Shah et al, 1986a and 1986b). Although these investigations revealed that the technology adopti on in the smallholder sector was generally low, specific constraining factors associated with technology adoption have no t been examined in detail. This study was undertaken to fill that informationgap-Objectives of the StudyThe general objective of the study was to determine the constraints faced by rubber smallholders in adopting a technology especially factors related to socio-cultural, economic, physical and technological attributes. Specifically, the study aimed to determine:

1. levels of technology adoption among smallholders; and

2. factors associated with technology adoption and rejection.

METHODOLOGYData for the study were collected from four main sources: (i) the smallholders; (ii) key informant smallholders and field officers; (iii) field observation; and (iv) office records.

The smallholders were grouped into three categories: (i) those having immature holdings (trees less than 7 years); (ii) those having mature holdings (trees 7 -2 0 years); and (iii) those having old holdings (trees more than 20 years). The rationale for grouping the smallholders according to age was that some technologies were unique to each category. For example, deep planting technology was applicable to the immature holdings while opening tapping panels was applicable to m ature holdings, and control upward tapping (cut) was applicable to old holdings.

Key informants were smallholder leaders and field officers. Smallholder leaders comprised progressive farmers, Village Development and Secu-

1. For a detailed discussion on each attribute of an innovation (technology), see Rogers and Shoemaker, 1971:135-155.



rity Committee Members, religious and social leaders and other influential individuals in the community. The Rubber Industry Smallholders Development Authority (RISDA) officers at district and sub-district (mukim) levels were key informants for officers.

Field observation was carried out mainly for verification purposes after the data had been collected through interviews, focus group discussions and office records. Data from office records were obtained from district officers of RISDA and the Rubber Research Institute of Malaysia (RRIM).

Location and Subjects of the Study For the purpose of this study, Peninsular Malaysia was divided into four regions namely, the northern region, the eastern region, the southern region and the central region. Within each region, areas with a heavy concentration of smallholders were identified and selected. The following areas were selected: Baling, Grik, Lenggong, Selama and Taiping for the n o rth e rn region; Tanah Merah, Machang, Besut and Setiu for the eastern region; Segamat, Muar and Jasin for the southern region; and Sepang, Seremban and Temerloh for the central region.

The main subjects of this study were the independent rubber sm allholders1 selected from the regions cited earlier. The selection of the independent smallholders was carried out with the help of the local RRIM and RISDA staff.

Field officers of RISDA and smallholder leaders were also the subjects of the study as they served as key informants on smallholders’ problems and needs. It was assumed that these key informants were knowledgeable about rubber cultivation, constraints facing the crop and the smallholders and the com m unity in which they themselves worked and lived.

Data Gathering TechniquesThe study relied heavily on the rapid rural appraisal techniques using semi-structured interviews a n d /o r discussions. Focus Group Discussion (FGD) was used to gather in-depth qualitative data from smallholders, smallholder leaders and field officers. A focus group comprising eight to 10

smallholders with a hom ogeneous background and interests was formed. For example, a group of smallholders with immature holdings formed a focus group. A similar focus group was formed for mature and old holdings.

U nder the guidance of a trained research facilitator, smallholders in each FGD group were encouraged to discuss and in teract with each other, to comment on the major themes presented to them, to question researchers and to respond to each o th e r’s com m ents on the issues that emerged out of the discussions. In each FGD, there was at least one biologist, one extension/ com m unication specialist and one process observer who also acted as a recorder.

The FGD used prepared guidelines on major and m inor themes for discussions. In this study, critical technologies for immature, mature and old holdings were prepared to guide the discussion. Also prepared were major extension/com m unication themes especially those that have a direct b ea rin g on d issem ination tech n iques, p ro grammes of smallholders development and constraints facing the extension system.

Focus G roup Discussion was also used to gather information from the key informants. In these groups, general themes such as level of technology adoption among smallholders, problems associated with technology adoption, strengths and weaknesses of extension services, etc. were discussed.

Field observation verified inform ation gathered through FGD and office records. Selected holdings of smallholders in the identified areas were visited to see whether or not certain technologies were practised and to note the conditions of the general surroundings especially those related to infrastructure, planting of other crops and other economic activities carried out by smallholders.

Technology adoption and rejection by smallholders was the focus of data analysis. Reasons for rejecting a technology were isolated and screened and this was done for the three types of holdings i.e., immature, mature and old holdings. In all cases, percentages and means were used for the analysis.

-------------------------------------------------- /1. Independent rubber smallholders refer to the “unorganized” and scattered smallholders under the jurisdiction of RISDA. The term “independent” is used to denote unorganized versus organized smallholders of FELDA and FELCRA.



RESULTSThe following outlines the status of adoption of some recom m ended practices by smallholders at different holding stages.

Immature HoldingsNew and proven technologies, when applied during the immature years of rubber, would result in the enhancem ent of productivity and efficiency when rubber entered its mature stage. Responses from 66 FGDs, involving a total of 173 smallholders with immature rubber, are summarized in Table 1.

When land clearing is carried out manually, any tree stumps left behind should be poisoned with 2, 4, 5-T or Garlon 250. The cut surface should immediately be treated with creosote to prevent root disease spore colonization. This study indicated that only 17% of respondents practised stump poisoning after felling the trees. Many respondents were not aware that root diseases affecting young rubber were caused by rotted stumps of roots that were not poisoned.

The contribution of legume cover crops in controlling weeds and the returning of nitrogen to the soil has been widely accepted. In the im

TABLE 1Percentage adoption of recommended technology by smallholders

Holdingstage

recommended technologyNorth

* (n=16)%

% adoption East South

(n=16) (n=16)% %

Central(n=16)

%Mean

%Immature 1. Stump treatment 8 12 24 22 17

2. Planting legume cover 14 16 18 13 153. Rock phosphate application 86 100 94 100 954. Sulphur application 6 0 10 10 75. Deep planting 1 9 21 16 126. Weed control7. Fertilization frequency

21 41 70 36 42year 1 0 0 0 0 0year 2 year 3 -5

0 0 0 0 013

8. Control pruning 20 13 22 10 169. Root disease treatment 24 25 36 42 32

Mature 1. Chemical weed control 50 60 76 64 622. Fertilization (rate) 40 77 51 64 583. Root disease treatment 20 24 20 14 204. Yield stimulation5. Tapping aids

2 0 0 0 1RRIMGUD 0 2 3 0 1JUS Knife 0 0 0 0 0

Old 1. Fertilization (rate) 59 42 5 NA 352. Pest and disease treatment 29 19 20 10 203. Yield stimulation4. Tapping aids

12 0 5 2 5RRIMGUD 0 0 0 0 0JUS Knife 0 0 0 0 0Motorray 0 0 0 0 0

*n represents num ber of focus group discussions carried out in the district. Each focus group comprised 8 -10 smallholders.NA - not available



mature holdings studied, establishing and maintaining the common legume co’/er crops such as Calopogonium caeruleum, Centrosema pubescens and Pueraria phaseoloides was practised by only 15% of the respondents. To supplement their incomes, the majority of the smallholders planted cash crops. R esearch ca rrie d o u t by RRIM on intercropping showed that intercrops did not affect rubber growth if adequate fertilizers were used.

To prom ote root growth at planting, rock phosphate mixed in the planting hole is recommended. As rock phosphate was normally provided in the replanting programme, this practice was carried out by the majority (95%) of the respondents. The use of sulphur for prevention of root disease was limited to only 7% of the respondents. Sulphur was not easily available in the local market and neither was it supplied in the replanting programme.

An increase in the yield of rubber as the tapping panel approaches the stock-scion union could be made possible by the absence of the union through deep planting. This technique also provides firm ground anchorage for the tree. Despite its advantages, deep planting was only adopted by 12% of the respondents. Many were not aware of the technology; others complained that extra labour was required to dig the deeper holes. Where holes were dug mechanically, the technology incurred extra costs.

Proper and scheduled manuring practices are necessary as adequate amounts of nutrients are required to maintain a high level of growth and productivity. The frequency of application recommended by RRIM is 7 times during the first year, followed by 5 applications in the second year and 3 times each in the third and fourth year respectively. The frequency is further reduced to only twice a year in the following years.

Smallholders are advised to plant legumes in their immature holdings. Under such a practice, the frequency of weed control is not fixed. If the legume cover crop is not established, weeding should be done before fertilizing about 3 times a year, using the strip, circular or interrow methods. This recom m ended frequency by RRIM was practised by 42% of the respondents.

The frequency of fertilizer application was reported to be far less than that recom m ended. None exercised the recom m ended frequency

from the first to second year while 13% fertilized their third to five years old rubber three times a year. The m anuring frequency did not change much with age of trees; one or two applications per year was the usual practice adopted. According to the respondents, fertilization practice was based on their own judgem ent and observation and advice from friends or neighbours. Advisory service on proper m anuring application schedules was minimal.

Research findings conducted by RRIM have shown th a t low branch ing trees have h igher growth rate and girthing than high branching trees. Better growth and thus girthing rate can be induced by low and controlled pruning methods as opposed to estate pruning. This study dem onstrated that only 16% of the respondents practised controlled pruning.

The most common disease in young rubber is white root disease which can cause widespread damage if it is not controlled effectively. The disease was reported by 46% of the respondents. Only 32% of the smallholders with holdings affected by the disease treated the problem (either by root exposure, or root exposure and chemical treatm ent).

Mature HoldingsThe status of technology adoption for mature rubber was determ ined through 64 focus group discussions involving 178 smallholders who operated mature holdings in the fifteen districts sampled. The technology adoption level of this category is presented in Table 1.

Based on RRIM recommendations, m ature rubber should be fertilized at least once a year at recom m ended rates, based on soil foliar analysis. The practice of m anuring mature rubber at 1 - 2 times a year was reported by 86% of the respondents. The rate of application varied from, 1 bag per hectare to 8 bags per hectare without due consideration to the num ber of tappable trees and soil types. Only 58% of the respondents applied 2 - 4 bags per hectare of fertilizer to their holdings. The field visits showed that the rubber planting density varied from as few as 200 tappable trees to 370 tappable trees/ha. Thus, if smallholders practised what they reported, it would mean that in some cases they used insufficient fertilizer while in other cases they used fertilizers excessively.



To facilitate tapping, application of fertilizer and minimizing competition with the crop for nutrients, weed control is recom m ended. When chemical is used, it should be applied at the frequency of 1 - 2 times per year. About 62% of the smallholders studied practised chemical weeding; 24% used a combination of chemical and manual weeding and 14% relied solely on manual weeding. When manual weeding was adopted, it was carried out minimally, sufficient enough to facilitate tapping operations. Safety precautions when weedicide was used were not stressed by extension agents. A partial adoption of the technology was attributed to insufficient subsidies for weedicide, financial constraints and a dearth of labour due to old age.

To increase latex production, smallholders are advised to practise yield stimulation. Unaware of the technique and its benefit, almost all operators (99%) of mature holdings did not adopt the technology. Similarly, non-utilization of tapping aids such as RRIMGUD and JU S knife has been altributed to poor advisory service or the less frequent visits to the mature holdings by extension agents.

Old HoldingsA total of 178 smallholders of old rubber were interviewed in 64 focus group discussions to determine the status of adoption of technologies recom m ended for old holdings. The responses obtained are shown in Table 1.

Fertilization was carried out only when cost of fertilizers was subsidized. Only about 30% of smallholders applied fertilizer to their old holdings. When fertilizers were used rate, m ethod and frequency of application differed among holdings. Only some (35%) fertilized their holdings at the rate of 2 - 4 b ag s/h a /year as recom m ended by RRIM.

Old rubber were reported to be affected by a num ber of diseases which include white root disease, pink disease, secondary leaf fall, panel disease and brown bast. Among these, the most common was white root disease (reported by 69% of the respondents). Termites and wild aminals were two pests identified by the sm allholders (7%). Control measures against both pests and diseases were relatively low (20%). Non-adoption of pest and disease control was attributed to unawareness of the problem, not knowing appropriate techniques

for treatment, unavailability of fungicide/pesticide locally, exhorbitant price of some chemicals and the techniques involved were too laborious (for root disease treatm ent).

In relation to tapping, no tapping aids such as RRIMGUD and JUS knife were used. The practice of yield stimulation was adopted by only about 5% of the smallholders.

Factors Associated with Non-Adoption of Technology from the Perspective of Key informant SmallholdersTo gain a better insight into the reasons for nonadoption of recom m ended technology, selected sm allholder leaders were interviewed. Focus group discussions with these informants were carried out in 11 districts: Baling, Grik, Taiping, Tanah M erah, Besut, Segam at, Muar, Jasin , Temerloh, Seremban and Sepang. Table 2 summarizes the perceptions of the key informants for the low level of technology adoption among rubber smallholders. The factors were grouped into the following sub-headings:

AttitudinalInformants from all the districts rationalized that often a recom m ended technology was not adopted when the practice called for laborious and time consuming work. For example, they reported many smallholders did not carry out root disease treatm ent by trenching due to the above factor.

Another often (91% of the focus group) cited reasons that have to do with the availability of subsidy. Many smallholders indicated that a recom m endation like fertilizer and pesticide application was practised only when the relevant inputs were subsidized. To quote a respondent from Jasin “smallholders generally do not apply fertilizer or chemical unless subsidy is available”.

The focus group interview in nine of the districts (see Table 2) revealed that some smallholders were not interested in any technology recom m ended to them . N either were they interested in attending courses, demonstrations, etc. which were carried out to deliberate on the technology. Others showed some interest in a certain recom m ended practice. However, being doubtful about the expected benefits and concerned with costs involved, these smallholders decided not to adopt the technology.



TABLE 2Factors associated with non-adoption of technology from

the perspectives of key informant smallholders

Atdtudinal

boGgCQ

ibe S•S -s .9- I £ (2 UPQ <uC/3

JC'u<uS£

Gcd£S<L>cr>

beGnCuCJC/3

**

Preference for less labourious and time x X X X X X X X X X X 100consuming work

Dependency on subsidy X X X X X X X X X X 91Not interested in the technology X X X X X X X X X 82‘Wait and see’ attitude X X X X X X 56Used to ‘old’ practice X X X 27

PhysicalUneconomic size of holdings x X X X X X X X X 82Scattered holdings X X X X 36Hilly holdings X X X 27Accessibility of holdings X X X X X 45Technology not available locally X X X X X 45

LabourOld age X X X X X X X X X X 82Insufficient labour X X X X 36

Holdings OwnershipMulti-ownership X X X X 45Absentee landlord X X X X 36

FinancialTechnology recommended expansive X X X X X X X X X X X 100Insufficient money to purchase input X X X X X X X X 73

Appropriateness of Technology X X X X X X X 65Extension Service 8c Communication Support

Not aware of technology X X X X X X X 64Insufficient information on technology X X X X X X X X 73Lack of advisory service X X X X X 45

% : Percentage of response “yes” of the total num ber of districts, x : Responses given by the districts.

Lack of interest in a recom m ended practice seems to be related to the contribution rubber makes to a smallholder’s income. When rubber was not the main source of income, or the income derived from rubber could not m eet family requirem ents, then sm allholders were less concerned with investing in a recom m ended technol

ogy. This was m entioned by smallholders in seven districts visited by the research team.

The rate of technology adoption was also affected by the smallholder’s inclination to see the impact of a practice before trying it out. This ‘wait and see’ attitude was reported in six districts.



Informants from Grik, Taiping and Temerloh reported that many smallholders had been operating their holdings in the same m anner for many years. Over time they had adopted certain practices which have become an acceptable m ethod of managing their holdings. Familiarity with the ‘old’ practice was one of the reasons cited for nonadoption. For example, it had been an “accepted” practice for many smallholders in Taiping not to fertilize their mature and old rubber. W hether or not fertilization was done made little difference in terms of impact to rubber yield or their income.

PhysicalSeveral physical problems were associated with non-adoption of technologies (see Table 2). The main physical constraint identified by the key inform ant smallholders (from 9 out of 11 districts studied) was uneconomic holding size. They felt that small holding size rendered it uneconomical for smallholders to adopt certain high cost technologies such as tapping aids, weedicide and pesticide treatm ent and fertilizer application.

Remoteness and inaccessibility of some of the holdings due to hilly terrain, swamp and river barriers prevent sm allholders from practising technology such as mechanical land preparation. The cost of bringing a tractor from town to a remote holding would inflate the capital cost. Hilly holdings also faced water accessibility problems. As water is required to dilute pesticide, fungicide and weedicide, a lack of it affects the adoption of those practices (Grik, Besut and Temerloh).

Informants from Baling, Grik, Tanah Merah, Besut and Temerloh reported that some chemicals and tapping aids were not available in their areas. Unavailability of 2, 4, 5-T and Garlon 250 in the local stores affected stump treatm ent practices. Similarly, difficulty in acquiring sulphur, herbicide and fertilizers limits adoption of practices that call for their inputs.

LabourA ccording to the m ajority o f the inform ants (see Table 2), most of the smallholders were old. They associated advancing age with lack of strength to practise intensive labour like weeding, pest and disease treatm ent and thus the low adoption of these practices.

The younger generation showed little interest in operating smallholdings. Labour shortage forced many smallholders (Tanah Merah, Besut, Temerloh, Sepang) to resort to hired labour or to leasing their holdings to other operators. Hired labour and leased operators were more interested in maximizing profit from the holdings. Thus, they were often not interested in practices which might raise expenses financially or physically. Neither did they care much about the long term effect of cultivation practices on the rubber trees.

Holdings OwnershipA sizeable percentage of holdings was owned by m ore than one owner. Multiple ownership of holdings was reported (by informants from live districts) - (see Table 2) to be a deterrent to technology adoption.

Technology adoption was also affected by absentee landlords (Besut, Segamat, Seremban, Sepang). In the absence of owners, hired workers concentrated more on the latex extraction than on maintaining the holdings.

FinancialThe majority of the informants (from all the districts studied) attributed non-adoption of a technology to financial constraints. According to them, smallholders could not afford a technology due to two main reasons:

(i) recom m ended implements like RRIMGUD, Motoray, and chemicals (weedicide and fungicide) were too costly (reported by informants from all the districts studied).

(ii) Insufficient money to purchase inputs (reported by informants from 8 districts).

