Refining English language tests for university admission: a Malaysian example / Arshad Abd Samad, Sharifah Zainab Syed Abd Rahman, and Siti Norbaiti

English has now become the lingua franca of much of technological, business and academic endeavours. Consequently, learning the English language is now seen as vital, especially at the university level where proficiency in the language has become a selection criterion. At present, the Malaysian University English Test (MUET) has been adopted by Malaysian public universities as an indicator of English language proficiency. A student’s overall result depends
on all the four language components of the MUET and often
determines the number and nature of the English language courses
he or she has to attend at university. This study seeks to examine
whether MUET is an accurate predictor of performance and success
at university and how the MUET can be finetuned as an entry level
English language test. It was carried out among 52 third year
undergraduates of the Faculty of Educational Studies, Universiti Putra
Malaysia, Serdang, Malaysia, admitted into the Teaching English as
a Second Language (TESL) programme. The findings of the study do
not offer conclusive evidence about the validity of MUET as a predictor
of academic success. However, six models of various combinations
of scores on language components on the MUET scores are examined
in terms of their effectiveness in increasing the accuracy in selecting
students for the TESL programme. The correlations obtained using
these models indicate that the combination of various components of
the MUET can be used to more accurately predict student achievement
at tertiary level than the cumulative MUET score itself. The results of these correlations and their implications in using language tests as
admission requirements in general are also discussed.