"Family Ties: Examining the Development of Children’s Math Self-Concept Using a Latent Change Score Approach"
SRCD 2019 Biennial Conference
March 21-23, 2019
Title: Family Ties: Examining the Development of Children’s Math Self-Concept Using a Latent Change Score Approach (Poster)
Session: Education, Schooling
Authors: Lara Turci Faust, Jessica Gladstone, Gregory Hancock, Allan Wigfield, Jacquelynne Eccles
Abstract: Research based in Expectancy-Value Theory (EVT; Eccles-Parsons et al., 1983) has established that students’ self-concept of ability for math predicts math achievement outcomes and choices regarding participation in math courses or careers (Meece, Wigfield, & Eccles, 1990; Simpkins, Davis-Kean, & Eccles, 2006). However, students’ developmental trajectories of their math self-concept show overall declines from first to 12th grade (Musu-Gillette, Wigfield, Harring, & Eccles, 2015).
In turn, the parent socialization model embedded within EVT (Eccles, 2007; Jacobs & Eccles, 2000) suggests that parents’ perceptions of their children’s academic ability influence children’s self-concept and thus their achievement-related choices cumulatively over time (Simpkins, Fredricks, & Eccles, 2015). Although research has found that children’s math self-concept is directly and positively related to parents’ perceptions of children’s math abilities (e.g., Häfner et al., 2016), most of this work has focused on the sole influence of mothers’ beliefs. A recent study by Gladstone, Häfner, Turci, and colleagues (2018) found that both mothers and fathers’ beliefs about their child’s ability in math uniquely contributed to children’s math self-concept, but was cross-sectional and unable to explore bidirectional effects between parents and children over time.
The present study extends previous work on this topic by exploring dynamic and longitudinal associations between mothers’ and fathers’ perceptions of their child’s math ability and children’s math self-concept from 4th through 6th grade. Specifically, we explored how parents’ and children’s scores from 4th through 6th grade predicted subsequent change in the other groups.
Data for the current study are from a subset of the Childhood and Beyond longitudinal study (Eccles et al., 1993). Participants included 205 mothers, 137 fathers, and 396 children (49.2% male). Data were collected while children were in 3rd through 5th grade. Families were predominantly White and middle class.
Proportional change parameters were significant and negative for children and fathers from grades 4 to 6, indicating a slowing in growth with respect to higher levels of perceived ability or self-concept (i.e., regression to the mean effect). Only mothers showed significant differences in proportional change over time (∆ß = -.92, p < .05). Coupling parameters show that while mothers’ early changes in ability perceptions were positively predicted by children’s math self-concept (∆ß = .18, p < .01), later changes were so strongly negatively predicted by fathers’ ability perceptions (∆ß = -.60, p < .01) that mothers’ proportional change was non-significant. Although parental perceptions were not significant leading indicators for children’s early changes in self-concept, later changes in children’s self-concept were also positively predicted by fathers’ ability perceptions (∆ß = .44, p < .01). By contrast, fathers’ early changes were positively predicted by mothers’ perceptions (∆ß = .34, p < .01), whereas their later changes were positively predicted by children’s self-concept (∆ß = .27, p < .01). Results build upon previous work to illustrate the complex familial processes that contribute toward the downward developmental trajectory of children’s math self-concept. In particular, our findings highlight the importance of exploring mothers and fathers in the same developmental model. Other theoretical as well as practical implications will be presented.