Appropriateness of TechnologyThe appropriateness of a recom m ended technology had been questioned by a num ber of the informants from seven districts. They viewed tapping implements such as Motoray and RRIMGUD, and JUS Knife to be inappropriate as most smallholders have limited resources and operate their holdings on a small scale basis. Based on their observation and experience, respondents from Besut, Seremban and Dengkil claimed that con



trolled upward tapping (CUT) resulted in tappers suffering neck pain. Motoray was reported in Besut and Seremban to be damaging to the tree while yield stimulants caused the bark to burst.

Extension Service and Communication Support To many informants (from 7 districts - see Table2), smallholders failed to adopt a technology when they were not aware of it. This was especially so in the case of non-adoption of tapping aids, and pest and disease control.

Even though smallholders were aware of a technology, insufficient information on the practice m ight discourage them from adopting it (Baling,Grik, Tanah Merah, Besut, Segamat, Jasin, Temerloh and Sepang). For example, most smallholders did no t know that controlled pruning could induce the plant to grow faster. The lack of information on the advantage of control pruning made them believe that it was an obstacle to getting a smooth cylindrical tree trunk in the future, thus discouraging them from doing it. Similarly, it was reported that smallholders did not adopt yield stimulation technology because they associated it with the final procedure before the old rubber trees were due for replanting within a year or two.

Some of respondents reported that they had heard of the availability of tapping aids such as Motoray from the television o r /a n d extension agents. Some even had the advantage of observing the change agents demonstrate how it worked. However, even after attending the m ethod demonstration, many sdll did not use the aid. They complained that they had litde opportunity to try the implement themselves, and were thus not sure of its practicality.

A nother factor often cited as being responsible for low technology adoption was inadequate communication support strategies. To these informants (Table 2) the weakness and lack (in certain areas) of the communication support strategies prevented smallholders from practising a recommended technology. For example, although fertilizers were given to smallholders at the replanting stage, advice on correct timing, frequency and rate of application of the fertilizer was minimal. Such smallholders had to resort to friends and fertilizer dealers, or use their own judgem ent.

A num ber of respondents complained about the difficulty in meeting extension agents, particularly after their rubber reached maturity. The frequency of the extension agents’ visits was closely associated with the age of the rubber trees. Most visits by extension agents were made during the immature stage (replanting grant period). More often these visits have to do with the inspection of the holdings for replanting grants disbursement rather than for educational and technology transfer purposes. As the trees reached maturity, the frequency of visits by extension agents declined until finally there was minimal extension activities once the trees were 20 years old and more.

Since many respondents associated technology adoption with the quality of extension service, they were encouraged to share their view on the agents’ competency. Many felt that the extension personnel skill and knowledge on current rubber technology were considerably low. Similar observations were made by the respondents on the change agents’ skill and knowledge on extension m ethod and delivery techniques.

Factors Associated with Non-adoption of Technology from the Perspective of Key Informant OfficersTo further understand the rationale for the relatively low level of technology adoption among the rubber smallholders, focus group discussions were also conducted with selected field officers in the areas studied. These interviews revealed that from their perspective, non-adoption could be attributed to various reasons. The factors identified by them are almost similar to those provided by the key inform ant smallholders. The reasons could be classified into five categories namely, attitudinal, physical, labour, ownership, and financial.

AttitudinalThe smallholders were perceived to be contented with their old practices and were not keen on new ways of m anaging their holdings. Many were doubtful about the appropriateness and benefits of the technology. Others feared the risk involved in practising new m ethods. T heir preference for less labour and less time-consum ing work and dependence on subsidies (e.g. fertilizers,



chemicals) were the other main factors identified by the informants as hindrances to adoption of a technology.

PhysicalSize, location and accessibility of a holding affected the rate of technology adoption. The officers reported that about 60% of the holdings were less than two hectares. The small holdings made it uneconom ical for many smallholders to adopt some technologies. Remoteness and inaccessibility of holdings, because of hilly terrain and swampy areas, further hindered technology adoption. The absence of a local oudet where smallholders could purchase inputs (for those who could afford it) was another factor for non-adoption of technology.

LabourPoor technology adoption rate had often been associated with smallholders’ age. According to the informants, the majority of smallholders (60% - 70%) were more than 50 years old. To assist them with farm work, these smallholders employed hired labour or leased their holdings to o ther operators. Technology adoption rate in holdings utilizing hired labour or leased workers was reported to be much lower.

Land ownershipThe officers estimated that a sizeable percentage (18% - 20%) of holdings had multiple ownership. It was observed that adoption of new practices in these holdings was slower. A similar phenom enon was reported in cases where the owner stayed far away from the holding. In the absence of the owner, it was extremely difficult to organize activities for transfer of technology.

FinancialMany inform ants felt that financial constraint was another major obstacle in technology adoption. This constraint was due to the nature of smallholders and their holdings; being relatively poor and drawing a marginal income from their small sized holdings. With limited cash, relevant inputs required for the adoption of some practices become too expensive for many smallholders to afford.

To provide a comprehensive picture on the factors associated with non-adoption, this study utilized three sources of data; the smallholders, key inform ant smallholders and key inform ant officers. Table 3 summarizes these factors.

The three groups of respondents identified six common problems related to the adoption of a recom m ended technology: physical, labour, holdings ownership, financial, appropriateness of technology and attitudinal. Although key infonn- ant smallholders and officers identified a longer list of attitudinal problems (see Table 3), the smallholders themselves reported only two. Constraints related to extension and communication support were identified only by the smallholders and key inform ant smallholders.

DISCUSSIONDecision-making in technology adoption is a complex process and requires an individual to select from a series of choices based on information h e /sh e receives. These series of choices are based on information on technology attributes, cost, labour, and appropriateness of a technology or they could be circumferential factors such as credit, market price of product and land tenure situations. For a ru b b er sm allholder, these choices becom e extrem ely difficult especially when h e /she is constrained by inadequate labour, in a d e q u a te fin an ce , sm all-sized o p e ra tio n , remoteness of holding, and inadequate information on the benefits of a technology. This is further aggravated by the fact that the outcome of adopting a technology in rubber cultivation usually takes from six m onths (effect of fertilizer) to 10 or 15 years (yield perform ance of a clone) and this extends the period of uncertainty and therefore, increases the complexity and difficulty in making decisions.

Closely related to the physical and circumferential factors is the attitude of rubber smallholders towards adopting a technology. This study identified several factors related to attitude of smallholders that affect decision-making whether to adopt or to reject a technology. Examples of these attitudinal factors are smallholders’ dependence on subsidy for adoption of a technology, smallholders preference for less laborious and less time-consuming work and ‘wait and see’ attitude.



TABLE 3Factors associated with non-adoption of technology from the perspectives of smallholders,

key informant smallholders and key informant officersPerspectives

FactorsKey Key

Smallholders informant informantsmallholders officers

AttitudinalPreference for less laborious and time consuming work ✓ S ✓Dependency on subsidy V V ✓Not interested in technology S ✓‘Wait and see’ attitude SUsed to ‘old’ practices S ✓Unwilling to take risks ✓

PhysicalUneconomic holding size S V ✓Scattered holdings S V ✓Hilly holdings S S SAccessibility of holdings V V STechnology/inputs not available locally s S S

LabourOld age ✓ V SInsufficient labour ✓ V S

Holdings OwnershipMulti-ownership V SAbsentee Landlord s V V

FinancialTechnology recommended expansive s S VInsufficient money to purchase input s V V

Appropriateness of technology S S

Extension service and communication supportNot aware of technology s SInsufficient information on technology s VLack/insufficient advisory service s V

The physical, circumferential and attitudinal factors combined to compound the complexity of the decision-making process and technology adoption among rubber smallholders.

What is portrayed here is no t a totally hopeless situation, but rather a reality facing the rubber smallholder sector which provides cues for research and extension organizations to plan and im plem ent strategies to overcom e such constraints.

Adoption or rejection of a technology by clients is a conscious and deliberate effort to change from existing practice to a new one. Therefore, a decision to adopt a technology is a decision to change. For a change to occur, clients must be provided with as much information as is possible so that they are fully aware, able to diagnose and evaluate the change according to their situations. With rubber smallholder^, information dissemination becomes a critical prerequisite for tech


Rahim M, Sail and Mazanah Muhammad

nology transfer. The information is usually relayed by the extension service.

With a gamut of external and internal factors surrounding a change situation facing the smallholders, the extension service has to be competent to disseminate information. And this is not an easy task especially when extension agents themselves also face numerous constraints in their effort to disseminate technology inform ation. This study has found that an agent spends most of his time on enforcem ent and administrative activities rather than on extension activities (advisory service, m ethod dem onstration). Examination on the extension agents’ working schedule revealed that only about one-fifth of their time was spent on extension activities (Rahim M. Sail et al. 1990).

Change is a slow and complex process similar to that of the decision-making process. With the constraints faced by smallholders and the extension agents, change becomes even slower and more complex and this accounts for some of the delays and rejection o f technologies recom m ended to rubber smallholders through the extension service. In line with this, we can establish a strong premise that it is not only the smallholders’ knowledge, beliefs and practices in technology that need to be changed, but also those of the extension agents. For example, this study showed th a t m any ex ten s io n ag en ts n e ed ed to be equipped with knowledge and skills of current technologies as well as effective dissemination strategies. Change is, therefore, basically to reduce restraining forces while at the same time to increase prom otional forces that affect both the smallholders as well as the extension service.

RECOMMENDATIONSSeveral recommendations could be forwarded for the consideration of research and extension organizations to further improve existing practices in technology transfer and adoption. These recommendations are as follows:

1. Training of trainers programmes (from research to extension) should cover all technology attributes in detail besides discussing relevant circumferential (e.g. costs, markets and credits) factors. The ‘how-to’ and ‘principle- in form ation’ should be the focus in these

training programmes as this would enhance the processes of decision-making and change in technology transfer.

2. In developing technologies for smallholders, attributes such as simplicity, cost and appropriateness to smallholders’ farming situation and cultural practices should be given priority and careful consideration by research organizations.

3. Publication m eant for extension agents and smallholders should be simple in the form of ‘how-to’ to follow steps about a technology. If inputs are required, information on cost, how and where the inputs can be obtained should be furnished.

4. In situ training and lecture/discussion should be followed by dem onstration of ‘how-to’ to increase effectiveness of technology transfer. Ideally, the dem onstration o f ‘how-to’ should be repeated a num ber of times by the demonstrator. The participants should try out the skills themselves with close guidance by the demonstrator. This process is slow but effective as it reinforces knowledge, skills and builds up confidence in the participants about a new technology. This is applicable to both processes of information flow from research to extension as well as from extension to clients.

5. Extension effort (particularly for smallholders with m ature holdings) has to be increased in order to increase the level of technology adoption among smallholders. This could be done through reducing the administrative and enforcem ent functions of field extension officers while at the same time increasing field extension effort of supervisory officers.

6. In-house training programmes of extension organizations should include, besides the detail technical aspects of current practices in rubber cultivation, HRD aspects (e.g. problem solving skills, decision-making skills, leadership skills and motivational skills) which would put the process of technology transfer in a proper perspective am ong extension agents. This is also applicable for smallholder training program m es. The inculcation of HRD aspects is im portant to ensure an overall success of technology transfer and technology adoption.



7. Inputs for adoption of a new technology must be made available in local areas and where inputs are expensive, they must be made available on easy term credits to smallholders.

8. To increase the level and effectiveness of technology adoption among smallholders, extension effort should be directed towards group approach as in the group replanting programmes. The group approach would overcome some of the constraints faced by individual smallholders who operate small-sized farms.

REFERENCESA h m a d H a s s a n O sm a n and G.S. T a n . 1988. The

Progress and Development of Rubber Smallholders. In Proceedings of the Sixth Seminar, Palembang, Indonesia. Kuala Lumpur. The Association of Natural Rubber Producing Countries, Malaysia.

A lang P. Za in u d d in and R ahim M. Sa il . 1988. Extension work on processing and marketing of smallholders’ rubber: an opportunity for assessing staff performance and commitment. In Proceedings of the Sixth Seminar, Palembang; Indonesia. Kuala Lumpur. The Association of Natural Rubber Producing Countries, Malaysia.

Economic Report. 1993/94. Kementerian Kewangan, Malaysia.

R a h im M . S a il , R . B a d r u l S h a h R .S .K ., A . G h a n i Ibrahim , A . F ua d M u h a m m a d , M a z n a h M uh a m m a d , M . H anim M . T a h ir , M . Y u so f M . N o o r and W. M a n so r W. Sal l e h . 1990. From Smallholder's Practicesy Constraints and Needs to Appropriate Technol

ogy Generation, Development and Adoption: The Case of Rubber Smallholdings in Peninsular Malaysia. Kuala Lumpur. Rubber Research Institute of Malaysia and Universiti Pertanian Malaysia.

Raja B adrul S hah Kobat , L eo ng M .W . and Srri M ariam . 1986a. The Adoption of Technology on Immature Independent Rubber Smallholdings. Rubber Research Institute, Kuala Lumpur.

Raja B a d rul S hah K o bat , L e o n g M u n W ai and S uarti M a n so r . 1986b. The Adoption of Technology on Mature Independent Rubber Smallholdings. Rubber Research Institute, Kuala Lumpur.

R og ers , E. M. and F. F. S hoem aker . 1971. Communication of Innovation: A Cross Cultural Approach (2nd Edition). New York. The Free Press.

Rubber Statistic Handbook. 1992. Jabatan Perang- kaan Malaysia, Kuala Lumpur.

S ulaim an M o h d . Ya ssin , R ahim M d . Sa il , A lang P erang Za in u d d in , M o h d . Fadzilah Kam sah , B ahari Yatim , N apsiah M ah fo z , A hm ad F uad M u h a m a d , B aham an A b u Sam a h , M o h d . A riff H u sse in , M o h d . G hazali M o ha yiddin , Zainal A bidin K idam , W ilfred P. L eo n G uerrero a n d Srn K h air o n M o h d . S hariff. 1985. A Review of Extension Activities for Rubber Smallholder, Malaysian R u b b e r Research a n d D e v e lo p m e n t B o a r d a n d U n iv e r s it i P e r ta n ia n Malaysia, K u a la L u m p u r .

Zaltm an , G. 1973. Processes and Phenomena of Social Change. New York. John Wiley and Son.

(Received 19 April 1993)

Pertanika J. Soc. Sci. & Hum. Vol. 2 No. 1 1994

/

41


The Stability and Predictability o f Betas: Evidence from the Kuala Lumpur Stock Exchange

SHAMSHER MOHAMAD and ANNUAR MD. NASSIR D epartm ent of Accounting and Finance Faculty of Economics and Management

Universiti Pertanian Malaysia 43400 UPM Serdang, Selangor Darul Ehsan, Malaysia.

Keywords: Beta, stability, portfolio, securities, expected returnsABSTRAK

Beta mengukur risiko bersistem atau risiko tidakboleh di pelbagai sesuatu sekuriti. Para pelabur inginkan nilai beta yang stabil (yang boleh diramal) untuk membolehkan mereka menilai pulangan teijangka atas pelaburan mereka dengan tepat. Nilai beta yang tidak stabil akan menyebabkan nilai pulangan teijangka kurang tepat dan seterusnya tidak dapat menunjukkan prestasi sebenar sesuatu pelaburan. Kajian ini meniliti soal kestabilan dan kebolehramalan beta Fowler-Rorke (beta ini mengambilkira soal ketipisan dagangan kebanyakan saham-saham di Bursa Saham Kuala Lumpur) bagi 148 firma yang tersenarai di BSKL. Penemuan kajian ini menunjukkan bahawa beta saham individu dan portfolio adalah pegun mengikut masa. Seperti dijangka beta portfolio adalah lebih stabil dari beta saham individu dan cara pembentukan portfolio mempengaruhi kestabilan beta portfolio. Kestabilan beta portfolio tercapai dengan sekurang-kurangnya 15 saham atau lebih, tidak kira cara mana portfolio itu dibentuk. Secara keseluruhannya, penemuan ini mencadangkan bahawa tahap kebolehpercayaan terhadap beta saham individu dan portfolio yang dianggarkan di BSKL tidak boleh diperkecilkan (diendahkan) oleh para pelabur apabila membuat keputusan pemilihan portfolio dan pelaburan.

ABSTRACTBeta measures the systematic or undiversifiable risk of a security. Investors desire stable ( and hence predictable) measures of beta to enable them to accurately estimate the expected returns on their investment. Instable betas lead to inaccurate estimates of expected returns over time and hence provide misleading signals on performance of investments. This study examines the stability and predictability of the three leads/lags version of Fowler- Rorke betas (unlike OLS betas, these betas address the problem of thinness of trading peculiar to the KLSE) of 148 firms listed on the Kuala Lumpur Stock Exchange (KLSE). The findings suggest that the beta of both individual securities and portfolios are quite stationary over time. As expected the portfolio betas are relatively more stable than individual securities betas. Furthermore, the method of portfolio formation affects the relative portfolio beta stability. However, portfolio beta stability is achieved with 15 or more securities, irrespective of method of portfolio formation. Overall, the findings indicate that investors can reliably utilise estimated individual security and portfolio betas for their portfolio selection and investment decisions.

INTRODUCTIONSince the introduction of the systematic risk coefficient or beta in Capital Market Theory (Sharpe, 1964) as an im portant stasdstic in estimating the returns on assets and consequently in the making of investment decisions, considerable effort has been expended on obtaining empirical estimates of betas (Wallace, 1980). It is generally accepted that the total risk of an asset consists of diversifiable and non-diversifiable portions. Diversifiable risk is attributed to factors which are specific to the asset and can be eliminated through diversifica

tion. Non-diversifiable risk is due to factors which influence all assets in the market and hence constitute the only relevant risk for each asset. Investm ent theory suggests that beta is the appropriate m easu re o f risk fo r d iversified p o rtfo lio s (Jensen, 1968) and the efficient m arket theory suggests that the am ount of return above the riskfree rate an investment manager can expect depends solely upon beta or the sensitivity of the investment’s return to the changes in market returns. Sharpe (1964) and Lintner (1965) developed the Capital Asset Pricing Model (CAPM) to

Shamsher Mohamad and Annuar Md. Nassir

explain the relevant risk and returns on assets in the market. The CAPM states that the risk premium for any asset is related to the risk premium of the m arket expressed in the following way;

E (R )= R r +B i(E(Rm) - R r)

Where E(R) is the expected return on asset i,Rf is the risk-free rate,Bj is the beta coefficient.

