Print   |   Save

Learners’ Perceptions Towards The Teaching Of
Science Through English In Malaysia
A Quantitative Analysis

Sopia Md Yassin, Sultan Idris Education University (Malaysia)
David Marsh, University of Jyväskylä (Finland)
Ong Eng Tek, Sultan Idris Education University (Malaysia)
Lai Ying Ying, Kemabong Secondary School (Malaysia)


This article describes the development and validation of a questionnaire, PATSIE (Perceptions and Attitudes Towards the Teaching of Science in English), which compares Limited English Proficient (LEP) and non-LEP (NLEP) pupils’ perceptions towards the teaching of Science through English. It was found that PATSIE, with a Cronbach’s alpha measuring at 0.8, consisted of four dimensions, namely Attitude Towards Science in English, Usage of English, Support in Learning English, and Problems in Learning English. Data collected were analyzed using multivariate analysis of variance to gauge main group (LEP versus non-LEP) differences and any interactional effects in relation to gender.  Results of this study indicate that the main group effect was significant across the first three dimensions. However, there was no interactional effect for group and gender. This study seeks to make an important contribution to the issues surrounding the teaching of Science through English in the Malaysian education system as PATSIE reveals previously unreported aspects of pupils’ perceptions of their experience of learning through English.


Keywords: LEP, PPSMI, Science, CLIL, quantitative analysis


Teaching of Science and Mathematics in English TeSME, more frequently referred to as PPSMI (Pengajaran dan Pembelajaran Sains dan Matematik dalam Bahasa Inggeris) is a policy involving the change of medium of instruction for these two subjects from Malay which is the National Language of Malaysia to English. The implementation of PPSMI seems to have resulted in several issues and problems that have required attention. This article reports on part of an on-going study that focuses on the demographic profiles and perceptions of learners with limited English proficiency.

PPSMI was implemented in stages beginning with the 2003 schooling session with Year 1 (primary school), Form 1 and Lower 6 (secondary school) learners. The scale by which this policy has impacted on the education system is considerable with approximately 5 421 158 learners in the primary and secondary schools (EPRD, 2007) directly involved.

Several reasons underpin this major change in the language policy. The ability to compete in the era of globalization; the government’s concern aboutthe nation’s human resource capital in the knowledge economy society; the knowledge and information explosion in science and technology with English as the most important global lingua franca (Gill, 2005); and the nation’s quest to become an education hub in the region, were some of the pressures to which the government was responding in 2002. Gill (2007) quotes former Prime Minister Tun Dr Mahathir Mohamad on the reasons for the change in policy:

We need to move from the extreme of nationalism which concentrates on being a language nationalist only, not a knowledge nationalist, not a development oriented nationalist. I feel that we should be a development oriented nationalist. We want our people to succeed, to be able to stand tall, to be respected by the rest of the world. Not to be people with no knowledge of science and technology, very poor, very backwards, working as servants to other people. If we have no knowledge we will be servants to those with knowledge.

(Gill, 2007: 119)

Page 54
Page 55

The enshrinement of Malay as the National Language in the Constitution of Malaysia in 1981 was successful in establishing Malay as the national language in education, administration and regional communication. However, competence in English, especially amongst the Malay population in rural areas was considered weak (Abdullah, 2005) whilst the non-Malays become bilingual (e.g. Tamil-English, or Mandarin-English), or more often trilingual (e.g. Tamil-Malay-English or Mandarin-Malay-English). This linguistic disadvantage facing some Malay graduates is reported to have led to a problem in their obtaining jobs in the private sector (Gill, 2007).

Six years after its implementation, PPSMI continues to receive strong opposition from various groups which are adamant about reverting the policy. By early 2009 the Ministry of Education is expected to have made a critical decision about whether or not to revert back to the use of Malay language as the medium of instruction for Mathematics and Science. Several series of expert and other stakeholder discussions were held during 2008 (Hishammuddin, 2008), alongside the analysis of research findings. These have raised issues on a range of key educational and socio-political concerns which have led to both positive and negative positions being taken towards the use of English as a vehicular language. Some of these are now considered in the following section.

Issues and Challenges of PPSMI

There is a considerable amount of research reported on the teaching of Science through English, with a substantial body of literature available. Whilst not being exhaustive, this article includes relevant reporting relating to discussion concerning PPSMI drawn from conferences, proceedings, journal articles, and unpublished postgraduate theses. An examination of the research reviewed (based upon the frequency of reporting and consistency of topic), points to a set of issues and challenges which includes the language competency of both teachers and learners; teacher professional development; use of purpose-designed teaching courseware; and achievement levels and support.

Studies have been carried out on teachers’ understanding and awareness of the purpose of PPSMI at various levels (primary and secondary) in different settings (urban and rural) and among different groups (pre-service and in-service). These studies indicate that teachers appear to accept the purpose and implementation of PPSMI. Teacher-readiness is at an intermediate to high level despite average English language competency with some even admitting their lack of proficiency (Norzita, 2004; Kon, 2005). Teachers are said to generally understand the English language, but lack the oral skills to teach subject matter through English (Kon et al., 2005).

A number of articles discuss the language problems faced by secondary students. These include students’ language needs (Chan, 2003), lack of vocabulary and confusion with certain words (Hashimah, 2003) and difficulty in understanding non-scientific terms in the scientific context (Saidi, 2004). Testimonies from practitioner’s on coping strategies to overcome these problems are described. Findings from the research on this issue are discussed by Noraini (2006) on the need for understanding learner problems in using the English language.

Studies conducted on learner English language competencies have also been comparedto parallel performance in Science. A recent large scale study conducted by Isahak (2008) involving 3 903 Year 5 pupils in their fifth year of PPSMI reveal the following:

  • 75% do not or barely comprehend teaching in English and find it difficult to learn
  • 80% of teachers used code switching
  • performance in Science and English was poor with an average score of 4.08/14.0 and 11.87/31.0 respectively.

