Self-Regulation and Classroom Disciplinary Climate as Predictors of Mathematics Achievement: A Cross-National Multilevel Study in Latin America
Ricardo L. Gómez, Walter F. Castro
Source record
Source: Crossref
Published: Sep 11, 2026
DOI: 10.21203/rs.3.rs-10407142/v1
Open original source ↗Source abstract
Abstract Self-regulation is widely regarded as a foundational resource for academic learning, yet its relationship with mathematics achievement varies considerably across the diverse institutional contexts of Latin American schools. This study examined whether classroom disciplinary climate moderates the association between self-regulation and mathematics achievement at the school level, using data from the 2019 Regional Comparative and Explanatory Study (ERCE 2019) conducted by UNESCO-LLECE. The analytic sample comprised 70,914 sixth-grade students nested within 3,640 schools across 16 countries. A four-model sequence combined random-effects multilevel estimation for variance partitioning and cluster-robust regression for fixed-effects inference with all plausible values combined via Rubin’s rules. The intraclass correlation coefficient for mathematics achievement was 0.506, indicating that approximately half the total variance in achievement was attributable to schools. After adjustment for student-level socioeconomic status, gender, grade retention, and preschool attendance, self-regulation remained a significant positive predictor of mathematics achievement (B = 0.769, SE = 0.053, p < .001). At the school level, poorer disciplinary climate was associated with lower mathematics scores (B = -8.24, SE = 2.09, p < .001) even after adjustment for school socioeconomic composition, school type, and location. A significant cross-level interaction emerged (B = 0.288, SE = 0.127, p = .024): the self-regulation effect was 58% stronger in schools with poor disciplinary climate than in schools with better climate, consistent with a compensatory model of resilience. Country-level variation in self-regulation slopes (range: 0.58–2.53) indicates contextual moderators worthy of future investigation. These findings have implications for education policy aimed at reducing inequality in Latin America.
Evidence graph
No public relationships recorded yet.
Integrity note: This page is a factual metadata record created by deterministic ingestion. It is not a claim that the work moves a mathematical frontier or has been independently verified.