The importance of beta as a tool for making investment decisions has been increasingly recognised and some local brokerage firms regularly provide information on the beta coefficients of a large num ber of stocks listed on the Kuala Lumpur Stock Exchange. Institutional investors desire to have the best possible estimates of ongoing or possibly changing betas of competitive funds to develop investment strategies relative to expected perform ance of their portfolios and those of their competitors’ in the market cycle ahead. These estimates will allow inferences to be made with respect to their competitors expected market outlook. For example, a gradually increasing fund beta would indicate a bullish outlook on the part of the competitor. A stable and predictable beta estimate will also enhance the validity of the beta- based investment performance ranking tools such asTreynor’s (1965) and jen sen ’s (1968) perform ance measures. The portfolio m anager’s ability to select securities is reflected in the sum of returns above or below expectations as defined by beta. If beta values are not accurate measures of risk, they will provide misleading signals about the portfolio m anager’s performance.

The estimation of betas is im portant for understanding the risk-retum relationship in a thinly traded market. The investment community desires an efficient m arket (at least in a semi-strong sense), as it will help to optimally allocate scarce resources (Fama 1970). To ascertain the efficiency of the market, the information-content or event- type of studies in finance and accounting rely on the ordinary least squares (OLS) beta to forcast returns. Excess returns based on the OLS beta estimates then form the basis for testing the capital market effects of various accounting information or signals. Under this approach, any fluctuation in beta is captured in the OLS residuals, a

characteristic that adversely affects the efficiency estimates and the power of the tests. The stability of the beta is im portant in this case, and if the beta is not stable, then appropriate adjustments need to be made to the beta to enhance the power of conventional capital market tests of information content. The concern over the accuracy of beta is its non-stationarity beyond some acceptable level.

Though there is evidence that beta is a robust description of how investors analyse risk under conditions of uncertainty and how they expect their portfolios to perform under these conditions (Sharpe and Cooper, 1972), Fama and French (1992) observed that the positive relationship between OLS beta and the average returns on US stocks observed by Black, Jensen and Scholes (1972) and Fama and MacBeth (1973) does not seem to hold anymore, and the book-to- market value of equity is a better explanatory variable of the cross-section of average returns. However, the existing tests cannot conclusively indicate whether this anomaly results from a faulty single-index asset pricing model a n d /o r a persistent mispricing of securities; hence market professionals and academics still use beta as a measure of market risk or market sensitivity. However, as long as securities continue to move in line with the broad market movements, portfolio managers must be concerned about the sensitivity of their portfolios to the general market, the stability of that relationship and the accuracy of the investm ent tools they utilise. The use of beta coefficients in security portfolio analysis and investments strategy selection will only be of value if betas are stable and thus predictable.

The OLS technique is usually used to estimate beta as the regression coefficient of a simple linear regression based on data of at least four to six years to ensure relative stability of the beta (Alexander and Chervany (1980)). However, the OLS beta for common stocks has been found to be unstable over time (Levy (1971), Sharpe and Cooper (1972)). The KLSE is a thinly-traded market and as such the beta in the market model is estimated with a serious bias leading to a non-synchronous trading problem (Ariff and Lim, 1989; Annuar and Shamsher, 1991). This problem arises when the market index at time t is based on stocks whose closing prices do not synchronise at t. Conse


The Stability and Predictability of Betas: Evidence from the Kuala Lumpur Stock Exchange

quently, any estimate of return or systematic risk on a thinly-traded share is strictly not comparable with another thinly-traded share or with continuously traded shares. Scholes-Williams (1977), Dimson (1979) and Fowler-Rorke (1983) suggest corrections to the OLS beta estimate to mitigate this non-synchronous trading problem.

The objective of this study is to examine the stability and predictability of the three leads/lags version of Fowler-Rorke betas of firms listed on the Kuala Lum pur Stock Exchange (KLSE). The Fowler-Rorke betas are used as there is evidence (Ariff and Lim, 1989) that these betas efficiently mitigate the thinness of the trading problem more effectively with the three leads/lags version.

The rest of the paper is organised as follows: Section 2 describes the data and the distribution of betas while Section 3 discusses the stability of individual security betas. Section 4 presents the stability of portfolio betas and Section 5 evaluates the predictability of betas. Section 6 concludes the paper.

DATA AND DISTRIBUTION OF BETASThis study covers the monthly returns of 148 ordinary stocks traded on the KLSE over the period from January 1975 to December 1989. The stability and predictability of the betas are studied over three 5-year periods: January 1975 to December 1979; January 1980 to December 1984; and January 1985 to December 1989, and two 7-year periods: January 1976 to December 1982 and January 1983 to December 1989. The Fowler-Rorke (FR) betas are estimated as suggested by Fowler-Rorke(1983)1 . The equally-weighted index of all listed stocks (compiled by the authors) on KLSE is used as a proxy for market index. The profile of the FR beta coefficients of the sampled firms for the 5 and 7-year estimation intervals are summarised in Table 1.

For both the 5 and 7-year estimation intervals there were negative betas; the minimum beta for the form er interval is -5.36 and for the latter interval -7.18. The mean of all betas is 1.00 for the 75/79 period with a standard deviation of 1.15,

whereas the mean beta for the 76/82 period is 0.96 with a standard deviation of 1.10. To ascertain the distribution of betas, the sample was categorised into six risk classes according to their beta values in ascending order. The frequency distribution is summarised in Table 2.

For the 5-year estimation interval, less than 40 stocks (27%) for periods 1 and 2 and 72 stocks (48.6%) for period 3 had beta values of less than 0.3, whereas 26 stocks (17.5%) of periods 1 and 2 and one stock of the third period had beta values greater than 1.8. The num ber of stocks outside the beta range of 0.3 to 1.8 was small. A similar distribution of betas was observed for betas in the 7-year interval.

STABILITY OF SECURITIES BETASThe stability of beta has received considerable attention in the literature of financial economics. Baesel (1974) provides evidence that the stability of beta is dependent upon the estimation period length. However, Theobald (1981) dem onstrated that beta stability does not increase indefinitely with the estimation period length, thereby implying an optimal estimation period. This study examines the stability of beta estimates for individual stocks and portfolios on the KLSE over five-year and seven-year periods using the transition matrix and product m om ent correlation. If the betas are stable and hence predictable, then investors can assess the future riskiness of their investments from past riskiness. The correlation analysis measures the strength of co-movement of the two variables (betas in this case) as well as test of significance2.

To test the stability of betas over different lengths of estimation period, the rates of return were first calculated for each stock and the market index. Then for the 15-year sample period (1975-1989), beta coefficients were calculated for the consecutive 5 years resulting in 3 betas per stock. Two betas per stock were calculated for two consecutive 7-year estimation intervals. The stocks were then ranked into different risk classes and the num ber of stocks that remained in the same risk class over the various estimation intervals were

‘The procedures to estimate the Fowler-Rorke (FR) betas are explained in Ariff and Johnson [1990].2The significance of correlation coefficient is evaluated using the F-statistics [Ramanathan, 1992 pp. 105-106].



TABLE 1Profile of the estimated FR beta coefficients for the sampled firms

Period Mean Std.Dev.

Min.Value

Max.Value

Std.Error

Variance

5-year Interval1 /7 5 -1 2 /7 9 1.00 1.157 -5.357 4.351 0.095 1.3391 /8 0 -1 2 /8 4 0.482 0.746 -1.100 8.426 0.061 0.5571 /8 5 -1 2 /8 9 0.621 0.658 -2.080 3.716 0.054 0.433

7-year Interval1 /7 6 -1 2 /8 2 0.960 1.100 -7.180 3.499 0.090 1.2101 /8 3 -1 2 /8 9 0.436 0.787 -3.241 7.334 0.065 0.619

TABLE 2 Frequency distribution of betas

Risk Class Beta Period 1 Period 2 Period 3Midpoint 75/79 80/84 85/89

Frq. Cum. Frq. Frq. Cum. Frq. Frq. Cum. Frq.(Low Risk)

1 0.2 40 40(27%) 40 40(27%) 72 72(48.65)2 0.4 18 58(39.2%) 18 58(39.2%) 63 135(91.2%)3 0.8 27 85(57.4%) 27 85(47.4%) 9 144(97.3%)4 1.2 25 110(74.3%) 25 110(74.3%) 2 146(98.6%)5 1.6 12 122(82.4%) 12 122(82.4%) 1 147(99.3%)6 2.0 26 148(100%) 26 148(100%) 1 148(100%)

(High Risk)7-year Interval

Risk Class Beta Period 1 Period 2Midpoint 76/82 83/89

Frq. Cum. Frq. Frq. Cum. Frq.(Low Risk)

1 0.2 29 29(19.56%) 38 38(25.67%)2 0.4 35 64(43.24%) 32 70(47.29%)3 0.8 31 95(64.14%) 28 98(66.21%)4 1.2 21 116(78.37%) 27 125(84.45%)5 1.6 18 134(90.54%) 13 138(93.24%)6 2.0 14 148(100%) 10 148(100%)

estimated. Tables 3 and 4 summarise the findings for five and seven year estimation intervals respectively.

The findings in Table 3 show that, on average, more than 50 percent of the stocks in the risk class one and 29 percent in risk class two remained in the same risk class over the three periods, whereas 7 percent (periods 1 and 2), 19 per

cent (periods 1 and 3) and 56 percent (periods 2 and 3) of the stocks in risk class three remained in the same risk class over the three periods. In general, there is substantial stability over the 5- year estimation interval in the low risk class memberships of individual securities. Similar results were observed for betas in the 7-year estimation interval (Table 4).



TABLE 3Stability of FR betas for individual securities (5 years)

Risk Beta Class Midpoint

Period 1 and 2 75/79-80/84

Number and % in the same risk class

Totalnumber

ofstocks

Period 1 and 3 75/79-85/89


Totalnumber

ofstocks

Period 2 and 3 80/84-85/89


Totalnumber

ofstocks

1 0.2 25(63%) 40 21(53%) 40 38(53%) 722 0.4 9(50%) 18 5(28%) 18 18(29%) 633 0.8 2(7%) 27 5(19%) 27 5(56%) 94 1.2 - 25 3(12%) 25 - 25 1.6 - 12 - 12 - 16 2.0 - 26 2(7%) 26 - 1

TABLE 4 Stability of FR betas for individual

securities (7 years)RiskClass

BetaMidpoint

Period 1 and 2 76/82-83/89

Number and % in the

same risk class

Total number of stocks

1 0.2 19 (66%) 292 0.4 11 (31%) 353 0.8 3 (10%) 314 1.2 - 215 1.6 - 186 2.0 - 14

STABILITY OF PORTFOLIO BETASSeven portfolios were formed each consisting of 21 randomly selected stocks. Table 5 shows the risk-class membership of the seven portfolios for the various betas. The findings show that irrespective of the estimation interval, portfolios in a lower risk class were more stable than those in a higher risk class. The risk class membership of portfolios is more stable than individual securities, as securities moving to higher classes are offset by those moving to lower classes.

PREDICTABILITY OF BETASThe reliability of beta coefficients as a tool for investment analysis is observed only if betas can be estimated with a certain degree of accuracy from historical values, and used as a reliable indicator for forecasting future beta values. The degree of association between past and future beta values is measured by product-moment and rank-order correlation coefficients3. The higher the association, the more reliable are the historical betas as estimates of future beta values. The product-moment and rank-order correlation analysis of the beta values of the sampled firms for the 5-year (Panel A) and 7-year (Panel B) estimation intervals are presented in Table 6. The results in Panel A indicate that there is significant rank-order and product- m om ent correlation between beta values for the three periods. Similar results were observed in Panel B. This implies that the beta coefficients of the 148 stocks could be predicted from those of the previous periods.

To ascertain the predictability of portfolio betas, the product-moment and rank-order correlation coefficients of beta values of 7 portfolios were observed. Each portfolio consists of 21 stocks randomly chosen in the first period (historical betas) with their corresponding values in the following period (future betas). The correlation coefficients are summarised in Table 7. The results show that

3For a discussion on product moment and rank order correlation, see Siegel [1956].



TABLES Stability of beta portfolios

Risk Beta Period 1 and 2 Period 1 and 3 Period 2 and 3Class Midpoint Number Total Number Total Number Total

and% number and % number and % numberin the of in the of in the ofsame portfolios same portfolios same portfoliosrisk risk riskclass class class

5-year estimation interval1 0.2 2 (50%) 4 2 (50%) 4 2 (67%) 32 0.4 2 (100%) 2 1 (100%) 1 1 (100) 13 0.8 - 1 1 (50%) 2 2 (67%) 3

7-year estimation interval76/82 - 83/89

1 0.2 1 (100%) 12 0.4 2 (50%) 43 0.8 2 (100%) 2

TABLE 6Product-moment and Rank-Order Correlation of security betas

(Prob > R under Ho: 4. = 0, N = 148 individual securities)Rank-Order Correlation Product-Moment Correlation

Panel A: 5-year estimation intervalPeriod 2 Period 3 Period 2 Period 3

Beta Beta Beta BetaPeriod 1Beta 0.395 0.450 0.342 0.406

(0.001) (0.002) (0.004) (0.002)Period 2Beta - 0.107 - 0.248

(0.031) (0.023)Panel B: 7-year estimation interval

Period 2 Period 2Beta Beta

Period 1 0.353 0.204Beta (0.001) (0.035)

there is significant correlation between portfolio betas in the various periods. However, these portfolio betas had a lower degree of stability compared to betas of individual securities. These findings are inconsistent with those of Blume (1971) and Levy

(1971) whose data were based on developed markets.

This inconsistency could be due to the use of Fowler-Rorke betas adjusted for thinness of trading in this study instead of Ordinary Least



TABLE 7Product-Moment and Rank-Order Correlation of Portfolio Betas

(Prob > R under Ho: = 0, N = 7 portfolios)Panel A: 5-year estimation interval

Rank-Order Correlation Product-Moment CorrelationPeriod 2 Period 3 Period 2 Period 3

Beta Beta Beta BetaPeriod 1Beta 0.286 0.357 0.084 0.096

(0.001) (0.001) (0.072) (0.068)Period 2 0.035 0.359Beta (0.261) (0.011)

Panel B: 7-year estimation interval —

Period 2 Period 2Beta Beta

Period 1 0.081 0.178Beta (0.069) (0.05)

Squares (OLS) betas, as the OLS portfolio betas (not shown in the text) showed a higher degree of stability than those of individual securities. This is to be expected since portfolio betas encountered less serious non-synchronous trading problems (Annuar and Shamsher, 1991).

Porter and Ezzell (1975) and Alexander and Chervang (1980) provide evidence th a t the m ethod of portfolio formation and level of diversification affect the stability of portfolio betas. In view of this evidence, portfolios consisting of 5, 10, 15 and 20 stocks were formed both randomly and by a ranking procedure based on securities beta coefficients over the five and seven-year estimation intervals. For the ranked portfolio formation, individual stocks were ranked in ascending order of their beta coefficients, and sequentially n num ber of portfolios were formed. The portfolio beta coefficients were calculated as the average of the betas of stocks in the portfolios. After determ ining the portfolio’s beta coefficients for the succeeding 5 and 7-year periods, the product- m om ent correlation, rank-order correlation and mean absolute deviations were computed.

Table 8 presents the descriptive summary of the 5 year (Panel A) and 7 year (Panel B) sets of portfolio betas. The findings show that the stand

ard deviation (SD) of the randomly formed portfolio beta coefficients in Panel A and Panel B decrease in value as the num ber of securities in the portfolio increases. A similar observation for the ranked portfolios is not as apparent. The decrease in SD is most noticeable when the portfolio size increases to 15 securities in Panel A and 10 securities in Panel B with relatively little improvement thereafter. These findings imply that the beta stability of random portfolios increases with the increase in the num ber of securities, although the rate of increase decreases after 15 (10) securities for the 5 (7) year estmation intervals. This shows that shorter term portfolio betas are more stationary with a larger num ber of securities in the portfolio.

Table 9 presents the correlation and mean absolute deviation values of portfolio betas from alternative measures of portfolio formation for the three 5-year periods (Panel A) and the two 7-year periods (Panel B). The findings in Panel A and B show that both the rank-order (RO) and product-m om ent (PM) correlations increase as the portfolio size is increased for the ranked portfolios, suggesting greater beta stability for m ore diversified portfolios. The alternative



m easure of beta stability, m ean absolute de- random portfolios, consistent with the findings in viation, suggests a decrease in value as the size of developed markets (Porter and Ezzel 1975, andportfolios is increased for both the ranked and Blume 1975).

