Genetic and Environmental Contributions to Math Cognition

• In Event: Interdisciplinary and Cross-Cultural Investigations Into Cognitive and Motivational Underpinnings of Mathematical Development

Abstract

Recent evidence suggests that, although mathematics is associated with reading skills, individual differences in mathematics may also be due to fundamental, domain-specific deficits in numeric processing. From the behavioral genetic perspective, as was the case in the reading literature, understanding the genetic and environmental variance and covariance among measures of numeric processing may provide important theoretical insight into the etiology of the mechanisms influencing complex math outcomes, as typically assessed in classroom and testing settings. Moreover, theoretically-derived measures of math processing may better explain the significant genetic and environmental overlap/independence between math and reading outcomes. Therefore, we examined math cognition using 436 pairs of 10 – 13 year old same-sex monozygotic and dizygotic twins drawn from the Western Reserve Reading and Math Projects. Twins were assessed on a 4 hour battery of reading and math skills in their homes by separate examiners. As part of this assessment, we examined the following measures of math cognition: Addition Strategies (Geary et al., 2007), Decomposing Numbers (Mazzocco and Hanich, 2010), Fast Math (Mazzocco et al., 2008), Problem Verification (Murphy & Mazzocco, 2008), Number Line Estimation (Siegler & Opfer, 2003), and Dots (Halberda et al. 2008). Measures of accuracy and speed were obtained from each scale. We also examined Calculation and Calculation Fluency measures from the Woodock-Johnson Test of Achievement (2001). Confirmatory factor analyses suggested that these measures were best explained by a three-factor model including Accuracy, Reaction Time, and Number Estimation Factors (Chi-Square = 259.8, DF = 51, CFI = .89, TLI = .86). We then simultaneously estimated the genetic (A), shared environmental (C), and nonshared environmental (E) contributions to the variance/covariance among latent Accuracy, Reaction Time, and Number Estimation Factors. Results suggested that genetic influences accounted for approximately 60% of the variance in Accuracy, Reaction Time, and Number Estimation math factors. Shared environment accounted for 20%-30% of the variance among these factors. Roughly 10%-20% of the variance in math factors was accounted for by the nonshared environment. Turning to the associations between Accuracy, Reaction Time, and Number Estimation Factors, results suggested that genetic factors were largely overlapping for Accuracy and Reaction Time, and largely independent for Number Estimation. Shared environmental factors were almost completely overlapping, with little evidence for independence among math factors. Finally, nonshared environmental factors were largely independent. These results represent the first confirmatory factor analysis of a diverse set of math cognition measures as identified by the larger mathematics literature. Our results suggest that Accuracy, Reaction Time, and Number Estimation factors are not only correlated because of overlapping genetic and shared environmental factors, but also distinct because of independent genetic and nonshared environmental factors (which, because they are latent factors, do not include measurement error). Additional analyses are currently being conducted examining these measures in the context of reading and language skills.

Author

• Stephen Petrill, The Ohio State University - Columbus