Findings from this study indicate that there is a sizeable group of pupils who have low proficiency in English along side not performing well in Science. These figures give cause for concern and are quite different from the recent announcement of the Primary School Assessment Test of 2008 which is a national standardized test held at the end of six years of primary schooling. The 2008 result formed the first cohort who had undergone PPSMI for six years. Results showed a 4.4% increase (compared with the national average for the past five years) in the number of pupils who scored an A for English. It reports that 46.6% of pupils have confidence in answering questions in English for Mathematics (31, 1% for Science). This compares to 0.2% for Mathematics and 0.3% for Science the year before (Chong, 2008).

Page 55
Page 56

Apart from English, the pupils did significantly better in composition with their first languages namely Malay, Chinese and Tamil. There was also an increase in the number of pupils scoring A in all subjects involving a total of 46 641 out of 518 616 pupils who sat for the exam. Results for the Mathematics and Science papers, however, showed a 2.2% decrease among pupils who scored A, B or C, while Science scores dropped by 0.7% compared with the average over the last five years. As evident from the issues discussed so far, there may be weaknesses in the implementation of PPSMI.

To facilitate the change in the medium of instruction, the English Language Training Centre (ELTC) has, since 2003, been given the responsibility for developing and conducting an English Language enhancement programme known as English for the Teaching of Mathematics and Science (ETeMS). ETeMS is regarded as an urgent interim measure, besides several other support mechanisms to ensure that mathematics and science teachers have the basic capacity to use English as the medium of instruction (Sharifah Maimunah, 2003).

This is complementary to other on-going professional development courses for pre-service and in-service teachers involved in PPSMI. Chan (2003) recommends that both language and content teachers work together to facilitate the teaching of Science through English. A survey on those teachers who had undergone pre-service and in-service training found that a substantial majority agreed that the training they received had prepared them to speak in English, and understand science reading materials in English (Noraini, 2007). Further analysis however, revealed that both their pre-service (44.3%) and in-service (31.4%) training could not develop their confidence in speaking English. About 84.7% of the respondents also reported the need for training in helping the learners to learn through English. Although teachers perceived that they are professionally prepared to teach Science through English, they still report that they need more preparation in overcoming learner’s difficulties in the language, especially for those who are weak in English or Science, or both.

Besides on-going ETeMS courses the Ministry of Education has had teaching courseware developed for the primary, secondary, Form Six and matriculation levels. Guidelines for developing this courseware, which has to be in tandem with the current curriculum specifications, incorporate a focus on pedagogical principles and thinking skills. Related research mainly focuses on the developed teaching courseware and on innovations using other forms of technology to cater for variations in learning styles. Surveys on the overall usefulness of this courseware found that the majority of teachers agreed that it is effective and assists them in coping with teaching Science through English (Peh, 2003; Noraini, 2006). However, some aspects and components were lacking in relation to higher order thinking skills and appropriate learner knowledge levels. Observation of 21 teachers’ instructional practices revealed minimal interaction between teachers and learners (Koh, 2006). Teachers were found to be mainly using the passive ‘click and show approach’ to explain science concepts using the voice-over found in the courseware. Some teachers who were not using the courseware cited reasons such as it slowed down their teaching and not having enough time to cover the syllabus. However, other researchers discuss the potential of the courseware materials (Kuldip, 2003) as tools for interactive online teaching-learning. Esther (2007) describes the potential for improvement of communication skills through the use of mobile phones, on-line forums and email.

Continuous support from administrators has contributed to the smooth implementation of PPSMI and helped sustain both policy and the demands and expectations of various stakeholders. Interviews and a survey of school principals conducted by Aminuddin (2003) and Noraini (2006) respectively, report their willingness to implement the change in policy successfully. These school principals were found to not only to play a vital role in the change process, but also serve as role models for teachers and learners. Although parents were found to support their children’s learning in English, their level of commitment varies according to different socioeconomic background (Noraini, 2006).

Page 56
Page 57

Need For Research on Limited English Proficient (LEP) Pupils

The review of the research discussed so far has raised a major concern about the English language proficiency of both teachers and learners. This research has not given specific attention to problems faced by learners, or specific techniques for supporting groups of learners who have difficulty in using English to learn. This line of enquiry is relatively new within the context of PPSMI. Science teachers need to be cognizant of the perceptions brought by learners to the teaching and learning situation. This enables teachers to seek and implement ways to successfully respond to learners’ language related problems. In addition to learners’ perceptions, attitudes and problems towards the use of English and family support and background are found to be important factors for teachers to draw upon in order to make instruction more meaningful and relevant.

Since PPSMI was implemented by the Ministry of Education in 2003, focus has been given to the challenges faced by teachers. A large portion of an allocation of more than RM 5 billion (approx. EUR 1.05 billion) was used for ETeMS training, developing teaching courseware, monetary incentives and provision of new teaching and learning materials such as laptops, LCDs and textbooks. These initiatives were devised to facilitate and assist teachers while less attention was given to issues faced by learners. This study takes into account the learners’ perceptions towards learning Science through English in view of the finding that the mere equalization of educational resources is not enough to achieve parity of results.

Although teacher readiness is recognized as critical in preparing for the transition from Malay to English-medium, an equally important factor concerns specific types of learners such as LEP which can enhance or hinder the teacher’s ability to achieve the goals of PPSMI. This study aims to develop and validate an instrument (PATSIE) that can be utilized to describe and understand learners’ perceptions towards their own experience of learning Science through English. It is hoped that by combining this study’s findings on learners’ perspectives with findings from other studies that learning issues can be clarified.

