English Language Proficiency as a Predictor of Academic Performance in English-Medium Subjects

Evidence from Bhutanese Middle Secondary Students

Abstract

This study investigates the extent to which English language proficiency predicts academic performance in English-medium instruction subjects among Bhutanese middle secondary students. Drawing on data from 1,352 students across two urban and two rural schools during the 2023 and 2024 academic year, multi-linear regression analyses were conducted on the overall sample as well as on subgroups based on school location and gender. The results show that English proficiency is a strong and statistically significant predictor of academic performance across all groups, with the highest predictive strength observed among urban students (β0= 0.79, R² = 0.79) and female students (β0= 0.68, R² = 0.72). Notably, the model explained less variance in the performance of rural students (R² = 0.50), suggesting that other unmeasured factors may influence academic outcomes in these contexts. These findings reinforce existing evidence on the critical role of English proficiency in education systems where English is the medium of instruction. They also underscore the need for context sensitive policies that support language development. Such policies should aim to reduce rural-urban disparities and address unmeasured factors affecting rural students’ academic outcomes through targeted interventions.