Unlocking mathematical potential through school-based language learning: Insights from PISA 2018

Abstract This study explores the association between school-based foreign language (FL) instruction and mathematical achievement among 15-year-old students, using data from the 2018 Programme for International Student Assessment (PISA). Two complementary analyses were conducted: a large-scale model ( n = 300,656) examining the relationship between time spent in FL learning and maths performance across 73 countries and a machine learning (ML) approach (random forest (RF); n = 53,459) identifying specific programme features that most strongly influence this relationship. Results show that longer exposure to FL instruction was associated with a modest but statistically robust increase in maths scores ( β = 0.08, p < .001), even after controlling for socioeconomic and contextual factors. Among programme characteristics, the integration of multicultural curricula emerged as a prominent predictor of higher maths performance. These findings indicate that sustained, culturally enriched FL learning is positively associated with numeracy outcomes, with implications for equity in academic achievement and cross-disciplinary performance.