Eriksson, Kimmo , Vartanova, Irina & O. Helenius | 2026
Intelligence
What determines differences in student achievement? Obvious candidates include individual differences in learning abilities and educational contexts at multiple levels: countries, schools, and classrooms. A fundamental developmental question is whether the relative importance of these factors changes as students mature. We develop and test competing hypotheses: some predict amplification of system-level advantages over time, while others emphasize increasing individual differentiation and institutional sorting. Using a cross-sectional cohort design across three TIMSS cohorts (2011−2023), we compare achievement variance structures in Grade 4 and Grade 8 across dozens of countries. We find a systematic reallocation of achievement variance: the proportion attributable to countries decreases substantially, while the proportion attributable to classes increases consistently—a pattern uniquely predicted by our integrated Individual Differences framework. In a second analysis, we use students' relative standing within their country as a proxy for individual learning characteristics. A formal model predicts that if country-level variance decreases relative to individual-level variance, the slope relating individual characteristics to system quality should be shallower in Grade 8 than Grade 4. We confirm this prediction: students in weaker systems need a smaller compensatory advantage in individual characteristics to reach a given achievement level in Grade 8 than in Grade 4. Both analyses converge on the same conclusion: individual characteristics become increasingly important as students mature. These findings challenge assumptions of straightforward compounding of system-level advantages and illuminate a complex interplay between individual development and educational differentiation.
