Intergenerational Transmission of Inequality: The Role of Working Memory in the Socioeconomic Achievement Gap in College Completion

Background and Purpose: Childhood socioeconomic disadvantage is consistently associated with lower educational attainment.  Working memory, a component of executive function that is an important predictor of academic performance, has been hypothesized to contribute to this achievement gap in two possible ways: (1) mediation, in which working memory is part of the mechanism underlying disparities in attainment; and (2) moderation, in which working memory heightens or buffers the effect of early disadvantage.  Despite the differing implications of these alternatives for prevention, no studies have examined whether working memory is a mediator or moderator of the prospective associations between childhood disadvantage and attainment of a college degree.

Methods: This question was addressed using a longitudinal study of stress and child development (n= 221 with complete data from the final wave, 47.1% female, 91.4% White).  At age 9, the family income-to-needs ratio (ITNR; M= 1.8 SD= 1.2) and maximum parental education level (33.5% college or above) was measured. During young adulthood (M= 23.5 years) college attainment was measured via self-report and working memory was assessed with three tasks - object 2-back, spatial working memory, a nonverbal sequencing task - aggregated into a single factor.  Mediation was tested using Baron and Kenny’s approach. Moderation was tested by adding an interaction term between working memory and each indicator of childhood socioeconomic status. Ordinary Least Square- and logistic regressions were used as modeling strategies depending on the distributional nature of variables. All models controlled for age in the final wave, sex, race/ethnicity, and self-reported academic competence at age 9. 

Results: Higher parental education was associated with higher odds of attaining a college degree (OR= 4.23, p= .002), but it was not associated with working memory (β= 0.11, p= .20).  While increased working memory was associated with college attainment (OR= 1.79, p= .01), its inclusion in the model did not reduce the association between parental education and attainment (OR= 4.29, p= .003).  On the other hand, family ITNR was not associated with college attainment (OR= 1.39, p= .11), though it was positively associated with working memory (β = 0.26, p= .003).  There were no interactions between working memory and either parental education (OR= 0.98 p= .97) or ITNR (OR= 0.73, p= .14). Taken together, there is no evidence of moderation nor mediation for either family ITNR or parental education.

Conclusions and Implications: Our findings suggest that working memory does not function as a mediator or moderator of the association between childhood disadvantage and educational attainment in young adulthood. Rather, increased working memory and having a parent with a college degree are independent and additive predictors of graduating from college by early adulthood.  This suggests that social work efforts to promote educational attainment, and to reduce achievement gaps and prevent the intergenerational transmission of inequality, are likely to be most effective when they are multifaceted and focus both on supporting children’s healthy cognitive development and improving children’s early environments.