Bidirectional effects between working memory and math in children with and without math disabilities: Does working memory or math drive the system?

Inefficient working memory (WM) processes underlie poor math performance in children with math disabilities (MD). However, the assumed directional relationship between WM and math performance in children with MD has not been tested. Recent studies with large heterogeneous samples of elementary school children show a bidirectional relationship between working memory and achievement (math in this case) in which achievement (i.e., math) plays the dominant role in later WM development. In this study, we investigate the bidirectional relation between WM and math in three cohorts across three grades (Cohort 1 = grades 1 to 3; Cohort 2 = grades 2 to 4; Cohort 3 = grades 3 to 5) of children (N = 651) over three years. Within these cohorts, children with MD (n = 103) were compared to children without MD (n = 446) and high math achievers (n = 175). The results showed that: (1) the pattern of within-subject changes and between-subject changes varied between children with and without MD, (2) age-related increases in math proficiency drew extensively upon WM resources, and (3) model testing showed a directional influence of WM on math for children without MD and high achievers, but not for children with MD. Performance for children with MD reflected a weaker connection between WM and math at the initial stage than in the later stages of learning. Taken together, the results suggest that the dominance of math predictions on later development of WM is not necessarily due to prior math knowledge or a reduction in WM resources, but to the weaker growth of WM relative to math growth.