According to Sutman (1993), teaching science and English language skills simultaneously to Limited English Proficient (LEP) pupils is an excellent vehicle for second language development. Based upon initial statistics of pupils identified as LEP in this study, the actual number throughout the nation is significant standing at approximately 41.15% of the student population. This is a direct result of present practice such as streaming pupils based on academic achievement and a heavily exam-oriented system that tends to cluster pupils into low ability classes. Streaming from the age of 6-7 years can have an adverse impact on learners which reduces their ability to realize their academic potential and impacts directly on their ability to perform successfully in Science through English.

For many pupils, although English is learned as a second language from Year One, it can be considered a foreign language, especially among Malay pupils in the rural areas. Describing the profiles of these pupils and understanding their perceptions towards PPSMI is considered important in developing strategies that enhance overall learning.

Page 57
Page 58

Research Project

The focus of this study on LEP learners’ follows comments made by educators that one outcome of PPSMI is the existence of a group of learners who find their classroom experience incomprehensible because they fail to cope adequately with the use of English as the vehicular language. The question arises if it is the English language, or other factors, which affects their understanding of science; and consequently if they would learn more successfully if Malay were the vehicular language.

To understand this phenomena, a research grant for the year 2007 entitled “Content and Language Integrated Learning (CLIL): Teaching Science for Understanding to Limited English Proficient (LEP) Pupils” (Malaysian LEP CLIL project) was approved by the Ministry of Higher Education. One of the objectives of this research project is to identify and develop exemplary programmes that can offer LEP learners access to the same kinds of challenging curricula and learning as are available to pupils already proficient in English.

Content and Language Integrated Learning (CLIL)

PPSMI can be considered as a type of CLIL. CLIL is an educational methodology which has been developed throughout the European Union. This development has mainly focussed on raising levels of multilingualism. New working educational models have often involved a shift from fragmentation towards integration, resulting in examples of integrated curricula. For example, some topics within the sciences have evolved as cross-disciplinary curricular subjects such as the environmental sciences, which draw on a range of previously fragmented separate subjects such as mathematics, physics, biology, and chemistry, amongst others.

At the same time the influence of ‘constructivism’ has become increasingly significant in relation to teaching and learning methodologies. Although for many years it has been recognized that transmission of knowledge from adult to child represents a very poor educational model, this remains a typical teaching approach found in many countries. The principles of the constructivist approach, drawn largely from cognitive psychology, are summarized by Bruner (1973: 44):

  • Instruction must be concerned with the experiences and contexts that make the student willing and able to learn (readiness).
  • Instruction must be structured so that it can be easily grasped by the student (spiral organization).
  • Instruction should be designed to facilitate extrapolation and or fill in the gaps (going beyond the information given).

CLIL is one example of curricular integration which is heavily reliant on learning by construction rather than primarily through instruction. It is a dual-focused educational approach in which an additional language, such as English, is used for the teaching of both content and language. It specifically focuses on reducing the language burden which can seriously harm concept formation, and which is a common feature of classroom life where children learn through the medium of a language where individual linguistic competencies may be low. It does this through simultaneous focus on content, language, and thinking skills (cognition). In terms of language development it involves ensuring that learners can use the types of language structures required for achieving higher order thinking skills even if their overall linguistic competence is restricted.

The introduction of CLIL methodologies, which now extends beyond Europe, is directly relevant to the Malaysian context. In theory, the introduction of teaching Science and Mathematics through English in 2003 was based on achieving the type of systemic change which would be of benefit not only to the learning of Science and Mathematics, but also English. But as in Europe, sometimes what may be a sound educational blueprint for achieving constructive change may not permeate through to the interface between policy and practice, namely the classroom.

Page 58
Page 59

This problem in converting policy into practice has surfaced in some educational contexts where systemic weakness has prevented educational innovation from being realized at the teacher-learner interface. In describing how the implementation of CLIL needs to be carried out, Mehisto (2008) comments on the challenges resulting from disjuncture. CLIL programme implementation often causes disjuncture – a tension between one’s current way of doing things and a new approach. Disjuncture can serve as a learning opportunity or invoke defensiveness and rejection.

Many teachers find it difficult to apply a multiple focus on content and language, as well as on cross-curricular integration, cognition, and reflection. A climate of high-stakes exams that can contribute to a reduction in autonomous decision-making by teachers; a lack of knowledge about CLIL-specific strategies and their impact on learning, as well as on exam results; teacher mindsets; and, a need for better planning by teachers and government authorities all impact on CLIL. Any of these factors can knock a program off balance.

(Mehisto, 2008: 111)

In European CLIL contexts much work has been done on ensuring that LEP-type learners are able to benefit from the experience of learning through an additional language. For large-scale CLIL types such PPSMI to succeed it is considered essential that learners have access to the content-obligatory language for thinking about the content which fosters creative and critical thinking. This type of language is subject and content-specific. Different types of content require very specific types of English usage which need to be embodied into the teaching context, especially the materials, and used systematically by teachers and learners alike. If the learners do not have access to, and understanding of, how these words and phrases operate then they will not be able to use the language to successfully understand the content. This issue of ‘language formula for learning’ is summarized by Coyle (2005: 8) as follows:

  • Content matter is not only about acquiring knowledge and skills. It is about the learner constructing their own knowledge and developing skills;
  • Content is related to learning and thinking (cognition). To enable the learner to construct the content, it must be analysed for its linguistic demands;
  • Thinking processes (cognition) need to be analysed for their linguistic demands;
  • Language needs to be learned which is related to the learning context, learning through that language, reconstructing the content and its related cognitive processes. This language needs to be transparent and accessible;
  • Interaction in the learning context is fundamental to learning. This has implications when the learning context operates through the medium of a foreign language;

Against this background, the research project was designed to help educators effectively integrate language and content in the Malaysian context. The focus of CLIL towards content mastery and language development with a third aim on the development of learner’s thinking skills is reported as a key success factor in contexts similar to those of Malaysia (see Marsh, 2008). As a methodological approach, CLIL is directly relevant to the teacher professional development goals identified through the Malaysian LEP CLIL research project.

