Associations Between Family Income and Math School Readiness: Investigating the Role of Maternal Math Support

• In Event: 2-076 - Supporting Children’s Early Mathematics Learning: Understanding the Role of the Home Environment

Integrative Statement

Children’s early math skills are important indicators of their readiness for school with large and persistent differences in these skills evident between children in low- and middle-income families from early childhood through the schooling years (Currie & Duncan, 2000; Duncan et al., 2007; Duncan & Magnuson, 2011; Feinstein & Bynner, 2004; Reardon, 2011; Shonkoff & Phillips, 2000). While considerable work has been dedicated to uncovering the mechanisms that help explain achievement differences associated with family income, generally speaking (e.g., Conger, Conger, & Elder, 1997; Yeung, Linver, & Brooks-Gunn, 2002), little attention has been focused directly on the mechanisms underlying income-related disparities in early math skills.

The goal of this study was to test the extent to which maternal support of children’s early math learning accounts for the association between household income and children’s math school readiness. Data came from the Boston, Massachusetts site of the NICHD Study of Early Child Care and Youth Development (n =140). Three measures of maternal support of children’s math learning were observed when children were 36 months of age: numerical support assessed the labeling of sets of objects, spatial support measured spatial language and gestures, and planning support measured support of mini-goal identification and modeling steps to reach the goals. Each measure has been shown to be an independent predictor of children’s early math skills (Casey et al., 2018; Lombardi et al., 2017). Other measures included family income-to-needs (assessed at 1, 6, 15, 24, and 36 months) and a latent measure of maternal characteristics, comprised of maternal verbal skills and education. The outcome of interest, children’s math school readiness, was directly assessed when children were ages 4 ½ years and in first grade with the Woodcock-Johnson Applied Problems subtest.

Structural equation models (SEM) were used to estimate direct and indirect associations. Results from a measurement model indicated that maternal support of planning and spatial concepts loaded on a separate latent term from that of maternal support of numerical learning. With these constructs separated into two latent terms, the resulting factor loadings and global fit indices for the measurement model (Figure 1) and subsequent hypothesized structural model indicated a good fit to the data (Figure 2).

Findings from the structural model indicated that greater household income was directly linked with greater maternal numerical support and greater maternal numerical support was, in turn, predictive of children’s math school readiness. As expected, maternal characteristics were associated with children’s math school readiness, but there were no links with maternal math support. Finally, an examination of indirect effects between household income and children’s math school readiness indicated that both latent measures of maternal numerical and planning/spatial support were, in combination, a significant mediator of the association between household income and children’s math school readiness (β = .29, p < .05). The discussion will focus on implications for support targeting children’s early math skills in disadvantaged families.

Authors

• Caitlin McPherran Lombardi, University of Connecticut
  Presenting Author

• Eric Dearing, Boston College
  Non-Presenting Author