Multilingual Competency and Academic Performance: A Machine Learning-Based Analysis of the 2022/2023 Somaliland National Primary Exam Data

Abstract

This study assesses the influence of language proficiency on academic performance in primary education within Somaliland, utilizing data from the 2022/2023 National Primary Exams. Employing a dataset of 20,638 students and applying ten machine learning regression models, the research investigates the impact of Somali, Arabic, and English language skills on overall academic outcomes. The findings reveal that proficiency in these languages significantly contributes to overall performance, with English showing the strongest positive association, followed by Arabic and Somali. Additionally, the study highlights minimal gender disparities in academic performance, aligning with previous research from the region. However, the urban-rural divide in educational outcomes remains substantial, with urban students outperforming their rural counterparts. Machine learning models, particularly Polynomial Regression, outperformed traditional methods in predicting student success, showcasing the utility of advanced analytics in educational research. These findings offer critical insights for policymakers aiming to improve language education, reduce regional disparities, and promote equitable access to quality education across Somaliland.