Changing the language of instruction for Mathematics and Science in Malaysia: The PPSMI policy and the washback effect of bilingual high-stakes secondary school exit exams

This longitudinal study examines a change in the language of instruction for Mathematics and Science (M & S) subjects from Bahasa Malaysia (BM) to English in Malaysia. This policy, known by the acronym of PPSMI, has two objectives: to promote student learning of M & S, and also to increase students' proficiency in English. The Education Ministry also chose to create a washback effect by introducing a bilingual high-stakes secondary exit exam.;The results point to the complexity of educational change processes. They indicate that classroom implementation of this policy is affected by multiple factors such as MST's English proficiency, the language proficiency and academic ability of their students, MST beliefs about language in teaching content, MST perception of themselves as subject teachers and the school's linguistic environment. The study found that while the English SPM exams exerted a strong washback effect throughout the two-year cycle, the BM version created a washback effect specifically among out-of-town MST at the approach of the secondary exit exam. It also found that although PPSMI places the emphasis on both content and language, MST tended to focus on teaching content. The findings contribute to the literature on educational change, washback in language testing, and CBI and bilingual education.;The study uses a framework which combines the perspectives offered by literature on educational change, content-based instruction (CBI) and washback in language testing to examine PPSMI. The research questions are interested in exploring upper secondary Math and Science teachers (MST) perceptions of the policy and the bilingual exams; in finding out what changes in MST classroom practices, how the changes occur and why; and also in determining what the washback effects of the bilingual exit exam are. It also looks at whether there are significant differences between urban and out-of-town MST in relation to all these aspects. A mixed methods approach is used for data collection and data analysis.