TABLE 8Descriptive summary of beta coefficients

Panel A: 5-year intervalEstimation Number of Number of Random Rankedinterval securities portfolios portfolios portfolios

Mean SD Mean SD1975-1979 5 29 0.973 0.311 1.102 0.743

10 14 0.874 0.232 1.071 0.70015 9 1.128 0.200 1.115 0.67920 7 1.153 0.192 1.071 0.713

1980-1984 5 29 0.530 0.338 0.491 0.32610 14 0.387 0.114 0.500 0.21315 9 0.444 0.089 0.493 0.16620 7 0.495 0.166 0.500 0.164

1985-1989 5 29 0.429 0.281 0.632 0.29710 14 0.825 0.186 0.642 0.23415 9 0.500 0.163 0.613 0.21720 7 0.647 0.129 0.641 0.216

Panel B: 7-year estimation interval1976-1982 5 29 0.902 0.311 1.105 0.743

10 14 0.758 0.232 1.071 0.70115 9 0.914 0.211 1.115 0.67920 7 0.036 0.201 1.071 0.715

1983-1989 5 29 0.454 0.287 0.441 0.36710 14 0.536 0.164 0.413 0.23715 9 0.257 0.153 0.427 0.21820 7 0.465 0.132 0.423 0.135

TABLE 9Stability of portfolio betas

Panel A: 5-year estimation intervalEstimation portfolio number of correlation mean absolute

interval formation securities deviationmethod in portfolio RO PM

1975/1979to 1980/1984 Ranked 5 0.193 0.000 0.917

10 0.129 0.032 0.88915 0.317 0.057 0.87420 0.393 0.067 0.829

Random 5 0.454 0.238 0.56410 0.604 0.629 0.48715 0.251 0.417 0.68320 0.287 0.084 0.458



Table 9 (continued)1975/1979to 1985/1989 Ranked

Random

1980/1984 Rankedto 1985/1989

Random

5 0.415 0.384 0.76910 0.490 0.491 0.78615 0.667 0.778 0.75320 0.643 0.526 0.669

5 0.549 0.525 0.57510 0.112 0.192 0.27415 0.783 0.813 0.62920 0.357 0.096 0.406

5 0.631 0.461 0.29210 0.719 0.586 0.27815 0.817 0.764 0.24420 0.893 0.800 0.235

5 0.127 0.189 0.30310 0.068 0.181 0.43915 0.183 0.264 0.16120 0.036 0.359 0.257

Panel B: 7-year estimation interval1976/1982to 1983/1989 Ranked

Random

5 0.267 0.049 0.79010 0.490 0.364 0.72015 0.450 0.299 0.73120 0.643 0.684 0.6625 0.589 0.233 0.511

10 0.442 0.308 0.26915 0.300 0.096 0.65720 0.081 0.178 0.501

CONCLUSIONFor a sample of 148 securities traded on the Kuala Lum pur Stock Exchange (KLSE), the findings of this study indicate that the beta coefficients as a measure of systematic risk are relatively stationary over time. For example, there is stability of beta coefficients over the 5 and 7 year estimation periods in the low risk class membership of individual securities. The risk class membership of portfolios is relatively more stable than individual securities when this measure of stability is applied.

When standard deviation of beta coefficients is used as a measure of stability the results show that the beta stability of randomly formed portfolios increases with increase in the num ber of securities, although the rates of increase decrease after 15 securities for 5-year betas and 10 securities for the 7-year betas.

Using alternative measures of beta stability such as product-moment and rank order correlations the results indicate that beta coefficients of ranked portfolios are relatively more stable and are related to the num ber of securities in the portfolio and the converse is true for randomly selected portfolios. On average, there is substantial stability between betas of individual securities for the three 5-year and two 7-year periods. However, there is a lower degree of stability for portfolio betas. The use of correlation coefficient as a measure of beta stability masks the possibility of decrease in the magnitude of intertem poral changes in portfo lio beta coefficients as the num ber of securities in the portfolio increases, regardless of the m ethod of portfolio formation.

Therefore, the mean absolute deviation statistic can be used to measure time stability of beta



coefficients. The results show that time stability of portfolio beta coefficients are directly related to the num ber of securities in the portfolio and are significantly stable for portfolios of 15 or more securities, irrespective of the m ethod of portfolio formation.

Overall, the results indicate that the individual securities and portfolio betas are relatively stable and can be reliably used for portfolio selection and investment decisions.

REFERENCESAnnuar, M.N and M. Shamsher. 1991. Estimation of

unbiased systematic risk measures: The case of Kuala Lumpur Stock Exchange (KLSE). M ’sian Mgt Rev. 26(2): 35-50.

Ariff, M. and L.W. Johnson. 1990. Securities Markets and Stock Pricing: Evidence from a Developing Capital Market in Asia. Singapore. Longman Singapore Publishers (Pte) Ltd.

Ariff, M. and K M. Lim. Methods selection for correcting non-synchronous trading bias in beta risk of stocks. Working paper presented at Finance and Banking Seminar, National University of Singapore.

Alexander, G.J. and N.L. Chervany. 1980. On the estimation and stability of beta. J. Finan. and Quant. Analys. 15: 123-137.

B a e s e l , J .B . 1974. On the assessment of risk: some further considerations. J. Finan. 29: 1491-1494.

Black, F, M. Jensen and M. Scholes. 1972. The Capital Asset Pricing Model: Some Empirical Tests. In Studies in the Theory of Capital Markets, ed. M.C. Jensen, p. 79-121 New York. Praeger.

Blume, M.E. 1971. On the assessment of risk .J. Finan. 26: 1-10.

D im so n , E. 1979. Risk measurement when shares are subjected to infrequent trading./. Finan. Econs. 7: 197-226.

Fam a, E. F. 1970. Efficient market markets: a review of theory and empirical work. J. Finan. 25(2): 383-417.

Fam a, E.F. and KR. F r e n c h , 1992. The cross-section of expected stock returns. / Finan. 47: 427-465.

Fama, E.F. andj. M c B e th . 1973. Risk, return and equilibrium: empirical tests. / Polit. Econ. 71: 607- 636.

F o w le r , D.J. and C.H. R o rk e . 1983. Risk management when shares are subjected to infrequent tradin g ./ Finan. Econs. 12: 279-289.

J e n se n , M. 1968. The performance of mutual funds in the period 1946-1964./. Finan. 23: 389-419.

L evy, R.A. 1971. Stationary of beta coefficients. Finan. Analysts. J. 27: 55-62.

L inter , J. 1965. The valuation of risky assets and the selection of risky investments in stock profiles and capital budgets. Rev. Econs and Stats. 47 :1S- 37.

P h u o n g , T.H. and W.T. W e n g . 1986. Stability and Predicability of Betas. In Investment Analysis and Management, ed. Saw See Hock and Lim Choo Peng. Singapore: Longman Singapore (Pte) Ltd.

P o rter , R .B . and J R . E zzel. 1975. A note on the predictability of beta coefficients./ Bus. Res. 3: 367- 372.

R a m a n a th a n , R . 1992. Introductory Econometics with Applications. (2nd Edn.). Dryden Press.

S iegel, S . 1956. Non Parametric Statistics for the Behavioural Sciences. Tokyo. McGraw-Hill Kogakusha Ltd.

S h arpe , W.F. 1964. Capital asset prices: a theory of market equilibrium under conditions of risk ./ Finan. 19: 42S442.

S c h o les , M and J. W illiam s. 1977. Estimating betas from non-synchronous data. J. Finan. Econs. 5: 309-328.

S h a rp e , W.F. and G.M. C o o p e r . 1972. Risk-return classes of New York Stock Exchange common stocks. Finan. Analy.J. 38: 46-50.

T h e o b o l d , M. 1981. Beta stationarity and estimation period: some analytical results. / Finan. and Quant. Analys. 16: 747-757.

T reynor,J . 1965. How to rate management of investment funds. Harv. Bus. Review. 63-75.

W allace A. 1980. Is beta dead. Institutional Investor. (Jan/Feb) 23-30.

(Received 19 May 1993)



Determinants of Foreign Direct Investment in the Malaysian Manufacturing Sector

ZULKARNAIN YUSOP and ROSLAN A. GHAFFAR Faculty of Economics and Management


Keywords: Foreign Direct Investment (FDI), manufacturing, determinantsABSTRAK

Pelaburan langsung asing (PLA) telah memainkan peranan yang penting dalam perkembangan sektor perkilangan di Malaysia. Malangnya, corak PLA tersebut adalah tertumpu kepada beberapa industri tertentu seperti barangan eletronik dan tekstil yang kurang melibatkan tenaga buruh yang mahir, kurang intensif teknologi dan sangat bergantung kepada input yang diimport. Hal ini telah mengakibatkan beberapa masalah seperti asas industri perkilangan yang sempit, mudah dipengaruhi oleh pergolakan (ekonomi) luar negara serta nilai tambah yang rendah. Oleh yang demikian, adalah perlu untuk memperluaskan asas industri perkilangan dan juga menggalakkan lagi perkembangan industri yang berasaskan sumber (tempatan). Kajian ini melihat beberapa faktor (kuantitatif) yang mempengaruhi PLA di sektor perkilangan Malaysia. Keputusan kajian menunjukkan bahawa keadaan ekonomi negara, kestabilan matawang, kemudahan kewangan tempatan, kemudahan infrastruktur, kemudahan tenaga buruh serta insentif pelaburan adalah di antara faktor-faktor yang mempengaruhi PLA di dalam sektor perkilangan di Malaysia.

ABSTRACTForeign direct investment (FDI) has played a significant role in the development of the Malaysian manufacturing sector. Unfortunately, the pattern of FDI is unevenly concentrated in a few industries such as electrical and textile which involve less skilled labour, less intensive technology and are highly dependent on imported inputs. This has led to the problems associated with a narrow manufacturing base, vulnerability to external fluctuations as well as low value added. It is therefore necessary to promote a broader manufacturing base and to further encourage FDI in resource-based industries. This study attempts to look at several quantitative factors that influence FDI in the Malaysian manufacturing sector. The results of the study indicate that a nation’s economic health, currency stability, access to local financing, availability of adequate human and physical infrastructures as well as investment incentives are among the important factors influencing FDI in the manufacturing sector of Malaysia.

INTRODUCTIONForeign direct investment (FDI) has played a substantial role in the developm ent of Malaysian manufacturing industries. Many of the multinational corporations (MNCs) which are involved in direct investment in the Malaysian manufacturing sector have brought with them technological knowhow and business experiences that have contributed to the development of the manufacturing sector and the economy as a whole (Fong 1988 and Beaumont 1990). In addition, MNCs have also been credited for enhancing the com

petitiveness of the Malaysian manufacturing exports in the world market by improving product qualities (Yusof, 1990).

The government on its part has been encouraging private secto r investm en t (especially FDI) more actively since the mid-eighties when the country was experiencing one of its worst recessions. The Investment Act was introduced in 1986 to further stimulate investment activities in various manufacturing industries. This has generated positive results with more foreign investors coming into the country. The share of

*

Zulkamain Yusop and Roslan A. Ghaffar

foreign proposed capital investment in the approved manufacturing projects has increased from 19% in 1984 to 64% in 1992 of the total proposed capital investment. The trend in FDI (in terms of proposed capital investment) in Malaysian manufacturing indicates a sharp increase during the 1984-90 period (Table 1).

TABLE 1Proposed capital investment in approved

manufacturing projects in Malaysia (RM Million)Year Local Foreign Total1984 3083.1 718.0 3801.1

(81%)* (19%) (100%)1985 4727.6 959.3 5686.9

(83%) (17%) (100%)1986 3475.3 1687.9 5163.2

(67%) (33%) (100%)1987 1873.9 2060.0 3933.9

(48%) (52%) (100%)1988 4215.9 4878.0 9093.9

(46%) (54%) (100%)1989 3562.7 8652.7 12215.4

(29%) (71%) (100%)1990 10539.0 17629.1 28168.1

(37%) (63%) (100%)1991 13763.1 17055.3 30818.4

(45%) (55%) (100%)1992 10003.0 17772.1 27775.1

(36%) (64%) (100%)1984-92 55243.6 71412.3 126656.0

(43.6%) (56.4%) (100%)Source: MIDA, 1993’Percentage (rounded to the nearest digit)

Despite the overall increase in FDI, foreign participation (in terms of capital) across the manufacturing industries continues to be uneven. Some of the industries are over invested while others appear to be overlooked by foreign investors. For example, as at December 1988, in the electrical industry, foreign share constituted 81% of the industry’s total fixed asset, while for the beverages and tobacco; rubber products; and textile industries, the foreign shares were 70, 56 and 53% respectively. In contrast, foreign share in the total fixed asset for the wood and wood products; plastic; and chemical industries represented only 14.6%, 17.8% and 21% respectively of the industry’s total fixed asset (MIDA 1990).

The uneven pattern of FDI across the manufacturing industries has resulted in a narrow manufacturing base. At present, m anufacturing activities are highly concentrated in the electrical and textile industries, The narrow manufacturing base appears to be at odds with the country’s diversification policy originated in the 1960’s. The narrow m anufacturing base also has several negative implications in terms of the structure of employment, export and value added. The employment pattern in the m anufacturing sector is concentrated in the electrical and textile industries. As at December, 1991, the electrical industry provided the highest level of employment (22.7% of the total employment in the manufacturing) followed by the textile industry, 15.2% (Table 2). In terms of exports, electrical products (electronic, electrical appliances and m achinery) have accounted for the majority of the share of the export of manufactures. In 1992, the share of electrical products in the total manufactured exports was 57.5% (Table 3).

TABLE 2Malaysia: employment by industry as

at December, 1991Industry (SITC) Employment

(%>Food, Beverages 63,741& Tobacco (311-313) (9.2%)Textiles, Clothing 92,273& Footwear (321-323) (13.3%)Wood Products (331) 56,694

(8.2%)Rubber Products (355) 43,231

(6.2%)Chemicals 8c Petroleum 21,986Products (351-353) (3.2%)Non-Metalic 26,310Mineral Products (369) (3.8%)Metal Products (371) 39,678

(5.7%)Electrical 8c Electronic 211,046Machinery & Appliances (383) (30.5%)Transport Equipment (384) 26,098

(3.8%)Other Manufactures 111,225

(16.1%)Total 692,282

(100.0%)Source: MIDA, 1993


PertanikaJ. Soc. Sci. 8c Hum. Vol. 2 No. 1 1994

TABLE 3Malaysia: export of manufactured goods (RM millions)

1970 1980 1985 1989 1990 1991 1992Total (100) 612 (100) 6101 (100) 12111Food, Beverages (18.3) 112 (7.8) 475 (4.9) 594& TobaccoTextiles, Clothing (6.5) 40 (13.2) 806 (10.6) 1289FootwearWood Products (14.4) 88 (7.6) 467 (3) 363Rubber Products (2.8) 17 (1.4) 84 (1) 133Chemicals 8c (32.2) 197 (5.9) 361 (11.6) 1412Petroleum ProductsNon-Metalic (3.3) 20 (1) 61 (1.2) 150Mineral ProductsMetal Products (4.2) 26 (2-6) 161 (2.5) 300Electrical & Electronic (2.8) 17 (46.4) 2832 (49.8) 6028Machinery & Appliances

(100) 36567 (100) 46833 (100) 61427 (100) 38443(5.0) 1788 (4.4) 2062 (3.7) 2243 (3.6) 1365(8.7) 3190 (8.5) 3983 (7.8) 4805 (7.5) 2874(32)(3.1)(7.4)

118411482698

(2.9)(3.3)(6.8)

136215343174

(3.4)(2.8)(5.8)

206317493539

(3.8)(2.9) (6.4)

146611312461

(1.8) 658 (1.6) 771 (1.4) 888 (13) 509(4.0)

(56.9)1463

20799(3.5)

(56.6)1625

26504(3.1)

(57.9)1883

35602(3.1)

(58.4)1208

22440(3.2)(6.7)

11842455

(4.1)(8.3)

19273891

(5.4)(8.7)

32925363

(4.9)(8.1)

18943095

Transport Equipment (11) 68 (6.7) 407 (8.5) 1031Other Manufactures (4.5) 27 (7.3) 447 (6.9) 831Mote : Figures in parentheses denote percentagesSource : Malaysia, Ministry of Finance, Economic Report, various issues



The electrical industry has also accounted for the major portion of the m anufacturing value added. In 1990, electrical machinery contributed 21.5% of the total m anufacturing value added, while chemical and chemical products contributed 10.7% and food contributed 8.3% (Malaysia, Industrial Surveys, 1992).

The highly dom inant electrical and textile industries, in terms of employment, exports and value added, indicate the narrowness of the country’s manufacturing base and its vulnerablity to external changes in terms of world supply and dem and for the products of the two industries. Should there be any severe drop in the dem and for these products in the world market, the employment, exports as well as economic growth, in the country could be greatly disturbed.

In addition, most of the multinational corporations or MNO owned electronics and electrical appliances plants (which are mainly off-shore plants) are engaged in the assembly processes of integrated circuits (ICs), industrial equipment and consum er appliances like air-conditioners, rad ios an d television sets from com plete ly knocked-dow n im p orted un its for the local and export markets (Fong, 1988). The nature of th e M alaysian e lec trica l in d u s try (h ighly d e p e n d e n t on th e im p o rte d in p u ts ) has been the main reason why the industry value- added is relatively low (Chalmers 1991).

The purpose of this paper is to report on a study o f several c o u n try re la te d fac to rs that are considered im portant when decisions on FDI are made by MNCs. While this study may not provide specific recom m endations for the prom otion of a broader m anufacturing base and encourage FDI in the under invested industries, it would be able to provide a general redirection for the current FDI policies and incentives.1

SPECIFICATION OF FOREIGN DIRECT INVESTMENT FUNCTION

Attempts at explaining FDI have appeared in three major strands of theoretical framework. These different theories are im portant and helpful in understanding the international firms’ operations.

Basically, analysis of FDI has used the comparative advantage theory, industrial organization theory and investment theory (Rock, 1973).

In specifying the FDI function, various country re la ted aspects w hich in flu en ce FDI in Malaysian m anufacturing are considered. It is assumed that multinational firms take into account several aspects of the host country in their FDI decision process to reduce the investment risks, minimize the factor costs and maximize their profits. Selection of the variables is based on previous findings on the determ inants of FDI (Rock 1973, Goh 1973, Ahmed 1975, Schneider and Frey 1986 and Bardai 1989). For the purpose of this study, we are especially interested in several economic factors. Accordingly, an FDI function for the Malaysian m anufac tu ring sector can generally be written as:

Y = f(GNP, MOG, ER, INT, INF, TBA, CPFC, FGDE)

where Y is the dependent variable. Two different proxies for the dependent variable were chosen for this study, namely: (1) the level of total assets o f th e fo re ign co n tro lled firm eng aged in Malaysian m anufacturing (AFC) and (2) the level of investment in fixed assets by foreign controlled firm engaged in the m anufacturing sector of Malaysia (IFC). As defined by Oman (1984), a firm is considered to be involved in FDI if the foreign entity owns a majority of the equities and has direct managerial control in the firm. Thus, for the purpose of this analysis, a manufacturing company with at least 50% of the equities controlled by foreign entities is considered as an FDI company. The total assets of foreign companies include stocks and claims on branches as well as affiliated enterprises abroad, while investment in fixed assets includes acquisition of transport equipment, machinery and office equipment.

