Abstract
Mathematical word-problem solving requires strategic regulation and willingness to engage. This archived-data reanalysis integrated measurement evaluation and predictive analysis of Self-Determination Theory (SDT)-related perceptions, self-regulated learning (SRL), and achievement in 737 students in Grades 1–6 at one elementary school. Data included baseline and end-of-period performance assessments and an 18-item post-period questionnaire. Measurement models were evaluated, and path analyses were adjusted for prior achievement and grade. Among item-level models, the six-component model fit better than broader alternatives, but broad SRL and SDT-related factors were poorly differentiated (latent r = 0.905). The adjusted indirect association was small (β = 0.047), and questionnaire scores added negligible predictive information beyond prior achievement and grade. The weak empirical separation concerns this questionnaire and its historical scoring structure; these data cannot determine whether it reflects interconnected theoretical constructs, limitations of their operationalization, or both. Concurrent self-reports, the single-school setting, and incomplete archival documentation preclude conclusions about temporal ordering, causal mediation, developmental trends, or program effectiveness. Directionality remains an analytical choice. The study contributes a joint examination of measurement distinctiveness and incremental explained variance across elementary grades, clarifying the need to evaluate both when interpreting motivational and regulatory correlates of mathematical word-problem achievement.