Page 59
Page 60

This four-phased research project employs a quantitative and qualitative methodology incorporating a survey, classroom observation and interview. This article presents the findings of the exploratory phase of the research project, namely to develop an understanding of the demographic profiles and problems of LEP pupils. More specifically, this quantitative analysis focusing on learners’ perceptions towards the learning of Science through English aimed to find out the following:

  1. Identification of a single dimension or multiple dimensions underlying the 20 PATSIE items.
  2. The demographic profiles of LEP and NLEP learners.
  3. Any main group and gender effects, as well as group and gender interactional effect on each of the following factors:
  1. attitudes towards Science in English
  2. usage of English
  3. support in learning English
  4. problems in using English

Identification of LEP and NLEP Pupils

LEP learners’ performance has resulted in the development of special instructional programmes which focus on ways to assist learners in acquiring English language skills where English is the vehicle of instruction (Cornell, 1995). Unfortunately, the existing definition for classifying LEP and non-Limited English Proficient (NLEP) learners is not uniform and is based on different criteria in different locations (Anstrom, 1996; Liu et al., 1997; Abedi, 2004;). Comparisons of several definitions by Anstrom (1996) found a common criterion, in which a LEP learner is one who has a language background other than English, and whose English language skills deny him/her the opportunity to learn successfully in the English medium. These academically unsuccessful learners do not have any physical or physiological disabilities, but perform below their expected level of achievement. They are often faced with problems such as poverty; a home environment that does not encourage learning; or negative attitudes toward school (Valdez et al., 2002).

Disagreement on other criteria being used to identify LEP learners centresaround which skills (speaking, listening, reading, and writing) are used to assess English language ability and what English proficiency levels are necessary for learners to be classified as LEP. Some of these definitions do not indicate the cut-off point on specific assessment instruments, raising problems in data collection due to varying interpretations. The results of studies from six different areas by Abedi (2004) on the classification and inclusion of LEP learners suggest that using multiple criteria including valid and reliable measures of learners’ English proficiency, and achievement tests, provides a more consistent decision-making process for LEP inclusion.

Based upon a broad understanding of LEP from the literature, the LEP learners identified in this study satisfies the two most common criteria used by state government and researchers (Sutman, 1993; Liu et al., 1997; NCLB, 2001; Valdez et al., 2002; Ravi, 2007; Trowbridge et al., 2008). The respondents come from an environment where a language other than English, Malay is dominant. The first language of the respondents is either Malay, Chinese or Tamil.

Additionally, the School Based Oral Assessment (SBOA) and School Based Assessment (SBA) were used to gauge learners’ proficiency in both the English and Malay language. The SBOA (comprising constructs such as grammar and vocabulary, pronunciation and intonation, fluency and rhythm) has been used since 2003. It is a continuous assessment process (for Years 1-6) which aims at encouraging and enhancing language skills in both Malay and English. Learners are assigned grades A to D representing an excellent to a minimum ability in language skills. For example, the descriptors used for the construct grammar and vocabulary are:

Grade A: no grammatical errors, excellent use of vocabulary and clause in context.

Grade D: major grammatical errors, incorrect or inappropriate use of vocabulary and clause.

Based upon the descriptors used to measure different language proficiencies given on each construct, LEP pupils are identified as having grades C or D for SBOA in English language and A or B for Malay language, and vice-versa for NLEP pupils (refer to Table 1).

Page 60
Page 61

The SBA for English and Malay language is administered each semester. SBA complements the SBOA in assessing the language proficiencies of learners. Although respondents in this study represent different geographic locations of the country, they go through a common syllabus that determines coverage of learning outcomes with test items developed according to the Primary School Achievement Test format which is a national level standardized test in Year 6. The 40th percentile was used to identify LEP and NLEP pupils based on their performance in the SBA for Malay and English. Table 1 summarizes the criteria used to identify LEP and NLEP pupils in this study.

Table 1.

Summary of Criteria Used to Identify LEP and NLEP Pupils
Language Proficiency and Achievement Tests LEP NLEP
School Based Oral Assessment (SBOA) English language C or D A or B
Malay language A or B A or B
School Based Assessment
English language Below 40th
Above 40th
Malay language Above 40th
Above 40th

Data Collection

Prior to administering the questionnaire, class teachers from the schools in the zones identifiedcompleted the data on respondents’ performance on the SBOA and SBA for both Malay and English language. Once the class teacher had identified respondents that meet the criteria in Table 1, a date was set to administer PATSIE.

The questionnaire was administered by a team of researchers from 13 September - 7 November 2008. Respondents in this study were Year 4 pupils drawn from 16 different schools randomly selected to represent four different zones of Peninsula Malaysia (Table 2). These respondents were pupils enrolled in the national primary schools representing diverse backgrounds. Even though respondents were drawn from different zones, they followed the same content of the Primary Science Integrated Curriculum specified by the Curriculum Development Division. A total of 1 048 respondents were sampled as LEP and NLEP pupils. The significance of studying LEP pupils is considerable. For example, in the Malay majority state of Kelantan there are more LEP than NLEP pupils.

Table 2.

