Language Mismatch and Student Academic Achievement: Evidence from Southeast Asian Mainland

Abstract

Language mismatch in academic assessment — the mismatch between
the student's home language and the test language — is a common
situation that that may act as a barrier to educational equity in
multilingual, multiethnic societies. Extensive literature has shown that
language mismatch is associated with poorer academic performance,
particularly among young immigrant students, though factors such as
language proficiency and exposure to the second language can ameliorate
this influence. However, the issue of language mismatch in mainland
Southeast Asia appears to garner limited attention despite this region’s
rich linguistic diversity and various ethnicities. This study addresses this
gap by examining the association between language mismatch and
academic achievement among primary (Grade 5) and secondary (age 15)
students in mainland Southeast Asia, specifically Vietnam, Laos,
Cambodia, Myanmar, and Malaysia (SEA-PLM 2019), and Vietnam,
Cambodia, Thailand, and Malaysia (PISA cycles 2018 and 2022). Using
ordinary least squares regression with school-fixed effects and controlling
for individual and familial confounders, this study estimates the net
association between language mismatch and test scores in reading,
writing, mathematics, and science. The findings show that language
mismatch is associated with lower scholastic achievement, with larger
gaps found in linguistically demanding subjects like reading and writing.
Notably, Malaysia (the only country in the sample with a formally
multilingual primary education policy), exhibit larger gaps between
matched and mismatched groups than those with monolingual education
systems. This pattern does not imply that multilingual education is
inherently disadvantageous; rather, it likely reflects the abrupt transition
in the medium of instruction between primary and secondary levels in
Malaysia, which might place additional linguistic burdens on languagemismatched
students. In addition, the negative associations between
language mismatch and student achievement are more evident among primary students, and the gaps appear smaller among secondary-level
students in the PISA dataset compared with primary-level students in
SEA-PLM, though this pattern should be interpreted cautiously given that
the two datasets test entirely different student cohorts.