Cognitive correlates of early math abilities and finger gnosis

Laburpena

We investigated whether finger gnosis were uniquely associated with different early math abilities in preschool children, whether the strength of that associationwas similar across early math abilities, and whether that association changed as a function of children’s age. We formulated a measurement model of the earlymath abilities that were evaluated. A one-factor CFA for early math abilities showed a poor fit (χ2(5) = 25.44, CFI = .90, RMSEA = .140). A two-factor model todifferentiate magnitude-based skills (F1) and counting/finger abilities (F2) showed a good fit (χ2(4) = 4.13, CFI = .99, RMSEA = .012). Finger gnosis showed alarge-size correlation with both factors, explaining (29% and 23% of the variance of F1 and F2, respectively). Including the domain-general covariates didchange the role of finger gnosis substantially and halved the correlation with both factors (χ2(22) = 33.27, CFI = .97, RMSEA = .050). The effect of finger gnosison each factor was substantially larger for younger children. Indeed, the effect of finger gnosis was not significant for older children (the 95% CI cross zero). Thisindicates that the association between finger gnosis and math vanishes over development. Figure 2 shows the interaction effect.