GNP (gross national product) reflects the general perform ance of the economy and provides an indication of the size of the local m arket for the m anufacturing outputs (Rock, 1973). A positive relationship is expected between GNP

‘Over the years, the government has enacted and implemented various policies to attract foreign investment in Malaysia. Such policies and their respective programmes are well documented and will not be discussed here. For reference see Yong (1988).



and FDI. MOG (Ratio of manufacturing output over GNP) measures the degree of the country’s industrialization. It also provides a good hint of the supply conditions of manufacturing inputs such as power, transportation, communication and labour. A higher ratio of manufacturing output over the GNP reflects a better supply of investm ent inputs. Thus, a d irec t re la tionsh ip between MOG and FDI is expected. The ER or Foreign Reserve position provides an indication on the exchange rate risks associated with FDI by showing the degree to which the country’s currency is under or over-valued. Foreign investors are likely to incur losses due to exchange rate changes if the currency is over-valued (Rock,1973). A decline in the foreign reserve position is expected to have a negative impact on FDI. Changes in the levels of external reserve do not directly influence FDI, but will first affect the exchange rate because the foreign reserve position actually provides information on the condition of exchange rate. It is the changes in the levels of exchange rates that will affect FDI. Thus, the effect of the foreign reserve position on FDI is not likely to be immediate. Accordingly, external reserve lagged one period is used for the study.

TBA (total asset of the banking system) provides an im portant barom eter on the availability of local financial services such as local credits and insurance services. FDI is expected to increase with TBA. CPFC or profitability indicates whether the activities of m ultinational corporations are affected by the opportunity for high profits. This idea is stressed by the investment theory of FDI. To check whether the level of profits in year t has any impact on the level of FDI in year t+1, the am ount of profits of foreign controlled companies in Malaysian manufacturing (lagged one period) is used for the analysis. INT (Interest rate) is normally associated with high cost of capital which often leads to the reduction in FDI (assuming that domestic financing is im portant to the foreign investors). However, in the case of a poorly developed financial market in which FDI projects are constrained by inadequate domestic savings, increase in the interest rates can help raise the domestic savings and finally promote FDI.

FGDE (federal government development expenditure) represents public investment rate. The impact of FGDE on FDI can be positive or nega

tive. FGDE which mainly involves spending on basic infrastructure such as education, transportation systems, water and sewage facilities can complement and thus, prom ote FDI. Conversely, public investment in projects in which the products compete (for source of funds and other factors of production) with those of the private sector can dam pen FDI to an extent that it substitutes the private projects. INF denotes inflation. In most cases, high rates of inflation provide a good indication of econom ic instability and failure on the part of the government to control the country’s macroeconomic environment.

DATA COLLECTION AND MODEL ESTIMATION

Data for the studies were collected from various sources including The Quarterly Economic Bulletin of Bank Negara Malaysia and the Financial Reports of Limited Companies (published by the Departm ent of Statistics). To eliminate the inflationary factor, the relevant data were deflated by the consumer price index (1967=100). Since Malaysia has been independent for only about 33 years and not until mid 1960s were most of the Malaysian data properly recorded, only a relatively small num ber of observations are available. For the above reasons, annual data covering the period of 1966 to 1988 are used.

A prelim inary estim ation suggested th a t heteroscedasticity was present in the models. In order to reduce the problem, all of the dependent and independent variables were transformed into log form. The two general equations representing the relationships between eight explanatory variables with FDI which are represented by two dependent variables may be expressed as:

LAFC = a 0 + otjLGNP + a^LER + a sLINT + a 4LINF + a.LMOG + a (.LTBA + a^C P F C + cx^FGDE + |i, 11 ]

LIFC = p0 + pjLGNP + pl2LER + P,LINT + P4LINF + p.LMOG + p6LTBA + pXCPFC + PgLFGDE + j l i 2 [2]

Estimations of equations [1] and [2] (Table 4) indicate the presence of a multicollinearity problem. This is supported by the fact that only



few of the explanatory variables are significant in spite of the high values of R2 0.9967 for equation[1] and 0.8642 for equation [2]). A partial correlation analysis shows that GDP is highly correlated to total asset of the banking system, and current profit of foreign controlled companies. External reserve is strongly related to the total asset of the banking system. Manufacturing output/G N P and interest rate are also highly correlated to the total asset of the banking system. To reduce the multicollinearity problem , the models were respecified with closely related variables appearing in different equations. The following regression equations were used:

For LAFC as the dependent variable:LAFC = 7i0 + ftjLGNP + ti E R + tu3LINT

+ 7t4LINF + m [3]

LAFC = T0 + TjLMOG + x^CPFC + t3LTBA + t4LFGDE + n4 [4]

For LIFC as the dependent variable:LIFC = oo0 + oo^GNP + oot2LER + oo^LINT

+ oo4LINF + ^ ~

LIFC = <|>0 + (t>jLMOG + ((^CPFC + <|>3LTBA + <|)4LFGDE + |Li6 [6]

All of the regression results for equation [3] through [6] are summarized in Table 4. It is shown in equation [3] th a t gross national p rod uc t (LGNP), external reserve lagged one period (LER) and interest rate (LINT) are positively related to the LAFC. The t values are significant at the 5% level for the GNP and LINT and at the 1 % level for LER. The coefficient for inflation (LINF) shows the correct negative sign as expected. However, its t value is not significant at the 10% level. Equation [4] demonstrates a significant positive relationship between manufacturing output/G N P (LMOG), total assets of the banking system (LTBA) and current profits of foreign controlled companies in m anufacturing sector lagged one period (LCPFC) with the dependent variable (LAFC). Their t values are all significant at the 1% level. Public investm ent (LFGDE) shows a significant negative relationship with the dependent variable. Its t value is less than the critical t value at a = 0.05.

Using LIFC as the dependent variable in equation [5], a significant positive relationship between LGNP, LER and LINT with the dependent variable was obtained. The t value for LGNP is significant at the 5% level while for LINT and LER, they are significant at the 10% level. Inflation (LINF) does not show any significant relationship with the dependent variable; however, its slope coefficient shows a negative sign.

Regressing LIFC on LMOG, LTBA, LCPFC and LFGDE in equation [6], it is found that three explanatory variables, namely LTBA, LMOG and LCPFC show significant positive relationship with the dependent variable. The t values for LMOG and LCPFC are significant at the 10% level while for LTBA its t value is significant at the 5% level.

SUMMARY AND POLICY IMPLICATIONSThe analysis presented in this paper reveals that gross national product (GNP), net external reserves, interest rates, manufacturing output/GNP, current profits of foreign controlled companies in Malaysian m anufacturing and total assets of the banking system are im portan t factors influencing FDI in Malaysian manufacturing. The significant positive relationship between GNP and the dependent variable is expected because a high level of GNP is normally associated with positive growth and a nation’s economic health. The level of external reserve (LER) which indicates the extent to which Malaysian currency is under or over-valued also has a positive impact on the activities of FDI. A low level of external reserve discourages FDI because it leads to overvaluation of a country’s currency which can increase the risk of investment activities through the currency exchange loss. The level of interest rate also plays a significant role in FDI decision. In the Malaysian case, increase in interest rate is associated with the increase in FDI. The d irect relationship between interest (deposit) rate and savings has been su p p o rted by several econom ists, for instance Shaw (1973). According to them, increase in deposit rates can lead to an influx of deposits into commercial banks which ultimately expands the real size of the banking system and hence raises the net flow of real bank credit to finance investment. Inflation does not seem to be a significant determ inant for FDI in Malaysia. A careful review on the trend and degree of infla-


Pertanika J. Soc. Sci. 8c Hum. Vol. 2 No. 1 1994

TABLE 4Results of regressions

Equation

Dep. Constant LGNP LER(-l) LINT LINF LMOG LTBA LCPFC(-l) LFGDE AdjustedR2

Durbin-Watson

F-value

[1] LAFC 1.4337 0.0968 0.0172 -0.0533 0.0393 0.5347 0.4499 0.3357 -0.1101 0.9967 2.4627 673.11(4.4981)*** (0.8709) (0.5929) (-1.1447) (2.7812)** (7.9965)*** (8.4658)*** (8.3840) (-5.0633)***

***

[2] LIFC -1.9022 -2.0671 0.5764 1.0397 0.1265 -0.8364 1.7358 0.2538 0.2114 0.8643 2.3873 15.16(-0.4998) (-1.4203) (1.6269) (2.0114)* (0.8250) (-.9570) (2.6154)** (0.6037) (-0.8069) ***

[3] LAFC 0.6694(0.6921)

0.3642(2.3765)**

0.2340(3.2192)***

0.2064(2.7512)**

-0.1165(-0.5599)

0.9923 1.7650 515.43***

[4] LAFC 1.9049(12.863)***

0.4866(5.4876)***

0.4293(22.965)***

0.4437(15.165)***

-0.1191(-4.2023)***

0.9932 2.0037 585.98***

[5] LIFC -8.0931(-6.0935)***

0.9444(2.6405)**

0.7060(1.7986)*

0.9803(1.7612)*

-0.0544(-.5291)

0.8716 1.8934 18.93***

[6] LIFC 1.0430(2.0281)*

0.4749(2.5419)**

0.5902(1.9058)*

-0.1296(-0.4335)

0.7883 1.6986 19.61***

Figure in parentheses denote t-values: * significant at 10% level of confidence** significant at 5% level of confidence

*** significant at 1% level of confidence



tion in Malaysia shows that in general, inflation in Malaysia has been well under control and has not been as severe as in many other developing countries (Yusof, 1985).

The posidve relationship between (LMOG) and FDI implies that certain hum an aspects and physical infrastructure related to manufacturing such as skilled and cheap labour; reliable supply of water and energy; and adequate transportation system for dissemination of the products are im portant in the eyes of foreign investors. Furtherm ore, it suggests that the host country’s conducive policies can contribute to the increase in FDI. A greater percentage share of manufacturing ou tpu t in the gross national p roduct also reflects the society’s acceptance of the “discipline” of manufacturing industry. Profitability is another im portan t consideration in the FDI decision process. This is confirmed by the strong positive relationship between the lagged value of current profits earned by foreign controlled companies in the Malaysian m anufacturing sector and the dependent variable. This relationship shows that foreign investors are quite sensitive to investment incentives which can increase the possibility for higher profits. Accordingly, this finding suggests the possibility of controlling FDI through investm ent incentives. Furtherm ore, the analysis demonstrates that local funds and other local financial services including insurance are im portant factors influencing FDI. The fact is supported by the strong positive relationship between the level of assets of the consolidated banking system and the dependent variable. This finding provides a base for guiding FDI operations in Malaysian manufacturing through credit control or credit allocation policies. In fact, a study by Liete and Vaez-Zadeh (1986) on credit allocation and investm ent decision in the Korean m anufacturing sector has concluded that limitations on credit availability tend to affect investment decision direcdy ra ther than through interest rate movements. Finally, the results show a significant negative impact of public investment rate on FDI. Public investment can either be in basic infrastructure (such as education, transportation systems, water and sewage facilities) which would facilitate FDI or in projects for the public sector enterprises where the products compete (for input factors as well as market) with those of the private sector.

In Malaysia, the public sector enterprises have been involved in both types of investments. The two different elements of public investments have two different directions of impact on investment activities. Thus, the negative relationship between public investment rate and FDI suggests that public investment activities on balance have an adverse effect on FDI. The negative consequences of government activities via the public sector have been noted by a num ber of observers. Among these are the reduction of fair competition among the private firms in Malaysia (Ghazali 1988) and the generation of macroeconomic problems such as inflation and balance of payment difficulties (Fischer, 1988).

Several aspects should be considered in the policy making process in order to improve the pattern of FDI in the various m anufacturing industries in the country.

First, since certain hum an aspects and physical infrastructures related to m anufacturing are im portant in the eyes of foreign investors, the provision of a well trained and efficient labour force, special or subsidized industrial sites, and other infrastructural facilities can attract foreign investors to certain designated industries.

Next, the allocation of special funds or credit facilities for firms which participate in industries which are under invested can also increase the flow of FDI into those industries because many foreign firms regard the possibility for local finance as one of the im portant considerations in their FDI decisions.

In addition, policy makers may also utilize tax incentives in order to improve the pattern of FDI since our analysis shows that increase in profitability is quite an im portant elem ent in the FDI decision. The use of tax incentives in guiding the FDI activities should however be blended with other incentives (other than tax) in order to be more effective.

It should be borne in mind that the better way to design policies is to first maintain and further elevate the country’s attractiveness as an investment ground for foreign investors. Guiding the activities of FDI accross the various industries in the country would be the subsequent step. It is im portant that the government aim at keeping a steady economic growth, reduce the currency fluc-



tuation, control the inflation, increase the efficiency of the public sector and financial institutions, upgrade the quality of labour force and more importandy, maintain polidcal stability.

REFERENCESA h m e d , A. 1975. Determinants of direct investment,

with special reference to developing countries. Ph.D. Dissertation. University of Pennslyvania.

B ardai, B . 1989. Evaluation of Malaysian corporate investment incentives. Malaysian Institute of Economic Research (MIER). Paper No. 25 (August).

B e a u m o n t , P.O. 1990. The role of foreign investment in Malaysia JumalEkonomi Malaysia 21 & 22: 65- 77.

C halm ers, 1.1991. The political economy of the electronics industry in ASEAN. ASEAN Economic Bulletin. 8(2): 194-209.

F ischer , B. 1988. Public enterprises in developing countries: market structure competition, efficiency and the scope for privatisation. Malaysian Journal Of Economic Studies 25 (2): 45-61.

F o n g , C.O. 1988. Multinational corporations in ASEAN: technology transfer and linkages with host economies. Paper presented at the 13th Annual Conference of the Federation of ASEAN Economic Association. Penang. Malaysia.

G hazali,Y. 1988. Malaysia’s industrialization policies and strategies - Have they helped or retarded growth? A Paper presented at the Federation Of Malaysia Manufacturing (VMM) National Manufacturing Conference. Jan. 22, 1988.

G o h , K. 1973. Factors Influencing U.S. International Firms in Locating Export-Oriented Manufacturing Facilities in Singapore, Taiwan and Hong Kong. Ph.D. Dissertation. Indiana University.

L iete , S.P . and R. V aez-z a d eh . 1986. Credit allocation and investment decisions: the case of the manufacturing sector in Korea. World Development. 14(1): 115-126.

Malaysia, Bank Negara, Quarterly Economic Bulletin, various issues.

____________________________Department of Statistics. Industrial Survey,

various issues._______ . Department of Statistics. Financial Report

of Limited Companies, various issues._______ . Ministry of Finance. Economic Report, vari

ous issues.MIDA (Malaysian Industrial Development Author

ity) , Statistics on the ManufacturingSector. Various issues.

O m a n , C. 1984. New Forms of International Investment in Developing Countries. Paris : OECD.

R ock , M.1973. Cross country analysis of the determinants of d.s. foreign direct investment in manufacturing industry in less developed countries. Ph.D. Dissertation. University of Pittsburgh.

S c h n eid e r , F. and B.S. F rey 1985. Economic and political determinants of foreign direct investment. World Development 13: 161-175.

S haw , E.S. 1973. Financial Deepening in Economic Development. New York: Oxford University Press.

Yo n g , P. C. 1988. Latest Malaysian Investment Incentives. Kuala Lumpur: Longman Malaysia Sdn. Bhd.

Yu s o f , M. 1985. Inflation and controls: Malaysian experience in 1970’s. Journal Ekonomi Malaysia 11:35-54.

Y u s o f , M.I. 1990. Export competitiveness: objectives and performance of the Industrial Master Plan. Jumal Ekonomi Malaysia 21-22: 99-118.

(Received 25 May 1993)



Kajian Kesesuaian dan Keberkesanan Raven SPM untuk Mengenal Pasti Kanak-kanak Melayu yang Pintar Cerdas

ABD MAJID MOHD ISA Fakulti Pengajian Pendidikan Universiti Pertanian Malaysia,


Katakunci: p in tar cerdas, inventori, keberkesanan, ujian piawai, kecerdasanABSTRAK

Kajian ini bertujuan menentukan kesesuaian inventori Raven SPM sebagai ujian kecerdasan secara kumpulan untuk menapis kanak-kanak pintar cerdas (gifted). Seramai 100 pelajar Melayu tingkatan satu, Sekolah Menengah Datuk Abdullah, Rembau merupakan responden bagi kajian ini. Mereka terdiri daripada yang mempunyai pencapaian yang baik dalam Ujian Pencapaian Sekolah Rendah dan layak ke sekolah asrama penuh seluruh Daerah Tampin/Rembau. Ujian Raven SPM dan WISC-R diberikan kepada mereka. Responden yang mendapat skor IQ melebihi 120 dari ujian WISC-R dikelaskan sebagai pintar cerdas. Korelasi antara skor Raven SPM dengan IQ WISC-R adalah tinggi dan kedua inventori ini berkongsi varians bagi konstruk kecerdasan (kriteria untuk pintar cerdas) sebanyak 65 peratus. Persamaan regresi untuk menganggarkan IQ (indeks kecerdasan) mendapati bahawa skor 50 boleh menganggarkan IQ 120 (pintar cerdas). Oleh kerana ralat bagi persamaan regresi ini tinggi, maka ramai kanak-kanak yang pintar cerdas mendapat skor kurang daripada 50 dan ini dinamakan false negative. Indeks keberkesanan dan kecekapan yang dikemukakan oleh Pegnato dan Birch (1959) diguna untuk menentukan sama ada skor 50 pada ujian Raven SPM dapat mengurangkan false negative. Oleh kerana skor yang kurang daripada itu akan menambahkan bilangan calon untuk diuji, satu skor yang ekonomikal dari segi kos perlu dikemukakan. Daripada dapatan penyelidikan ini, skor Raven SPM sebanyak 48 didapati dapat mengurangkan bilangan calon sebanyak 30 peratus dan skor ini mengandungi 85 peratus kanak-kanak pintar cerdas. Jika dibandingkan dengan skor yang lain, skor ini mempunyai indeks keberkesanan yang baik kerana seorang daripada tiga calon yang akan diuji seterusnya ialah kanak-kanak pintar cerdas.

