Predictors for Mathematics Achievement? Evidence From a Longitudinal Study


Delphine Sasanguie, Subfaculty of Psychology and Educational Sciences, Department of Psychology, KU Leuven Kulak, Etienne Sabbelaan 53, BE - 8500 Kortrijk, Belgium; e-mail:


Numerical processing has been extensively studied by examining the performance on basic number processing tasks, such as number priming, number comparison, and number line estimation. These tasks assess the innate “number sense,” which is assumed to be the breeding ground for later mathematics development. Indeed, several studies have associated children's performance in these tasks with individual differences in mathematical achievement. To date, however, most of these studies have cross-sectional designs. Moreover, the few longitudinal studies either use complex tasks (e.g., story problems) or investigate only one of these basic number processing tasks at a time. In this study, we examine the association between the performance of children on several basic number processing tasks and their individual math achievement scores on a curriculum-based test measured 1 year later. Regression analyses showed that most of the variance in children's math achievement was predicted by nonsymbolic number line estimation performance (i.e., estimating large quantities of dots) and, to a lesser extent, the speed of comparing symbolic numbers. This knowledge about the predictive value of the performance of 5- to 7-year-olds on these markers of number processing can help with the early identification of at-risk children. In addition, this information can guide appropriate educational interventions.