Math anxiety and math achievement in primary school children: Longitudinal relationship and predictors

Background: Math anxiety (MA) and math achievement (MATH) are related, but the direction of their relationship and their predictors are still unclear. Aims: We tested whether MATH predicts MA (Deficit Theory), MA predicts MATH (Debilitating Anxiety Theory), or whether MA and MATH have reciprocal relationships (Reciprocal Theory). Further, we established whether MA at T0, T1, and T2 and MATH at T1 and T2 were predicted by gender, general anxiety, fluid intelligence, verbal and visuospatial working memory, and symbolic and non-symbolic numerical representations tested at T0. Sample: We tested 337 Polish primary school children. Methods: We analyzed longitudinal data at three time points: T0 - the beginning of the first grade, T1 - the end of the first grade, T2 - the end of the second grade. Results: MATH at T1 predicted MA at T2 and MA at T1 predicted MATH at T2 (supporting the Reciprocal Theory). Additionally, MATH at T1 was predicted by fluid intelligence, visuospatial working memory, and symbolic numerical representation; MATH at T2 by fluid intelligence, verbal working memory, and MATH at T1. MA at T0 was predicted by general anxiety and symbolic numerical representation, MA at T1 by MA at T0 and gender, and MA at T2 by MA at T1. Conclusions: The results support the Reciprocal Theory of the MA and MATH relationship. MA is predicted by general anxiety, knowledge of mathematical symbols and gender. MATH is predicted by fluid intelligence, working memory and knowledge of mathematical symbols.