ABSTRACTThe objective of this study is to determine the suitability of Raven’s SPM as a group intelligence test for the screening of gifted children. One hundred Form One Malay students from Sekolah Menengah Datuk Abdullah, Rembau were the respondents in the study. They were among the high achievers in the 1990 Ujian Pencapaian Sekolah Rendah at Tampin/Rembau District who qualified to be emplaced at residential schools. The Raven’s SPM and WISC-R were administered. Students with IQ scores of more than 120 on the WISC-R are classified as gifted. The correlation between the Raven’s SPM score and the IQ score derived from the WISC-R was high and seemed to share about 65 percent of the ‘intelligence’ costruct. From the regression equation, the gift5v;ctness construct could be drawn from a Raven’s SPM score of 50. However, the standard error for this equation was also notably high. Therefore, many of the gifted children may be among those with a Raven’s score of less than 50 and this is known as false negative. The effectiveness and efficiency indexes forwarded by Pegnato and Birch (1959) were used in determining whether a score lower than 50 on the Raven’s SPM would reduce the false negative. As the lower score would increase the number of children to be tested, the score recommended must be cost effective. A score of 48 was recommeded as it reduces the number to be tested to 30 percent, and includes 85 percent of the gifted children. Also, the efficiency is high compared with other scores. Using this procedure, it is estimated that one among three children studied is gifted.

Abd Majid Mohd Isa

PENGENALANBagi mencapai Wawasan 2020, Perdana Menteri Malaysia, Majlis Amanah Rakyat (MARA), sebuah agensi k e ra jaan yang d itu b u h k a n u n tu k m em ajukan Bumiputra, telah m engum um kan untuk menghasilkan 1000 pintar cerdas (gifted) Melayu pada tahun 2020 (Utusan Malaysia, 11 November, 1991). Salah sebuah daripada 18 Maktab Rendah Sains MARA (MRSM) yang ada pada masa ini, akan menjadi sekolah bagi kanak- kanak p in tar cerdas. Dalam masa yang sama, Kementerian Pendidikan yang sedang mengkaji Akta P end id ikan , 1961 ju g a akan m em beri tum puan khas kepada kanak-kanak pintar cerdas dan program pendidikan ini akan dijalankan melalui sekolah berasrama penuh (Awang Had Salleh, 1990).

Kanak-kanak p in ta r cerdas (gifted) ialah kanak-kanak yang mempunyai kecerdasan yang tinggi. Kecerdasan ini diukur melalui skor IQ yang dihasilkan oleh ujian kecerdasan (intelligence test). Kanak-kanak yang mempunyai skor IQ yang tinggi selalunya menjadi orang dewasa beijaya: m en d ap a t kelayakan akadem ik yang tinggi, menjadi profesional yang menghasilkan reka cipta (patent) dan mendapat pendapatan yang lumayan (Oden, 1968; Martison, 1961; Reynold, 1962). Oleh itu, semua institusi pendidikan kanak-kanak pintar cerdas memilih calon mereka berdasarkan skor IQ.

Skor IQ daripada duajenis ujian kecerdasan secara individu yang sering digunakan un tuk memilih kanak-kanak pintar cerdas ialah Weschler Intelligence Scale for Children-Reuised (WISC-R) dan Stanford-Binet. Bancian yang dijalankan oleh Karnes dan Collins (1987), m endapati WISC-R lebih popular dan banyak dikaji terutam a bagi kanak-kanak yang tidak berun tung (diadvantaged) dan tidak m enggunakan bahasa Inggeris sebagai bahasa utama. Ujian kecerdasan secara ind iv idu hanya bo leh d ig u n ak an o leh ahli psikologi yang bertauliah. Untuk seorang kanak- kanak, ujian ini memerlukan lebih daripada 60 m in it. Sudah pasti ianya tidak ekonom ikal digunakan untuk semua calon.

Selalunya, terdapat sekurang-kurangnya dua proses kanak-kanak pintar cerdas ini dikenal pasti. Proses pertama ialah proses tapisan atau screening. Alat a tau m aklum at yang d igu nakan u n tu k

menapis calon ini selalunya ialah penamaan guru, pencapaian ujian akademik, ujian kreativiti dan ujian kecerdasan secara berkumpulan. M enurut Pegnato dan Birch (1959), ujian kecerdasan secara berkum pulan adalah alat tapisan yang paling efektif, sementara penamaan guru selalunya akan menyingkirkan ramai calon pintar cerdas. Proses kedua ialah proses pengesahan ataupun proses final confirmation. Dalam proses ini, kanak-kanak yang berjaya dalam proses tapisan tadi akan melalui ujian kecerdasan secara individu sama ada Stanford-Binet ataupun WISC-R.

Didapati korelasi skor ujian kecerdasan secara kumpulan dengan skor ujian kecerdasan secara individu adalah tinggi (Covin, 1977; Lowrence dan A nderson, 1979; Rust dan Lose, 1980). Ujian kecerdasan secara kumpulan tidak memerlukan ahli psikologi bertauliah untuk melaksanakannya. Selain daripada itu, ianya boleh juga dijalankan serentak dan mudah untuk dikira skornya.

Oleh itu, ujian kecerdasan secara kumpulan ham pir sama dengan ujian pencapaian seperti U jian P encapaian Sekolah R endah (UPSR). Perbezaannya dengan UPSR hanyalah dari segi item ujian di m ana ujian kecerdasan secara kumpulan tidak dipengaruhi oleh pengalaman sekolah dan latar belakang kanak-kanak. Terdapat sa tu k e lem ah an u jian kece rd asan secara ku m pulan den g an u jian kecerdasan secara individu iaitu ujian kecerdasan secara kumpulan gagal u n tu k m em bezakan IQ di ka lang an kum pulan kanak-kanak yang kecerdasannya tinggi. U ntuk m enen tukan kesesuaian dalam memilih kanak-kanak cerdas pintar, ianya perlulah dikaji terlebih dahulu.

Raven's Standard Progressive Matrices (Raven SPM) adalah antara ujian penapisan yang sering digunakan untuk memilih kanak-kanak pintar cerdas yang kurang beruntung (disadvantaged). Di Malaysia, kajian ten tang keberkesanan Raven SPM untuk menapis kanak-kanak pintar cerdas belum lagi dijalankan. Jika didapati ujian Raven SPM mempunyai keberkesanan yang baik, maka, kos untuk mengendalikan ujian individu sama ada Stanford Binet atau WISC-R bolehlah dijimatkan dan ram ai p in ta r cerdas dapat dikenal pasti. Objektif utama kajian ini ialah untuk meninjau sejauh mana berkesannya ujian Raven SPM untuk menapis calon pintar cerdas.


Kajian Kesesuaian dan Keberkesanan Raven SPM

METODOLOGI KAJIANPensampelanKajian ini m erupakan satu penerokaan. Untuk mengurangkan kejadian Ralat I dan Ralat II, saiz sampel yang secukupnya perlu dikenal pasti. Saiz sampel yang minimum bagi kajian ini ditentukan dengan m enggunakan form ula yang dikemukakan oleh Cohen, iaitu,

Sumber: Robert S. Barcikowski (1987) pi 1-12

di mana:a =15 (sisihan lazim bagi WISC-R) ap = .05 (iaitu tahap signifikan)Zi_ap = 1.645 (skor z pada persentil 95)|l0 =100 (min bagi WISC-R)(3 = .85 (1-kuasa)jlxa = 1 2 0 ( s k o r m i n i m a p in t a r c e r d a s )Z1 _p =1.04 (skor z pada persentil 85)

Berdasarkan kepada perkiraan di atas, saiz sam pel yang m inim um bagi kajian ini ialah sebanyak 40 sahaja.

RespondenU n tu k m en in g g ik an lagi kesah an d ap a tan penyelidikan ini, seramai 100 (53 lelaki dan 47 perem puan) pelajar tingkatan satu tahun 1991 Sekolah M enengah Datuk Abdullah, Rembau, ada lah re sp o n d e n kajian in i. K esem uanya berbangsa Melayu dan mereka terdiri daripada pelajar yang m endapat gred UPSRyang baik pada tahun 1990 di Daerah Tam pin/Rem bau. Oleh kerana tempat di sekolah berasrama penuh adalah terhad, m ereka dikumpulkan untuk belajar di seko lah in i yang m em punyai asram a dan kemudahan yang dianggap baik.

T iad a seb aran g piawai u n tu k m engkla- sifikasikan kanak-kanak pintar cerdas. Terman (1925) m enggunakan skor 140 ke atas ujian Stanford Binet untuk mengklasifikasikan kanak- kanak pintar cerdas. Sejak akhir-akhir ini, kanak- kanak pintar cerdas diklasifikasikan sebagai yang tergolong 10 peratus teratas. Ini ialah kerana skor ujian IQ berbeza mengikut latar belakang kanak- kanak. U n tu k kajian in i, re sp o n d e n yang m endapat skor 1.33 sisihan lazim di atas min iaitu

skor 120 Full IQ WISC-R ke atas dikelaskan sebagai kanak-kanak pin tar cerdas. Seramai 12 orang responden mempunyai IQ 120 ke atas.

Instrumen KajianKajian ini m enggunakan dua instrum en ujian kecerdasan iaitu WISC-R (ujian kecerdasan secara individu) dan Raven SPM (ujian kecerdasan secara kumpulan).

a. WISORUjian WISC-R ataupun Wechsler Intelligence Scale for Children-Revised terbitan ‘The Pschological Corporation’ pada tahun 1974 adalah semakan semula WISC, yang pertama kali digunakan pada tahun 1949. Ia mempunyai 10 bahagian (sub-tests); information, similaritiesy arithmatic, vocabulary, comprehension picture completion, picture arrangement, block design, object assembly, dan coding. WISC-R melaporkan tiga skor IQ iaitu verbal IQ (terdiri dari jum lah 5 bahagian iaitu information hingga comprehension) , performance IQ (jumlah bagi picture completion hingga coding) dan Full IQ (jumlah kesemua 10 bahagian).

Semua item dalam ujian WISC-R diteijemah ke Bahasa Melayu m eng iku t p ro sed u r yang dicadangkan oleh buku panduan (Wechsler,1974). Setelah diteijemah, item ini diuji dengan menggunakan 20 (10 lelaki dan 10 perem puan) kanak-kanak kakitangan Universiti Pertanian Malaysia yang berum ur berum ur 12 tahun. Min, sisihan lazim dan reliabiliti (ketekalan dalaman dan ujian semula) hasil teijem ahan ini adalah seperdjadual 1.

Pengukur IQyang penting; Verbal IQ Performance IQ dan Full IQ tidak mempunyai perbezaan yang signifikan dengan kum pulan bandingan seperd dalam buku panduan (manual). Skor Full IQ digunakan untuk memilih calon pintar cerdas mempunyai min sebanyak 103 dan sisihan lazim leb ih ku ran g 15. Ini tidak berbeza den gan kumpulan bandingan di mana minnya ialah 100 dan sisihan lazimnya 15.

Form ula Cronbach-Alpha digunakan un tuk m en g u k u r k e tek a lan da lam an WISC-R m enunjukkan ketekalan dalamannya bagi tiga skor IQ ialah antara .89 hingga .91. Dari indeks ini, Standard Error of Measurement (SEM) bagi Full IQ ialah ±6. Ini bermakiva, jika seseorang kanak- kanak itu m endapat skor 110, maka, skor sebenar


Abd Majid Mohd Isa

JADUAL1 Reliabiliti WISC-R dalam Bahasa Melayu

WISOR Min SL Cronbach Alpha SEM Ujian Semula*Verbal IQ 49.1 11.3 .89 5.17 .90Performance IQ 55.3 11.3 .87 5.57 .91Full IQ 103.1 14.9 .91 6.19 .91Nota: SL = sisihan lazim

SEM = Standard Error of Measurement *selepas 30 hari

(true score) kanak-kanak itu ialah di antara 104 hingga 116 (pada 95% paras keyakinan). Ini menggambarkan bahawa WISC-R dalam Bahasa Melayu m em punyai ketekalan dalam an yang tinggi. Korelasi antara ujian pertama dengan ujian kedua selepas 30 hari pula sekitar .90. Oleh itu, kestabilan item WISC-R dalam Bahasa Melayu juga tinggi.

b. Raven SPMRaven SPM m engandungi 60 item berben tuk matrik yang mana kanak-kanak memilih matrik yang ganjil atau tidak sesuai. Selain daripada a ra h a n sebelum kanak-kanak d ib e n a rk a n menjawab, ia tidak perlu diteijem ahkan ke dalam Bahasa Melayu. M engikut buku panduan atau manual ujian tahun 1977, ketekalan dalam an (menggunakan kaedah split-half ialah antara .60 hingga .97. Korelasi ujian semula selepas setahun ialah di antara .55 hingga .84.

Pengumpulan DataData untuk kajian ini mula dikumpulkan pada b u lan Ju la i, 1991 ia itu se lepas re sp o n d e n m enduduki peperiksaan pertengahan tahun dan sebelum cuti penggal kedua bermula. Pada hari pertama, dengan bantuan dua orang kaunselor sekolah tersebut, responden dikum pulkan di dewan sekolah dan ujian Raven SPM diberikan. Masa untuk responden menjalankan ujian Raven SPM tidak dihadkan. Pelajar yang telah selesai dibenarkan balik ke bilik daijah mereka. Dari pemerhatian, masa yang diambil ialah di antara 30 hingga 45 minit.

Selang beberapa hari selepas itu, responden dipanggil seorang demi seorang untuk menjalani ujian WISC-R. Ujian ini dijalankan oleh penyelidik di b ilik kaunse ling . Secara p u ra ta seo rang responden mengambil masa 70 m init (sisihan

lazim 12 minit) seorang untuk menjalani ujian WISC-R. U ntuk waktu persekolahan iaitu dari pukul 8 pagi hingga 1 tengahari hanya 4 orang responden dapat diuji. Penyelidik m engambil masa 30 hari persekolahan untuk tujuan menguji WISC-R kepada 100 responden.

DAPATAN KAJIANKorelasi antara skor Raven SPM dan skor WISC-R (verballQ , performancelQ dan full IQ) ialah antara .74 hingga .82 (Jadual 2). Dengan korelasi ini jum lah varians (ganda dua r) yang dikongsi oleh kedua ujian ini ialah lebih kurang 65 peratus. Oleh itu, dapadah dirumuskan bahawa kedua ujian ini m engukur konstruk yang sama iaitu konstruk ‘kecerdasan’.

JADUAL2Matrik Korelasi Raven SPM dengan WISC-R

Ujian WISC-R IQverbal performance full

Raven SPM .74 .78 .82WISC-R:

IQ verbal .67 .92IQ performance .89

Oleh kerana korelasi antara skor Raven SPM dengan IQ adalah dnggi dan perkaitannya pula linear; maka IQ boleh diramalkan melalui skor Raven SPM. Persamaan untuk m eramalkan IQ responden berdasarkan skor ujian Raven didapad melalui analisis regresi seperd berikut:

IQ= 1.56(Raven SPM) + 40.32

di mana, ganda dua R ialah .5429 dan ralat (SE) = 8.46. Persam aan ini d idapati signifikan berdasarkan pada nilai F (rujuk jadual 3).


Kajian Kesesuaian dan Keberkesanan Raven SPM

JADUAL 3 Analisis Varians bagi Regresi

Sumber df SS MS F pRegresi 1 14663.2 14663.2 194.5 .0001Ralat 98 7385.3 75.4Jumlah 99 22048.5

Berdasarkan persamaan ini, skor dari ujian Raven SPM boleh menganggarkan IQ (jadual 4). Dari formula ini, kanak-kanak pintar cerdas (yang m em punyai IQ 120) bo leh dipilih dari yang m endapat skor Raven SPM sebanyak 50. Oleh kerana persamaaan ini mempunyai ralat, maka ada responden yang IQnya tinggi tetapi mendapat skor Raven SPM kurang daripada 50 dan ini d in am ak an fa lse negative. R esp ond en yang m endapat skor 50 atau lebih, tetapi IQnya tidak mencapai 120 pula adalah tergolong sebagai false positive. M enurut Fineman dan Carran (1986), kejadian false negative adalah lebih serius daripada kejadian false positive. Usaha perlulah dijalankan un tu k m engurangkan kejadian false negative supaya tiada kanak-kanak yang benar-benar layak tidak terpilih.

JADUAL4 Anggaran Kecerdasan (IQ) berdasarkan

Skor Raven SPMRaven SPM Anggaran

IQ (WISOR)Skor %52 12 121.550 22 118.348 34 115.246 40 112.1

Untuk m engurangkan bilangan kanak-kanak yang IQnya melebihi 120 tidak dipilih (false negative) , maka skor Raven SPM mestilah dikurangkan. Pengurangan skor Raven akan m enam bahkan bilangan calon yang mesti diuji dengan WISC-R Ini akan m enam bahkan kos. Untuk mencari titik perseim bangan di antara kos dengan pengurangan false negative, indeks keberkesanan dan kecekapan yang dikemukakan oleh Pegnato dan Birch (1959) perlu diambil kira. Keberkesanan ialah peratus kanak-kanak pin tar cerdas yang dapat dikenal pasti dan kecekapan pula ialah

bilangan kanak-kanak pin tar cerdas yang ada dalam kanak-kanak yang dicadangkan untuk diuji seterusnya.

Pengiraan kedua^dua indeks ini adalah seperti berikut. Andaikan, dalam satu sekolah ada 50 kanak-kanak dan 10 orang dikenal pasti sebagai pintar cerdas. Seramai 20 pelajar m endapat gred A dan 8 orang yang m endapat gred A ialah pintar cerdas. Keberkesanan gred A ialah 8 /10 x 100% = 80% dan kecekapan pula ialah 8 /20 x 100% = 40%.

Jadual 5 menunjukkan indeks keberkesanan dan kecekapan bagi skor 46, 48, 50 dan 52 Raven SPM. Daripada persamaan regresi, skor yang boleh menganggarkan pintar cerdas ialah 50. Jika skor sebanyak 50 ini dicadangkan untuk proses tapisan, keberkesanannya hanya menghampiri 60 peratus. Pada skor ini, lima orang kanak-kanak pintar cerdas tidak dipilih {false negative) untuk ke proses final confirmation.

JADUAL5 Keberkesanan dan Kecekapan

beberapa Skor Raven SPMRaven SPM Pintar Keberkesanan Kecekapan Skor N Cerdas (N=12) (%) (%)52 12 2 16.7 16.750 22 7 58.3 31.848 30 10 83.3 33.346 40 12 100.0 30.0

Berdasarkan kepada maklumat dalam Jadual4, skor yang dicadangkan untuk proses tapisan (,screening) ialah 48. Pada skor 48, didapati ham pir semua (83%) kanak-kanak pintar cerdas dipilih dalam proses tapisan. Dari segi kosnya pula, indeks keberkesanan pada skor 48 pula ialah 33.3 peratus. Ini bermakna, seorang daripada tiga orang kanak- kanak yang dipilih untuk diuji dengan WISC-R ialah kanak-kanak pintar cerdas.

