Poster #34 - Cognitive Predictors of Exceptionally High and Low Mathematical Performance in 4-5-Year-Olds

• In Event: 1-061 - Poster Session 02
  In Poster Session: PS 02 Section - Cognitive Processes

Integrative Statement

Before children start receiving formal mathematics education there are already large individual differences in children’s mathematical skills (Dowker, 2005). Several domain-specific (Merkley & Ansari, 2016) and domain-general (LeFevre et al., 2010) factors have been found to be associated with both typical and atypical mathematical development. For example, in elementary school children, working memory plays a role in developmental dyscalculia (e.g., Geary, 2004) as well as mathematical precociousness (e.g., Swanson et al., 2006). Little research is available on younger children who have not received formal mathematics education. At preschool age, we aimed to characterize children who have difficulty with early mathematics or excel in early mathematics in terms of domain-specific and domain-general precursors. We compared the numerical abilities of children who showed consistent low, average, or high math achievement. We also examined whether different profiles of numerical competency also differed in terms of cognitive abilities that have been linked with individual differences in mathematical development, namely language, spatial ability, and working memory (e.g., LeFevre et al., 2010). As early numerical competency consists of several skills with different developmental trajectories (Lyons et al., 2014) there may be different patterns of strengths and weaknesses in emerging math competencies.
Participants were selected from a larger ongoing longitudinal study comprising 410 four- to five-year-old children attending preschool in Flanders. One hundred fifty-seven preschoolers that showed consistent low, average, or high math achievement at two time points in preschool were selected. All children completed eight numerical tasks (verbal counting, nonverbal calculation, object counting, numeral recognition, number order, symbolic comparison, nonsymbolic comparison, dot enumeration) at time point 1 and 3 (Spring 2017, Spring 2018; preschool). At time point 2 (Autumn 2017; preschool) they completed a battery of cognitive tasks, including measures of language (Peabody), spatial ability (Block design), and non-verbal working memory (Corsi blocks task). Composite z-scores for numerical ability at time point 1 and 3 were calculated using the average of the standardized scores of the eight numerical tasks. Children were included in the low-performing group, if they performed below the 10th percentile on the numerical competency test battery at time point 1 and 3 (n = 20). Children were included in the high-performing group, if they performed above the 90th percentile at both time points (n = 15). We further selected a control group (average) including children who performed between the 25th and 75th percentile at both time points (n = 122).
An ANOVA revealed significant group differences on all tasks (i.e. numerical and cognitive) as shown in Table 1. Post-hoc analyses revealed a similar pattern of results for all tasks: the performance of the control group was significantly above that of low-performing children, but significantly below that of high-performing children, except for symbolic and nonsymbolic comparison at time point 3. Binary logistic regression analyses with the cognitive variables as independent variables revealed significant associations between children’s language and non-verbal working memory ability and group membership. The data indicate that the same cognitive factors are important for low as well as high mathematical performance.

Authors

• Merel Bakker, KU Leuven
  Presenting Author

• Joke Torbeyns, KU Leuven
  Non-Presenting Author

• Bert De Smedt, KU Leuven
  Non-Presenting Author