Frequency and Percentage of Respondents According to Group and Location
Zone No. of Schools No. of Classes LEP NLEP Total
Kelantan 3 7 116 (52.5%) 105 (47.5%) 221
Melaka 3 13 103 (33.9%) 201 (66.1%) 304
Kedah 5 12 122 (41.2%) 174 (58.8%) 296
Perak 5 11 84
143 (63.0%) 227
Total 16 43 425 (40.6%) 623 (59.4%) 1048 (100%)
Page 61
Page 62

Development and Validation of PATSIE

The PATSIE questionnaire developed to measure learners’ perception on the teaching and learning of science through English involved 51 items. In the pilot study these were based on five initial hypothetical constructs, namely language, attitudes, practice, environment and instruction. These a priori constructs emerged from the research on the learning of science through languages other than the first language. The items were then checked for clarity and readability appropriate for primary pupils, before the 51-item questionnaire was administered to a group of 358 Year Four pupils from the state of Malacca.

The dataset was subjected to several factor analyses, each time specifying a different number of factors as suggested by Tabachnick and Fidell (1996). When a 3-, 4- and 5-factor model was explored, the adequacy and plausibility of extraction favoured the 4-factor model. With the removal of items that have double, triple or quadruple loadings and also items which have low item-total correlations (i.e., less than 0.3), the remaining 15 items loaded persuasively into 4 coherent factors. These were initially labelled as Attitudes Towards Science in English (3 items), Usage of English (6 items), Support in Learning English (3 items), and Problems in Using English (3 items).

Table 3.

Structure for the PATSIE Factors: Item No, Item Wording,
Factor Loading, Eigen Values and Cronbach’s Alpha
    Factor Loadings
Item No. Item Wording 1 2 3 4
A12 I hope to be able to learn Science in English in the university. .75      
A13 Learning Science in English is useful for my future. .71      
A14 I like learning Science in English. .68      
A16 I look forward to learn Science. .65      
A15 My English has improved because Science is taught in English. .62      
A17 I like collecting materials related to Science in English. .44      
U2 I understand television programs in English without reading the subtitles.   .70    
U1 I understand the speeches given in English during school assembly.   .68    
U6 I study on my own to understand Science.   .58    
U4 I like to read in English.   .54    
U3 I can read instructions and guides written in English.   .45    
U5 I take part in competitions or activities conducted in English.   .45    
S7 My parents have spent time helping me to improve my English.     .74  
S9 My parents are concerned about my problems in learning Science.     .67  
S8 My parents encourage me to read in English.     .54  
S11 My parents buy learning materials in English whenever I need it.     .52  
S10 I receive rewards whenever I get good grades for Science.     .50  
P19 I understand English but am too shy to use it.       .79
P20 I avoid using English because of fear in making mistakes.       .74
P18 I can understand what is taught but find it difficult to answer in English.       .68
Eigenvalues 4.54 1.93 1.37 1.21
% of variance 22.70 9.63 6.83 6.03
Cronbach’s Alpha 0.76 0.67 0.61 0.62

Extraction Method: Principal Component Analysis.
Rotation Method: Varimax with Kaiser Normalization.

Page 62
Page 63

A refined version of the questionnaire was administered to 1 048 Year Four pupils learning Science through English from 16 national primary schools representing four zones throughout Malaysia. The dataset collected was analysed to see if the four constructs established in PATSIE were further supported.

Principal Components Analysis (PCA) with varimax rotation was conducted to assess the underlying structure of the 20 items of the PATSIE questionnaire. Varimax rotation was chosen because it produces “factors that are unrelated to or independent of one another” (Bryman & Cramer, 1998: 284) and hence, “are easy to interpret” (Brace et al., 2003: 304). Four factors were requested simply because the items were designed to index four constructs which emerged out of the pilot study. After rotation, the first factor accounted for 22.70% of the variance; the second factor accounted for 9.63%; the third factor accounted for 9.83%; and the fourth factor accounted for 6.03%. Table 3 displays the items and factor loadings for the rotated factors, with loadings less than 0.40 omitted to improve readability and clarity.

The first group of six items is factorially distinct, all with loadings greater than 0.40 on Factor 1. Taken together, this group of items which forms Factor 1 reflects positive attitudes towards the learning of Science through English. Therefore we have retained its label as “Attitude Towards Science in English”. The second factor, which seems to index the usage of English in and outside of school, is composed of six items with factor loadings between 0.45 and 0.70. Therefore, the second factor retains its label as “Usage of English”. The third factor is most strongly associated with five items (items 7, 8, 9, 10 and 11) with factor loadings between 0.50 and 0.74. Taken together, this group of items reflects the parental support in the learning of English. Therefore, this factor retains the label of “Support in Learning English”. Finally, the fourth factor is strongly associated with 3 items with factor loadings of 0.79, 0.74, and 0.68. Taken together, these three items reflect the problems encountered while learning Science through English. Therefore, the fourth factor retains its label as “Problems in Using English”. The overall internal reliability, established using Cronbach’s coefficient alpha, was measured at 0.8, which can be claimed as a high value which also indicates that the 20 items in PATSIE have high internal consistency.

Statistical Analyses

This study used a survey method to gather learners’ perceptions towards the teaching of Science through English. Data collected from the questionnaire were analyzed using the Statistical Package for Social Sciences (SPSS) Version 14.0 software. Various statistical analyses were employed to aid the interpretation of responses from 1 048 respondents. Application of several statistical procedures was necessary to support the validation of conclusions from which to derive implications for improving teaching and learning practices based upon the perceptions held by learners in this study.

Descriptive statistics using frequencies and percentages were used to describe the distribution of samples according to the variables identified in Section A of the questionnaire. Since the data in Section A consisted of frequencies in discrete categories, the Chi-square test was used to determine the significance of differences between LEP and NLEP groups of learners. This was done to determine whether group was a significant factor in the home language used; use of bilingual and English dictionaries; use of extra tuition; and use of computers to learn English.

