How reading comprehension and age at arrival predict achievement among 1st generation immigrant students in Germany: evidence from PISA 2012–2022

Abstract Background Germany consistently shows one of the largest achievement gaps between immigrant and non-immigrant students across countries in PISA, with disparities persisting even after adjustment for socioeconomic factors. Yet international large-scale assessments typically treat immigrant students as a homogeneous group, potentially obscuring within-group heterogeneity—particularly regarding varying stages of second language acquisition—that could inform targeted intervention. Purpose Based on these considerations we investigate which migration-related factors—age at arrival, teacher-rated reading comprehension, and home language—predict academic achievement among 1st generation immigrant students in Germany, and whether these relationships remain stable across a decade of changing migration contexts. Methods Using German PISA data from four PISA cycles (2012–2022; N = 1,189 1st generation immigrants), we estimated multilevel models with students nested within assessment cycles. Achievement was measured across mathematics, reading, science, and a composite domain using 100 plausible values per student derived from IRT models. Exploratory analyses include arrival-age categories which were derived through recursive partitioning. Results Teacher-rated reading comprehension—a classroom-based functional literacy indicator distinct from PISA’s standardized reading score—was the strongest predictor of achievement across all domains. This measure predicted not only reading (R 2 = .506) but also science (.290) and mathematics (.217), indicating that reading comprehension is an essential prerequisite for understanding the text-heavy PISA tasks in science and mathematics. While teacher-rated reading comprehension emerged as the strongest predictor, age at arrival proved robust across all model specifications (β = -0.03 standard deviations per year), with cumulative effect of d = 1.15 between earliest and latest arrivers. Among socioeconomic indicators, only home possessions contributed significantly; parental education and occupational status showed no significant associations. Conclusions Reading comprehension and migration timing emerge as substantially stronger predictors of immigrant student achievement than socioeconomic background in German PISA data, while home language use shows negligible effects. These findings highlight limitations in current reporting practices: aggregating immigrant students into broad categories obscure the heterogeneity that actually drives achievement gaps. Disaggregating results by arrival age and reading comprehension would provide educators and policitymakers with more actionable insights.