Math self-concept in the transition to secondary school: Developmental trends, predictors, and educational implications among high-ability and average-ability students

This study examined the development of math self-concept during the transition to secondary school comparatively among high-ability and average-ability students in a largescale longitudinal sample in Flanders, Belgium (N = 5740 students; 49.5% males). Latent change models revealed that high-ability students, although maintaining higher relative levels of math self-concept, experienced steeper levels of math self-concept decline in the transition to secondary school than their average-ability peers, according to students' reports. Cognitive ability, measured with standardized intelligence tests, predicted steeper math self-concept decline within both ability groups, and in both groups stronger math self-concept decline was associated with (a) higher levels of teacher-reported underachievement in Grade 7 (r² = 0.13, p < .001), (b) lower standardized math achievement in Grade 11 (r² = 0.49, p < .001), (c) lower student-reported school well-being in Grade 7 (r² = 0.13, p < .001) and Grade 11 (r² = 0.04, p < .001), and (d) higher risk of grade repetition by the end of secondary school (r² = 0.11, p < .001). These findings establish that the level of math self-concept and the degree in decline in math self-concept can have negative implications for educational outcomes, even when math self-concept level remains high relative to peers.