Because of the multidimensionality of PATSIE, which consisted of 20 items in Section B, a multivariate analysis of variance (MANOVA) was chosen using a two-way 2 x 2 (group x gender) to gauge the main group (LEP versus non-LEP) difference and its interactional effects with gender. Further analysis using univariate ANOVA was conducted to indicate significant effects on the four factors of PATSIE.

Page 63
Page 64


Demographic Profiles of LEP and NLEP Pupils

The results in Table 4 indicate that group was a significant factor in relation to gender (χ2 = 4.77, p = 0.029). There were more female respondents in NLEP as compared to LEP. Group (LEP vs. NLEP) was also a significant factor in relation to the number of languages used (χ2 = 60.90, p < .001). More NLEP respondents use more than one language as compared to LEP respondents. Equally, group was a significant factor in relation to the use of a bilingual dictionary (χ2 = 7.82, p = 0.020) as well as solely an English dictionary (χ2 = 8.86, p = 0.012). More NLEP respondents use a bilingual dictionary and also solely an English dictionary as compared to LEP respondents.

In relation to English tuition, group was a significant factor (χ2 = 16.51, p < .001) where there were more LEP respondents who take English tuition as compared to NLEP respondents. Finally, group was also a significant factor in relation to the use of computers in learning English (χ2 = 24.47, p < .001). Here, more NLEP respondents use a computer to learn English as compared to LEP respondents. In summary, there were markedly more NLEP students as compared to LEP students in terms of using more than one language; using bilingual (English and Malay) and English-only dictionaries; and using computers to learn English. While there were more females amongst the NLEP respondents, more LEP respondents take private or out-of-school English tuition as compared to the NLEP respondents.

Table 4.

Crosstab and Chi Square by Selected Variables
Among LEP and NLEP Pupils
χ2 p
Item N % N %
Male 205 48.2 258 41.4 4.77 .029
Female 220 51.8 365 58.6
Home Language            
1 Language 352 82.8 375 60.2 60.90 .000
>1 language 73 17.2 248 39.8
Use Bilingual Dictionary            
No 79 18.6 107 17.2 7.82 .020
Yes 341 80.2 516 82.8
No response 5 1.2 0 0
Use English Dictionary            
No 252 59.3 350 56.2 8.86 .012
Yes 168 39.5 273 43.8
No response 5 1.2 0 0
Attend English Tuition            
Yes 274 64.5 348 55.9 16.51 .000
No 146 34.4 275 44.1
No response 5 1.2 0 0
Use Computer            
No 297 69.9 362 58.1 24.47 .000
Yes 123 28.9 261 41.9
No response 5 1.2 0 0
Page 64
Page 65

The analysis of the dataset, which was accessed from the school records, indicates that the NLEP students scored significantly higher than the LEP students in terms of science achievement (t=19.20, p < .001), English proficiency (t=10.82, p < .001), and Malay proficiency (t=41.65, p < .001).

Multivariate Analyses of Selected Variables

As shown in Table 5, the MANOVA indicates that there was a significant main effect for group (p = .000) on the overall perceptions and attitudes towards the teaching of science through English (i.e., the linear combination of four factors in PATSIE). The proportion of variance on the four-factor scores that can be accounted for by group is 7.3%. However, there was no significant main effect for gender (p = 0.07) on the combined factors. The two-way group and gender interaction was not significant (p = 0.239). Accordingly, discussions on the gender and its interactional effects on each of the factors could be ruled out.

Table 5.

MANOVA Summary of PATSIE by Selected Variables
Among LEP and NLEP Pupils
Effect Pillai’s Trace
F Hypothesis
p η2
Group 0.073 20.56 4 1041 .000* .073
Gender 0.008 2.16 4 1041 .072 .008
Group x Gender 0.005 1.38 4 1041 .239 .005
* significant at p < .05

Follow-up univariate ANOVA (see Table 6) on each factor using the Bonferroni adjusted alpha of .0125 indicate that the effect of group was significant for the first three factors, namely attitudes towards the teaching of science in English, usage of English, and support in the learning of English. The mean score from the self-rating of NLEP students was significantly higher than the mean score of NEP students across the first three factors (see Table 7). However, the main group effect was not significant for the fourth factor, namely, problems in using English.

Table 6.

Means and Standard Deviations of PATSIE as a Function of Group and Gender
Group N Attitude Towards Science in English Usage of English Support in Learning English Problems in Using English
Males 205 13.87 2.41 11.52 1.96 10.54 2.21 6.31 1.52
Females 220 13.68 2.34 11.20 2.07 10.71 2.32 6.26 1.51
Total 425 13.77 2.34 11.35 2.02 10.63 2.27 6.29 1.51
Males 258 14.37 2.36 12.55 2.16 11.29 2.04 6.62 1.68
Females 365 14.60 3.16 12.52 2.15 11.41 2.27 6.27 2.05
Total 623 14.50 2.86 12.54 2.15 11.36 2.18 6.62 1.91
Page 65
Page 66

Table 7.

Effects of Group and Gender on Four Factors of PATSIE
Source Dependent Variable df F η2 P
Group Attitudes towards Science in English 1 17.72 .017 .000*
Usage of English 1 79.24 .071 .000*
Support in Learning English 1 26.42 .025 .000*
Problems in Using English 1 2.04 .002 .153
Gender Attitudes towards Science in English 1 0.01 .000 .920
Usage of English 1 1.76 .002 .185
Support in Learning English 1 1.11 .001 .292
Problems in Using English 1 3.14 .003 .077
Group x Gender Attitudes towards Science in English 1 1.51 .001 .219
Usage of English 1 1.20 .001 .274
Support in Learning English 1 0.04 .000 .835
Problems in Using English 1 1.79 .002 .182
* significant at p < .0125


The study sought to document the experiences of learners and to ascertain their perceptions towards the teaching of Science through English in Malaysia. Perceptions were solicited from 425 and 623 Year Four pupils identified as LEP and NLEP respectively. Results of the present study have important implications from an instructional and social viewpoint for teachers, administrators and parents. The significance of this study also lies in the provision of insights into factors associated with LEP learners.