PERBIN CAN GAN DAN CADANGANUjian Raven SPM sebagai proses penapisan (screening) untuk mengenal pasti kanak-kanak Melayu yang pintar cerdas sebelum seseorang itu diuji dengan lebih intensif dengan inventori yang lain adalah sesuai. Penggunaan ujian Raven secara efektif (mengurangkan kejadian ‘false negative’) dapat mengurangkan bilangan calon kepada 30


Abd Majid Mohd Isa

peratus. Ini akan menjimatkan kos dan menjadikan proses pemilihan akhir dengan menggunakan ujian WISC-R atau Stanford-B inet dapat dijalankan dengan berkesan.

Kajian ini cum a satu p en ero kaan . U ntuk kesim pulan yang lebih konklusif, kajian yang serupa hendaklah dijalankan kepada kumpulan kanak-kanak yang lain. Ini adalah kerana ujian kecerdasan ini dikatakan peka kepada budaya dan latar belakang kanak-kanak itu. Alat atau inventori la in p e rlu d iu ji kesesu aian nya k e ran a ada kemungkinan alat atau inventori lain yang lebih e fek tif dari Raven SPM. Jug a , jik a p rog ram p e n d id ik a n kanak-kanak p in ta r cerdas ini dilaksanakan, penggunaan Raven SPM untuk setiap tahun tidak lagi efektif dan inventori gantian amadah diperlukan.

RUJUKANA wang H ad Sa l leh . 1990. Kertas Dasar Seminar Kanak-

Kanak Pintar Cerdas, di Universiti Malaya anjuran IPI dan Universiti Malaya.

B arcikowski, R.S. 1987. STATPRO Statistics Handbook: A guide for Data Analysis Part 2. Professional Publishing.

C o v in , T.M. 1977. Comparison of SIT and WISC-R IQs among special education candidates. Psychology in the Schools 14: 19-23.

F ine m a n dan C a r r a n , 1986. An epidemiologic approach to screening gifted students utilizing WISC-R subtests. Psychology in the Schools 23: 142- 147.

Karnes, F.A. dan C o l l in s , E.C. 1987. State definitions on the gifted and talented: A report and analysis. Journal for Education of the Gifted 22: 44-62.

L o w ren ce , D. dan A n d e r s o n , H.N. 1979. A comparison of the Slosson Intelligence Test and the WISC-R with elementary school children. Psychology in the Schools 16: 361-364.

M a r t iso n , R.A. 1961. California State Study of Educational Programs for Gifted Children. Sacramento: California State Department of Education.

O d e n , M.H. 1968. The fulfillment of promise: 40 year follow-up of Terman gifted group. Genetic Psychology Monographs 77: 3-93.

P eg n a t o , C. dan B irc h , J.W. 1959. Locating gifted children in junior schools: a comparison of methods. Exceptional Children 25: 300-304.

R aven , J.C. 1977. The Mannual for Raven's Standard Progressive Matrices. London: H.K Lewis.

R ey no ld , M.C. 1962. Early School Admission for Mentally Advanced Children: A review of Research and Practice. Washington D.C.:Council for Exceptional Children.

R u s t , J.O. dan L o s e , B.D. 1980. Screening for giftedness with Slossen and SRBCSS. Psychology in the Schools 17: 446-451.

T erm an , L.M. 1925. Mental and Physical Traits of a Thousand Gifted Children, Genetics Studies of Genius, Vol. I, Stanford University Press.

Temuramah Khas dengan Pengerusi MARA, YB Tamrin Ghafar. Utusan Malaysia, 11 November, 1991.

W echsler , D. 1974. Manual for WISGR. New York: Psychological Corporation.

(Received 17 September 1993)


PertanikaJ. Soc. Sci. 8c Hum. 2(1): 69-76 (1994) ISSN: 0128-7702 © Universiti Pertanian Malaysia Press

Malaysian English: Exploring the Possibility of Standardization

ROSLI TALIF and TING SU HIE Department of Languages

Faculty of Educational Studies Universiti Pertanian Malaysia


Keywords: Malaysian English, New English, language standardization, Standard English, varieties o f English, colloquial English, Educated Singapore English

ABSTRAKKertas keija ini disediakan untuk mempei temukan gagasan tentang Bahasa Inggeris Malaysia dengan gagasan bahasa Inggeris baku dan proses pembakuan bahasa. Petanda-petanda bahasa Inggeris telah dibakukan dapat diteliti melalui: (1) pandangan umum tentang perkembangan variasi Bahasa Inggeris lisan di Malaysia sejak pemerintahan Inggeris; dan (2) penelitian faktor-faktor yang berhubung dengan pembakuan bahasa: (a) penelitian peranan agen pengkodifikasian dalam kebiasaan penggunaan bahasa Inggeris Malaysia; (b) status Bahasa Inggeris Malaysia sebagai Bahasa Inggeris Baru; (c) keperluan terhadap pembakuan bahasa Inggeris Malaysia; dan (d) kerencaman linguistik Bahasa Inggeris Malaysia. Nampaknya, antara 3 variasi Bahasa Inggeris Malaysia, iaitu Bahasa Inggeris Malaysia Jenis 1, Bahasa Inggeris Malaysia Jenis 2 dan Bahasa Inggeris Malaysia Kolokial, Bahasa Inggeris Malaysia Jenis 2 mempunyai potensi dimajukan sebagai jenis Bahasa Inggeris Malaysia yang dapat diterima secara meluas. Kemungkinan Bahasa Inggeris Malaysia dibakukan masih kecil kerana keterbatasan keperluan terhadap jenis Bahasa Inggeris tempatan yang dibakukan untuk dijadikan model rujukan dalam pengajaran Bahasa Inggeris, di samping kedudukannya yang masih pada peringkat awal di dalam proses pembakuan bahasa. Oleh itu kajian deskriptif lanjut dan berterusan terhadap Bahasa Inggeris Malaysia adalah dirasakan perlu.

ABSTRACTThe notion of Malaysian English in relation to the concepts of standard English and language standardization is explored. An overview of the development of varieties of English spoken in Malaysia is discussed and the factors involved in language standardization such as the role of codifying agents in regulating the use of Malaysian English, the status of Malaysian English as a “New English”, the need for a standardized variety of Malaysian English and the linguistic complexity of Malaysian English are examined. Of the three varieties of Malaysian English, namely, Malaysian English Type I, Malaysian English Type II and colloquial Malaysian English, Malaysian English Type II has the potential to develop into a widely accepted variety of Malaysian English. Nevertheless, the possibility of Malaysian English being standardized is still remote due to the limited need for a standardized variety of Malaysian English as the reference model in language teaching, and its early stage in the process of language standardization. Further and on-going descriptive studies on Malaysian English are needed.

INTRODUCTIONThis paper attem pts to explore the notion of Malaysian English in relation to the concepts of standard English and language standardization. The discussion should serve as a take-off point for further empirical studies on Malaysian English, and also as reference for decisions regarding the place of Malaysian English in the ESL classroom.

Leslie, A. Oddities in Language

Rosli Talif and Ting Su Hie

Purposes of the StudyThe purposes of this study are to:(1) provide an overview of the development of

varieties of English spoken in Malaysia;(2) explore the possibility of Malaysian English

being standardized.

Development of Varieties of Malaysian EnglishIn Malaysia, English had a dom inant status during the British Administration; it was the language of the ruling class, the Christian religion and the administration (Bhathal 1990). From this setting, several varieties of Malaysian English have developed.

Until 1965 a com m on variety, Singapore- Malayan English, existed as both Malaysia and Singapore were under British rule (Platt and Weber, 1980). With the independence of Singapore, the development of Singapore-Malayan English reached an impasse due to differences in national policies regarding the status and functions of English.

In Singapore, English is not only the language of science, technology and international trade, but also a language for inter-ethnic communication and a dom inant language in the sphere of work. The 1990 Census of Population shows the literacy rate for English is highest (65%), followed by Chinese (61.5%), Malay (16%) and Tamil (3.4%) (Kwan-Terry, 1993). This is partly due to English being a compulsory language in all schools, and is one of the four official languages. Today, Educated Singapore English (EsgE, spoken by English-educated Singaporeans) is used in formal contexts (Tay 1982), and the colloquial variety, Singlish, is used informally.

O n th e o th e r h an d , in M alaysia today, “Malaysian English” can be discerned as a threetiered continuum. The impetus for the developm ent of these varieties of Malaysian English are the declaration of Bahasa Malaysia as the national and official language, and the change in the medium of instruction from English to Bahasa Malaysia in 1971 in West Malaysia, 1973 in Sabah and 1977 in Sarawak at Primary One level.

Malaysian English Type ISingapore-Malayan English is also referred to as Malaysian English Type I (ME Type I henceforth)

by Platt and Weber (1980) and is spoken by English-m edium edu cated Malaysians who were taught a British type of educated English. Baskaran (1987) describes this acrolect which is internationally intelligible as “standard Malaysian English (Gill, 1993).

A distinguishing characteristic of ME Type I is its phonology which resembles ESgE. The intonation is syllable-timed instead of stress-timed, and there is an absence of weak forms and liason (Tay 1982). Like ESgE-speakers too, ME Type I-speakers use a narrower pitch range; and are generally not aware of the fine shades of m eaning that can be conveyed by intonation in English (Tay, 1982). In addition, the pronunciation of some words in ME Type I differs from standard British English, possibly due to the influence of graphology; approximation in pronunciation (for example, / siks/ for /siksO/) and differences in how sounds of English words are perceived. However, the slight variation in phonology may no t h inder international intelligibility.

As far as syntax and formal use are concerned, adherence to a standard model of British or American English still prevails to a certain degree (Wong, 1978). For lexis in particular, there are items with a localized context such as kampung and makan, which are absent in British English. Some lexical items have different meanings. For instance, in Singapore and Malaysian English, missus is considered more polite than xvife whereas the form er is a low prestige word in British English (Platt,1980).

The use of ME Type I is on the decline. From 1962 to 1967, the enrolm ent in English medium secondary schools decreased from 90% to 69.1% (Platt 1980). Now ME Type I is used only by the older generation of English-medium educated Malaysians, and Malaysians educated overseas. More and more Malaysians are using Malaysian English Type II (ME Type II henceforth).

Malaysian English Type IIME Type II is spoken by Malaysians who are Malay- educated. With the change in the medium of instruction in 1971, there was a rapid increase in the e n ro lm e n t in M alay-medium secondary schools: 4.1% in 1956 and 30.9% in 1967. ME Type II has obvious features of interference of



Bahasa Malaysia, thus placing it further away in the continuum of international intelligibility as compared to ME Type I.

In ME Type II, the pronunciation of most words and even the spelling is sometimes influenced by Bahasa Malaysia words which originated from English words, such as akademik for academic and biskut for biscuit.

Where syntax is concerned, the word order of noun phrases in Bahasa Malaysia is used often. For instance, fonkad for phone card, and not enough tall for not tall enough. Variable marking of past tense in speech, a feature of ESgE described by Platt (1980) is found as well. For example, “Istart working here last year." It is difficult to ascertain from general observation w hether the deviant structures are merely occasional learners’ errors, or are fossilized as a feature of ME Type II. There is a high possibility of these structures becoming a perm anent feature of ME Type II as they are still intelligible.

ME Type II is making its impact in formal use such as in seminars and news broadcasts because the num ber of ME Type II-speakers in various professions is growing. W hether it would develop into the Malaysian English depends on education and language policies on the use of English.

Colloquial Malaysian English Colloquial Malaysian English (cME henceforth) is a local dialect having less complex speech forms and exhibiting m ore deviation from standard English in terms of phonology, grammar and vocabulary (Wong, 1978). cME is used by both ME Types I and II speakers in informal contexts. In more established varieties of English, it is usual for stylistic variations to occur within a speaker’s sociolect but for “New Englishes”, a sociolectal range exists instead. In Malaysia, the speech continuum ranges from ME Type I or ME Type II for formal use to cME for informal use.

The syntax of cME varies substantially from standard English. A feature common in both Malaysia and Singapore is pronoun copying like “My mother, she works very hard”. The use of fillers also p redom inates in cME such as “lah ” (“Come ^ J u r a s s ic Park is a good movie”), “one” (“The bus is always late one).” Tongue (1979) defines fillers as items of language which communicate no particular denotative meaning but which

are used to indicate affective attitudes of the speaker, or simply to fill a pause or in the stream of speech. In cME, fillers do fulfil these functions.

Code or language-switching between English and Bahasa Malaysia is a common phenom enon especially for ME Type II-speakers. For instance, “I nakpergi bank this afternoon.” Language switch is an avoidance strategy used by the learner for two purposes: (1) linguistic, that is, to avoid a difficult target language form or one that has not yet been learned, or (2) social, that is, a desire to fit in with on e’s peers (Tarone et al. 1983). As language switch in casual settings is common even among Malaysians who are proficient in English, it seems that language switch is used more for social acceptability, as has been observed by Lam, a Malaysian broadcaster (New Straits Times, 21 August 1993).

For lexis, many items are only used in the Malaysian context, such as “Please off thefan Many idiomatic expressions have been direcdy translated from the mother tongues of speakers, such as shaking legs (having a relaxing time), and spend someone (giving someone a treat). Some other slang expressions comprising Bahasa Malaysia words and phrases found in cME are koyak-lah and finish-lah to mean “I’m done for”. These features make cME internationally unintelligible.

Thus far, cME has only been used in friendship and transaction domains. It has not been used in the mass media as in the case of Singapore. Singapore Broadcasting Corporation has banned the use of Singlish in its commercials for fear of the detrimental effects of Singlish on the standard of spoken English (New Straits Times, 2 August 1993).

CONSIDERATIONS IN THE STANDARDIZATION OF MALAYSIAN ENGLISHThe preceding overview of the three varieties of Malaysian English provides a background for the discussion on the possibility of standardization of Malaysian English. To explore this possibility, the following factors are considered: (1) the role of codifying agents for Malaysian English, (2) the status of Malaysian English in relation to “New Englishes”, (3) the linguistic nature of Malaysian English, and (4) the need for a standard Malaysian English.



Before proceeding with the discussion, it is necessary to define “standard English”. Standard English with a non-localizable accent usually refers to Received Pronunciation (RP), also known as Educated Southern English, Oxford Pronunciation or Q ueen’s English, in Britain and General American in the United States (Strevens,1981). However, in Gill’s (1993) study on attitude towards suggested pedagogical models for English language teaching in Malaysia, only 49.9% of the respondents chose RP as a suitable model whereas “English spoken by an educated Malaysian with an unmarked accent, that is, an accent, which is neither strongly Malaysian nor strongly RP, and almost without grammatical mistakes” is considered suitable by 79.2%. This favoured model of English is comparable to ME Type I, but whether it could develop into the standardized variety of Malaysian English for general use depends on a multitude of other factors, which will be discussed below.

Role of Codifying AgentsFor a language to maintain its standard usage with some allowable deviance, language norms have to be enforced by various codifying agents. Examples of codifying agents are a central body like the I / Academie Francaise of France, the education system, publishers and broadcasters.

Malaysia does not have a central body set up expressly for the purpose of maintaining the standard of English. Neither does Britain. However, the education system and the mass media play an im portant role in propagating RP (Standards and Correctness in English Open University, 1982).

Where the Malaysian education system is concerned, the standard of spoken English imparted to students ranges from native-like English to cME depending on the educational background of teachers, in spite of the fact that textbooks and teaching materials are written in standard English. In Britain, teachers correct students who use language other than the prescribed norm but this is not so in Malaysia. Wong (1978) states that as there are no materials for teaching functional [Malaysian] English, standard English is still used as a model. However, this may no longer be true since the introduction of the Communicational English syllabus in 1975 (Kementerian Pelajaran Malaysia, 1975).

Besides the education system, the community in Britain exerts a form of standardizing pressure on schools to teach standard English. Parents want their children to speak proper English, and not slang or local speech (Ellis, Standard and Correctness in English Open University 1982). There is evidence that Malaysian parents are ju st as concerned. In a letter to New Straits Times dated 5 August 1993, a Malaysian parent expressed her concern over a trainee teacher who taught students to pronounce “leopard” as “lio-pat” and “thirsty” as “twisty”.

Educators may impart a variety of English to students, but the English spoken by Malaysian broadcasters is more uniform. It may serve as a model. However, like the education system, the Malaysian mass media is not prescriptive in propagating standards of oral English. In Britain the mass m edia plays a prescriptive role whereby agreed-upon pronunciation of place names, uncommon literary or scientific words, and words in common use were published in Broadcast English (Leitner 1982, p.96).

The use of written English is regulated to a certain extent by publishers, often using the dictionary as the reference source for standard usages of English term s and words. N ettle of Heinem ann Educational in “Standard and Correctness in English Open University” (1982) states that non-standard forms in manuscripts are often edited as the publishers do not want to appear “uneducated” to the reading populace. This is true of Malaysian book publishers as well.

Is Malaysian English a “New English ”?Malaysian English is briefly described in “The New Englishes” by Platt et al. (1984) but it is not cited among the list of well-known “New Englishes” such as the developing varieties of English spoken in India, Nigeria and Singapore. To find out if Malaysian English has the characteristics of “New Englishes”, the criteria put forth by Platt et al.(1984), and Foley (1988) are used.

A cco rd in g to P la tt et al. (1 9 8 4 ,)“New Englishes” are unique in the way they develop. ME Types I and II have developed in a m anner similar to other “New Englishes” as follows:

(1) they have developed through the educationsystem, through English-medium schools be-



fore independence (ME Type I), and later through Malay-medium schools (ME Type II);

(2) they have developed in an area where a native variety of English is not spoken by most of the population; Bazaar Malay was the lingua franca of Malaysians;

(3) they are used for a range of functions among those who speak or write it in a particular region, such as for communication with family members, friends and colleagues, in transactions, educational system, media, law and religion (Platt and Weber 1980);

(4) they have become localised by adopting certain phonological, lexical, syntactic or idiomatic elements from the mother-tongue languages of the region such as Bahasa Malaysia, and Indian and Chinese dialects, especially ME Type II.

