The effect of leadership in implementing English as the medium of instruction on academic performance: A comparison of public and private primary schools in Karongi District, Rwanda

This study examines the effect of leadership in implementing English as the Medium of Instruction (EMI) on academic performance, comparing public and private primary schools in Karongi District, Rwanda. Supported by Bandura's Social Learning Theory and Policy Implementation Theory, the research employed a descriptive survey design using a mixed-methods approach, combining both qualitative and quantitative methods. A total of 145 participants were targeted across six schools during the study through a universal sampling technique, including 139 teachers and 6 head teachers. Data were collected through questionnaires, classroom observation checklists, document reviews, focus group discussions with 18 parents, and interviews with 6 school leaders and 1 Sector Education Inspector. Sampling techniques included simple random, stratified, purposive, and quota methods. The data were analyzed using R-square, regression, and model summary, and hypotheses were tested through descriptive statistics. The findings revealed significant inequalities in EMI implementation between public and private schools. Quantitative results indicated that school leadership strongly influences learners' academic performance in English (R = 0.540, R² = 0.291, p = 0.000), particularly when the school vision is clearly communicated (Mean = 3.57). However, practices such as daily English use in meetings and activities recorded low scores (Mean = 2.64), indicating unreliable implementation. Teachers’ English proficiency did not significantly influence learner imitation behaviors (p = 0.784), suggesting that other factors—such as teaching methods, parental involvement, and school and classroom environment—also play a role. The availability of English language resources, including internet tools, significantly impacted learners’ proficiency (R = 0.761, R² = 0.580, p = 0.000), with regression results confirming a strong model fit (F = 97.292). Additionally, the use of English as the medium of instruction had a statistically significant effect on both learner participation and academic performance (F = 5188.326, p = 0.000). Qualitative findings from observations, document analysis, and focus group discussions further highlighted gaps in policy implementation, the dominance of mother tongue in both schools and classrooms, and poor resource allocation by school leaders, particularly in public schools. The study recommends enhancing school leadership, providing regular training, monitoring EMI policy implementation, fostering an English-speaking culture across all school levels, and utilizing idle capacity and time within schools.