The analysis provides strong evidence to support the multidimensionality of learners’ perceptions and the factorial validity of PATSIE. The data reduction of factors in PATSIE is used as a basis for further analysis in this study. PATSIE, as a valid instrument, can also be used to shed light on relationships that may exist between variables. Its user-friendly format gives PATSIE potential as a diagnostic tool, or an evaluative tool, in the identification and instruction of LEP and NLEP learners.

This survey is part of an ongoing Malaysian LEP CLIL research project on learning Science through English. The PATSIE, developed and validated in this study, is part of an effort to discern how learners perceive their experience at the end of their fourth year of learning Science through English in Malaysia.

The significant results among LEP and NLEP on questions related to respondents’ home language, use of dictionary, computers, and English-tuition, are useful criteria to better understand LEP students and how to support them.

Being the first of its kind in developing an instrument to gather Malaysian learners’ perceptions, this study also provides a method for identifying LEP learners. Besides increasing the validity of LEP classification, the use of learners’ background variables can be used as a guide in crafting lessons for increased participation and success in learning (Abedi, 2004). A significant percentage of learners belong to the LEP group, and it is hoped that the findings from this research may help educators not only understand them better, but also appreciate the complex influences underpinning appropriate educational solutions.

Results from the comparison of perceptions between LEP and NLEP pupils are significant. NLEP learners have significantly more positive attitudes towards Science in English, greater parental support, and experience of using the English language than LEP learners. However, both groups faced similar problems in using English.

Page 66
Page 67

Creating a non-threatening and conducive English speaking environment within the school community is crucial in encouraging pupils to use the language. It is a way of helping pupils to overcome psychological barriers to speaking in a second language. It can help build their confidence in using the language. This conducive environment can be created through various classroom and extra-curricular activities.

The significantly lower perception among LEP learners on three factors of PATSIE indicates the need for learning environments to involve pupils in inquiry based settings rather than traditional transmission modes. The challenge is for schools to provide LEP learners with appropriate instructional strategies. This involves shifting the mind-sets and expectations of teachers and administrators. Teachers need to gradually accommodate diverse levels of English competency among LEP learners with enough support to ensure their mastery of concepts in Science. This is best done when teachers take into account learners’ attitudes, and also group differences that impact on instructional effectiveness.

The Malaysian National Philosophy of Education (NPE) endorses the concept of developing an individual’s potential through education in a holistic and integrated manner (Ministry of Education, 2005). Despite the ideals stipulated in the NPE, current practices are generally neglecting the needs of LEP learners, including those in PPSMI. Learners’ limited proficiency in English should not be viewed as impeding their participation in PPSMI.


Abdullah, H.:2005, Language planning in Malaysia: The first hundred years, English Today, 21(4), 3-12.

Abedi, J. and California Univ., L.: 2004, Inclusion of Students with Limited English Proficiency in NAEP: Classification and Measurement Issues, CSE Report 629, Center for Research on Evaluation Standards and Student Testing CRESST, (ERIC Document Reproduction Service No. ED484039), Retrieved June 26, 2008, from ERIC database.

Aminuddin, M.: 2003, Administrators’ perceptions of the support they can offer to Mathematics and Science teachers teaching in English, Paper presented at the TED-ELTC ETeMS Conference 2003: Managing Curricular Change, Retrieved November 11, 2008, from conferences/8_Abstracts.pdf

Anstrom, K.: 1996, Defining the Limited-English Proficient Student Population, Directions in Language and Education. (ERIC Document Reproduction Service No. ED404874), Retrieved May 21, 2008, from ERIC database.

Brace, N. Kemp, R. and Snelgar, R.: 2003, SPSS for psychologists: A guide to data analysis using SPSS for Windows (2nd ed.), Basingstoke, Hampshire, Palgrave.

Bruner, J.: 1973, Going Beyond the Information Given, New York, Norton.   

Bryman, A. and Cramer, D.: 1998, Quantitative data analysis with SPSS for Windows: A guide for social scientist, London, Routledge.

Chan, K. F.: 2003, The global affair: ETeMS, Paper presented at the TED-ELTC
ETeMS Conference 2003: Managing Curricular Change, Retrieved November 11, 2008, from conferences/8_Abstracts.pdf

Chong, T.: 2008, Keputusan UPSR lebih baik [Better UPSR Results], Utusan Online, Retrieved November 14, 2008, from

Cornell, C.: 1995, Reducing failure of LEP students in the mainstream classroom and why it is important, Journal of Educational Issue of Language Minority Students, 15. Retrieved June 26, 2008, from

Page 67
Page 68

Coyle, D.: 2005, Developing CLIL: Towards a Theory of Practice, Monograph 6, APAC, Spain

Esther Gnanamalar, S. D.: 2007, Science teachers please SMS for soft skills: The changing interface for learning in Malaysian science teacher education, Paper presented at 2007 International Conference on Science & Mathematics Education: Redefining Learning Culture of Sustainability, Penang, Malaysia.

Educational Planning and Research Division.: 2007, Malaysian Educational Statistics 2007, Ministry of Education, Malaysia, Retrieved November 8, 2008, from

Gill, Saran K.: 2005, Language policy in Malaysia: Reversing direction, Language Policy, 4(3), 241-260.

Gill, Saran K.: 2007, Shift in language policy in Malaysia: Unraveling reasons for change, conflict and compromise in mother-tongue education, AILA Review, 20(1), 106-122.