Apart from the m anner in which a local variety of English develops into a “New English”, the particular variety of English must have attained a position of linguistic prom inence in the community before it can be termed as a “New English”. Based on Foley’s (1988) criteria, ME Types I and II can be termed as a “New English” because they fulfil the criteria in the following manner:

(1) English is only one of two or more codes in the linguistic repertoire of Malaysians, the others being Bahasa Malaysia and the Chinese and Tamil dialects,

(2) English has acquired an im portant status in the m ultingual com m unity of Malaysia. Though not as widely used as in the 1970’s, English still retains a place of importance in international relations, tertiary education, law and as a second language in schools.

Evaluating Malaysian English according to the criteria of Platt et al. (1984) and Foley (1988), it seems ME Types I and II have developed sufficiently for them to acquire some form of acceptance of the language norms within its community of users such that it can be accorded the status of a “New English”. Even then, the more crucial factor to consider in the standardization of Malaysian English is its linguistic property.

Linguistic Nature of Malaysian EnglishThe characteristics in h eren t in the language itself which differentiate it from dialects and even variants of the same language, also determ ine the possibility of the language being standardized to a certain extent. In this section, Malaysian English is examined to find out its standing in the process of language standardization as described byAgheyisi (1988):(1) Codification is “the process by which the lan

guage becomes enriched, stabilized and rendered comprehensively adaptable to the immediate and potential communication needs of its community of speakers” (Agheyisi, 1988). People who use language professionally and consciously are responsible for codification, and codification is presented to the speech community via grammars, dictionaries and spellers (written or oral), and finally the “standardized” variety is advanced via the governm ent, education system and mass media (Fishman 1970).

(2) “Intellectualization” of its lexicon and grammar takes place through a three-step scale, namely, (i) simple intelligibility as in conversational register, (ii) definiteness as in “workaday technical” register, and (iii) accuracy as in scientific register. Malaysian English has sim ple intelligibility but the dom ains of “workaday technical” register and scientific register are dominated by standard English. Garvin and Mathiot (1968) observe that languages which have yet to be standardized usually lack phases of the second and all of the third registers (Agheyisi 1988).

(3) The formulation of a written norm for the language, with a tradition of literary expression. In Malaysia and Singapore, it was only in 1945 that locally written English literature began to develop, the most common geYire being poetry (Platt 1980). Platt observes that the majority of writers have used standard English with few or no examples of typical Singapore- Malaysian features. Hence Malaysian English is just in the process of making its mark in the literary tradition.

It seems that Malaysian English has not undergone the essential steps in tlie process of language stand



ardization. Thus at this m om ent in time, it does not have the linguistic properties of a standard language, as described by Svejcer (1978). A standard language:

(1) has a complex interrelationship of written and oral forms. Without any systematic attempts at codification, there has only been a notion of what Malaysian English is;

(2) is the language of culture, science and journalism. In these domains standard English is used providing no room for usage and “intellectualization” of Malaysian English;

(3) possesses a norm for selected linguistic facts. Systematic codification of Malaysian English has yet to take place apart from an attem pt by Wong (1978).

4

It can be tentatively concluded that Malaysian English does not possess the linguistic properties and status of a standard language as yet. However, there is a potential for a standard Malaysian English to develop should there be a pressing need for it because it has developed far enough for it to be termed a “New English”. Ultimately, it is the language users who decide which variety of English they wish to speak, and not the standardized variety which is propagated through government policies, the education system or the mass media.

Is there a need for a Standard Malaysian English ?In order to examine whether Malaysians perceive a need for a standard variety of Malaysian English, the functions of English in Malaysia are compared with the functions expected of a standard language.

A standard language plays the f ollowing roles in the linguistic community:

(1) as a shared linguistic system for communication on codes of social conduct (Crewe 1977). English is used for a wide range of functions but only by a relatively small group of ME Type I- or ME Type II-speakers. Bahasa Malaysia is used more extensively for inter-ethnic communication.

(2) as a Prestige Variety for unifying speakers of various dialects, and for preserv ing the uniqueness of the language and its community of speakers vis-a-vis o ther related lan

guages and th e ir com m unities (Agheyisi 1988). In relation to this, Gupta (1988) states that the pre-requisites for standardization of “New Englishes” are: (i) local prestige usage (written, not informal); (ii) usage not locally stigmatized; and (iii) usage not internationally stigmatized (Foley 1988). It is uncertain whether there is prestige attached to the use of Malaysian English locally, needless to say, internationally. Instead, Bahasa Malaysia being the national language plays the role of a prestige variety in Malaysia.

(3) as a Reference Model in the Education System. During the British rule in Malaysia, RP was the official norm, and the model in the education system (Platt and Weber 1980). With the adoption of the communicative approach in the teaching of English, the aim is merely to enable students to speak intelligibly such that the communicational intent is successfully conveyed (English Language Syllabus, Kementerian Pelajaran Malaysia 1975,). In the Integrated Secondary School Curriculum (KBSM) syllabus for English language (Kem enterian Pendidikan Malaysia 1987), students are expected to speak and read using correct pronunciation, and with correct intonation, word stress and sentence rhythm. Although there is no prescriptive guideline as to what correct pronunciation and intonation is, pronouncing dictionaries like Gimson (1980) and jones (1977) are reference sources which are easily available. These sources are still useful even though according to Gill (1993), it is English spoken by an educated Malaysian with an unmarked accent (possibly ME Type I) which is favoured as a pedagogical model.

From the foregoing discussion on the perceived need for a standard Malaysian English, it is clear the need only arises in language teaching, but not as a prestige variety or for communication. Where English language teaching is concerned, English spoken by educated Malaysians with ethnic accents are considered unsuitable: 52% for Malay accent, 63.4% for Indian accent, but surprisingly 65.1% considered English spoken with a Chinese accent suitable (Gill 1993). However, Gill qualifies that this could be because the

74 Pertanika | . Soc. Sci. 8c Hum. Vol. 2 No. 1 1994


taped sample did not reflect a speaker with a strong Chinese accent, whrreas the Malay and Indian speakers had strong ethnic accents.

A lthough Malaysian English with an un marked accent is favoured as a pedagogical model in ELT, it is more likely for Malaysian English with ethnic accents (ME Type II) to develop into the standard Malaysian English, taking into consideration the trend of development of the adoptive variety of English in Jamaica (comparable to ME Type II). The adoptive variety is not overtly Creole but lacks international intelligibility due to differences in structure and meaning (Shields 1989) who argues that the ‘adoptive’ variety is likely to become the de facto target rather than the traditional Standard English (comparable to ME Type I) since mono-style speakers of the ‘adoptive’ variety are prom inent opinion makers, the press and the teaching profession. As ME Type II speakers are found in all sectors of the society, they too have the power to propagate the use of this variety of Malaysian English by consensus.

IMPLICATIONS AND CONCLUSIONSBy virtue o f the m anner of its development, its characteristics and its role in the multilingual community of Malaysians, Malaysian English Types I and II may justify a claim to the status of a “New English”. However, the utilitarian value of English has dim inished considerably with Bahasa Malaysia becoming the language of inter-ethnic com m unication and the nation’s prestige language. Thus the perceived need for a standard Malaysian English arises only in the education system.

As it is, the absence of a specified model for English language teaching has only drawn diff used concern due to the availability of various models of English, which are nationally intelligible. However, total public acceptance, especially among educationists, towards the use of Malaysian English in language teaching is still low.

W here in ternational intelligibility is concerned, it might be necessary to standardize a variety of Malaysian English so that changes in the language can be regulated for the benefit of Malaysians who are involved in international communication. However, due to the limited usage of English in Malaysia, the processes of language

standardization such as codification, “intellectual ization” of the language and formulation of a written norm are still rudimentary. Therefore the actual characteristics of Malaysian English are still unclear, resulting in a lack of uniformity in the variety of English imparted by the broadcasters, publishers and educational practitioners. The setting is such that standardization of Malaysian English is not possible, or even necessary as yet.

REFERENCESA gheyisi, R. N. 1988. The Standardization of Nige

rian Pidgin English. English World-Wide 9(2):227- 241.

B ath al , R. 1990. The Development of English in Malaysia: A Brief History. Southeast Asian Review of English 20:62-70.

C r e w e, W. (ed.) 1977. The English Language in Singapore. 208p. Singapore: Eastern Universities Press.

Fishman,J. A. 1972. Sociolinguistics: A Brief Introduc- tion. 126p. Rowley, Massachusetts: Newbury House Publishers.

F o le y , J. (ed.) 1988. New Englishes: The Case of Singapore. 234p. Singapore: Singapore University Press.

G ill,' S. K. 1993. Standards and pedagogical norms for teaching English in Malaysia. World Englishes 12(2):223-238.

G i m s o n , A. C. 1980. An introduction to the pronounciation of English. 3rd ed. 352p. London: Edward Arnold.

G upta , A. F. 1 9 8 8 . A Standard for Written Singapore English. In New Englishes: The Case of Singapore, ed. J. Foley, p. 27-50. Singapore: Singapore University Press.

J o n e s , D. 1977. Everyman's English Pronouncing Dictionary. 560p. London: Dent.

Kem enterian P eiajaran M aiaysia . 1975.English Language Syllabus in Malaysian Schools Tingkatan 4-5. 137p. Malaysia.

K em enterian P e n d id ik a n M alaysia . 1987. Sukatan Pelajaran Sekolah Menengah Bahasa Inggeris. 9p. Malaysia.

Kwan-T erry, A. 1993. Cross-currents in teaching English i n Si n gapore. World Englishes 12(1):75-84.

I a m , R. H o w well do we communicate?. New Straits Times. 21 August 1993: Appointments Page 1.

PertanikaJ. Soc. Sci. 8c Hum. Vol. 2 No. 1 1994 75


L eitner , G. 1982. The Consolidation of ‘Educated Southern English’ as a Model in the early 20th century. International Review of Applied Linguistics 20 (2) :91-107.

L eslie, A. 1981. Oddities in Language. 116p. Singapore: Federal Publications.

N ew Straits Times. 2 August 1993. “Singapore bans use of ‘Singlish’ on TV and radio”.

P. A zizah English teacher unfit to teach the language. New Straits Times. 5 August 1993:13.

P latt , J. T. 1980. Varieties and Functions of English in Singapore and Malaysia. English World-Wide 1(1): 97-121.

P latt , J. and W eber , H. 1980. English in Singapore and Malaysia: Status, Features, Functions. 292p. Kuala Lumpur: Oxford University Press.

Pia t i ', J., W eber , H. and Ho, M. k. 1984. The New Englishes. 225p. London: Routledge and Kegan Paul.

S h ie ld s , K. 1989. Standard English in Jamaica: A Case of Competing Models. English World-Wide 10(1): 41-53.

S m ith , L. E. (ed.) 1981. English for Cross-Cultural Communication. 248p. Hong Kong: The Macmillan Press.

Standards and Correctness in English Open University. Sound Recording Program me 08. 1982. Milton Keynes: Open University Education Enterprise.

Str evens , P. 1981. Forms of English: an Analysis of the Variables. In English for Cross-cultural Communication, ed. L. E. Smith, p. 1-14. Hong Kong: The Macmillan Press.

Svejcer, A. D. 1978. Standard English in the United States and England. 172p. The Hague: Mouton Publishers.

T a r o n e , E., C o h e n , A. D . and D um as , G. 1983. A closer look at some interlanguge terminology: a framework for communicative strategies. In Strategies in Interlanguage Communication, ed. C. Faerch and G. Kasper, p. 4-14. New York: Longman.

Tay, M. W. J. 1982. The Phonology of Educated Singapore English. English World-Wide 3(2): 135- 145.

T o n g u e , R. K. 1979. The English of Singapore and Malaysia. 128p. Singapore: Eastern Universities Press.

W o n g , I. F. H. 1978. English in Malaysia. ED 154624. Paper presented at Conference on English as an International Auxiliary Language, Honolulu, Hawaii, 1-15 April 1978.

(Received 22 January 1994)


Subscription RatesMalaysia/Singapore Overseas

Individual RM50 US$50Institutions RM90 US$60

ORDER FORMEnclosed is RM ___________ /U S $ _____________ for Vol_____________ No.of Pertanika Journal of Social Science and HumanitiesName:--------------------------------------------------------------------------------------------Organisation/Institution:--------------------------------------------------------------------

Address:

Country:-----------------------------------------------------------------------------------------------------------Overseas subscribers please include US$5 per issue for airmail surcharge. Cheques/Bank drafts should be made payable to Universiti Pertanian Malaysia Press, 43400 UPM Serdang, Selangor Darul Ehsan, Malaysia.

Preparation o f M anuscriptTyping and PaperThe manuscript should be typed double spaced on A4 paper with 4cm margins on all sides. It should be limited to 25 pages including tables, illustrations and references.Title PageThe tide of the paper, name of author and full address of the institution where the work was carried out should appear on the tide page. A short title not exceeding 60 characters should be provided for the running headlines.AbstractEnglish and Bahasa Melayu abstracts of 200 words each should follow the tide page. Papers from outside Malaysia may be submitted with an English abstract only.KeywordsAbout six to ten keywords are required and these should be placed direcdy above the abstract.TablesTables should be typed on separate pages and numbered using Arabic numerals. Each table should be referred to in the text, have a brief title and include explanatory notes, if necessary, below it. Vertical rules should not be used. Footnotes in tables should be designated by symbols or superscript small italic letters. Tables should conform to page size.EquationsThese must be clearly typed triple-spaced. They should be identified by numbers in square brackets placed flush with the right margin.Illustrations and PhotographsIllustrations including diagrams and graphs are to be referred to in the text as ‘figures’ and photographs as ‘p la te s ’ an d n u m b e re d consecutively in A rabic num erals. All photographs (glossy black and white prints) should be supplied with appropriate scales.Illustrations should be of prin t quality; outpu t from dotmatrix printers is not acceptable. Illustrations should be on separate sheets, about twice the size in print. All letters, numbers and legends must be included on the illustration with the au thor’s name, short tide of the paper and figure num ber written on the verso. A list of captions should be provided on a separate sheet.Spelling & MeasurementsThe Oxford English Dictionary should be consulted for spelling. M etric units m ust be used for em pirical measurements.Citations and ReferencesCitations to the literature in the text are to be indicated by the a u th o r’s nam e and year o f pub lica tion in parentheses, e.g. (Barnett and Lewis 1982). If an author is quoted in the text, the year of publication should immediately follow in parentheses, e.g. Barnett and

Lewis (1982) state that “........” Citation of a particularpage follows the date and is preceded by a comma, e.g. Humphrey 1990, p. 26-27.For works with multiple authors, the first au thor’s surname is used followed by et al. The full form of citation is used for two authors.If two or m ore works by the same au thor are cited together, the au thor’s name is followed by the various years of publication arranged in chronological order e.g. (Sulaiman 1979, 1980, 1981). In the case of citing an author with several works published in the same year, the works should be distinguished by the addition of a small letter e.g. Choa (1979a); Choa (1979b).When several authors are cited, they are to be arranged in chronological order and separated by semicolons, e.g. Zaharah 1960; Yong 1980; Lewis 1990.Serials are to be abbreviated as in the World List of Scientific Periodicals. The abbreviation for Pertanika Journal of Social Science and Humanities is Pertanika J. Soc. Sci. Hum.The following reference style is to be observed:BookShamsher M ohamed, Shamsuddin Ismail & Annuar M ohd. Nassir. 1989. Asas Belanjawan Modal. 197p. Serdang: Universiti Pertanian Malaysia Press.Chapter in Edited BookZahid Emby. 1990. The Orang Asli Regrouping Scheme - converting swiddeners to commercial farmers. In Margins and Minorities - The Peripheral Areas and Peoples of Malaysia, ed. V. T King & M.J.G. Parnwell, p. 94-109. Hull: Hull University Press.Unpublished Materials (e.g. theses, reports, documents) Shahwahid, H.O. 1989. Price competitiveness and demand behaviour of Malaysian Meranti lumber and hardwood plywood in the United States’ im port market. Ph.D. D issertation, State University o f New York, Syracuse.Ministry of National Unity. 1973. A socit>-6conomic survey on the new villages in Perak and Melaka. 67p. Malaysia.SerialsNoran Fauziah Yaakub. 1990. A multivariate analysis of attitude towards teaching. Pertanika 13(2): 267-273.Conference ProceedingsAmir Awang. 1992. Counselling, human resources development and counselling services. In Proceedings of Asia Pacific Conference on Human Resource Development, ed. Sulaiman M. Yassin, Yahya Mat Hassan Kamariah Abu Bakar, Esah Munji and Sabariah Mohd Rashid, p. 243- 246. Serdang: Universiti Pertanian Malaysia.

Pertanika Journal of Social Science 8c Humanities Volume 2 No, 1 March 1994

Contents

Projection of Sectoral Value-added: Com parative Analysis o f Alternative M ethods - Zakariah Abdul Rashid and Sabaruddin Ahmah Khair

L earning English in a non- supportive Environm ent am ong Malay Learners in Secondary Schools - Jamali Ismail

T he Relationship between Quality o f H om e Environm ent and M ental Scores o f C hildren A ttending the UPM Laboratory Preschool - Rohani Abdullah, Siti Nor Yaacob and Rozumah Baharudin

Factors Associated with N on-adoption of Technology by R ubber Sm allholders - Rahim M. Sail and Mazanah Muhamad

T he Stability and Predictability o f Betas: Evidence from the Kuala L um pur Stock Exchange - Shamsher Mohamad and A nnuar Md. Nassir

D eterm inants o f Foreign Direct Investm ent in the Malaysian M anufacturing S ec to r- Zulkam ain Yusop and Roslan A. Ghaffar

Kajian Kesesuaian dan Keberkesanan Raven SPM un tu k M engenal Pasti Kanak-kanak Melayu yang P intar Cerdas - Abd M ajid Mohd Isa

Malaysian English: Exploring the Possibility o f Standardization - Rosli Talif and Ting Su Hie

social pertanika journal of science a humanities papers/jssh vol. 2 (1) mar. 1994... · issn: 0128...

Documents