Hashimah, Z.: 2003, Dilemma of Form one students in learning science in English: my personal experience, Paper presented at the TED-ELTC ETeMS Conference 2003: Managing Curricular Change, Retrieved November 11, 2008, from conferences/8_Abstracts.pdf

Hishammudin, T.H.: 2008, Weighing all the pros and cons. Sunday Star, Retrieved December 16, 2008, from

Isahak, H. Abdul Latif, H. G. Md Nasir, M. Abdul Halim, I. and Mariam, M. N.: 2008, Kesan dasar pengajaran Matematik dan Sains dalam Bahasa Inggeris di sekolah rendah [The effects of using English as the medium of instruction for Mathematics and Science in primary school], (UPSI Research Code 03-12-95-05), Universiti Pendidikan Sultan Idris, Perak, Malaysia.

Koh, C. K.: 2006, A qualitative analysis on classroom practices of 27 PPSMI teachers in Sarawak. Unpublished PhD thesis, University of East Anglia, UK.

Kon, Y. H.: 2005, Teaching efficacy beliefs of primary school teachers in using English to teach Mathematics and Science, Jurnal IPBA, 3(2), 45-49.

Kon, Y. H. Low. B. Y. Chong, P. W. and Mohanakrishnan, K.: 2005, Think Science, speak English: A study of selected year one primary teachers’ English oral fluency in the teaching of Science, Jurnal IPBA, 3(2), 78-83.

Kuldip, K.: 2003, Learning objects in technology-related instruction, Paper presented at the TED-ELTC ETeMS Conference 2003: Managing Curricular Change, Retrieved November 11, 2008, from conferences/8_Abstracts.pdf

Liu, K., Thurlow, M., Erickson, R., Spicuzza, R., Heinze, K.: 1997, A Review of the Literature on Students with Limited English Proficiency and Assessment. State Assessment Series: Minnesota, Report 11. (ERIC Document Reproduction Service No. ED415641) Retrieved January 8, 2009, from ERIC database.

Marsh, D.: 2008. CLIL in primary East Asia Contexts: Primary Innovations in East Asia: Indonesia, Malaysia, Thailand, Philippines. British Council, East Asia.

Mehisto, P.: 2008, CLIL Counterweights: Recognizing and Decreasing Disjuncture in CLIL, International CLIL Research Journal, Vol.1, 96-117.

Ministry of Education.: 2005, Integrated Curriculum For Secondary Schools Curriculum Specifications Science Year Four, Curriculum Development Centre, Kuala Lumpur, Malaysia.

U.S. Department of Education.: 2002, No Child Left Behind Act of 2001.Retrieved June 26, 2008 from pg107.html#sec9101

Page 68
Page 69

Noraini, I. Loh, S. C. Norjoharuddeen, M. N. Ahmad Zabidi, A. R. and Rahimi, M. S.: 2006, Implementation of English in teaching mathematics and science, Retrieved July 4, 2008, from navigation/academics/faculties/FDD/awards/ceder/

Noraini, I. Loh, S. C. Norjoharuddeen, M. N. Ahmad Zabidi, A. R. and Rahimi, M. S.: 2007, The professional preparation of Malaysian teachers in the implementation of teaching and learning of Mathematics and Science in English, Eurasia Journal of Mathematics, Science & Technology Education, 3(2), 101-110.

Norzita, M. D.: 2004, Tinjauan terhadap kesediaan guru dalam pelaksanaan pengajaran dan pembelajaran Sains dan Matematik dalam Bahasa Inggeris. Unpublished master’s thesis, Universiti Kebangsaan Malaysia, Kuala Lumpur, Malaysia.

Peh, T. K. A.: 2003, A study on the perceptions of Form 1 science teachers in Melaka Tengah towards the quality of prescribed courseware, Paper presented at the TED-ELTC ETeMS Conference 2003: Managing Curricular Change, Retrieved November 11, 2008, from conferences/8_Abstracts.pdf

Ravintheran, K.: 2007, Effects of culture, environment and language on limited English proficiency (LEP) students, Unpublished master’s thesis, Universiti Pendidikan Sultan Idris, Perak, Malaysia.

Saidi, S. and Zurida, H. I.: 2004, Masalah pembelajaran bahasa bukan saintifik dalam pembelajaran sains, Diges Pendidik, 4(1), 32-43.

Sharifah Maimunah, S. Z.: 2003, The crucial role of English in the implementation of the “Teaching of Mathematics and Science in English” policy with highlights on support programmes in English, Paper presented at the TED-ELTC ETeMS Conference 2003: Managing Curricular Change, Retrieved May 15, 2006, from conferences/7_05DrSharifahMaimunah.pdf

Sutman, F.: 1993, Teaching Science Effectively to Limited English Proficient Students, ERIC/CUE Digest, Number 87, (ERIC Document Reproduction Service No. ED357113), Retrieved May 21, 2008, from ERIC database.

Tabachnick, B. G. & Fidell, L. S.: 1996, Using multivariate statistics (3rd ed.). New York, Harper Collins.

Trowbridge, L. W. Bybee, R. W. and Powell, J. C.: 2008, Teaching Secondary School Science: Strategies for Developing Scientific Literacy (8th Ed.), New Jersey, Pearson Education, Inc.

Valdez, G. Svedkauskaite, A. and McNabb, M.: 2002, Critical issue: Mastering the mosaic-framing impact factors to aid limited-English-proficient students in Mathematics and Science, Retrieved June 26, 2008, from 

Zaidi, Y.: 2003, Support for Science and Mathematics Teachers in The Implementation of PPSMI: Challenges Ahead and Strategies to Sustain The Momentum. Paper presented at the TED-ELTC ETeMS Conference 2003: Managing Curricular Change, Retrieved May 16, 2006, from