Poster #64 - Socioeconomic Status and the Development of Behavior Problems: A Second Order Meta-Analysis

• In Event: 2-055 - Poster Session 06
  In Poster Session: PS 06 Section - Developmental Psychopathology

Integrative Statement

Socioeconomic status is related to the development of behavior problems at both ends of the socioeconomic gradient (Luthar, Barkin, & Crossman, 2013; Reiss, 2013; Wadsworth, Evans, Grant, Carter, & Duffy, 2016). However, socioeconomic status and its components (e.g., income, educational attainment, occupational prestige) are likely dynamic and each component may contribute to the ontogeny and persistence of behavior problems through distinct mechanisms (Assari & Lankarani, 2016; Bauldry, 2015; Evans, 2017). These complexities can pose a challenge to scholars interested in precisely elucidating how socioeconomic status is linked to the development of behavior problems (Bradley, 2016; Diemer, Mistry, Wadsworth, López, & Reimers, 2013). In this study, we conducted, to our knowledge, the first second order meta-analysis on the relation between socioeconomic status and the development of behavior problems. We aimed to (a) estimate the overall magnitude of associations between components of socioeconomic status and behavior problems and (b) examine if the overall magnitude varied between adolescents and adults, males and females, and Whites and racial/ethnic minorities. We conducted a systematic search in PsycINFO, PubMed, Google Scholar, Web of Science, and ProQuest Dissertations and Theses Global, and examined specific journals and reference lists of included meta-analyses. We identified 12 meta-analyses, including 2 dissertations, with approximately 474 primary studies and over 300,000 participants. We employed a random-effects meta-analytic approach to compute each weighted average association. We weighted each effect size by their corresponding inverse standard error, which includes both within (v) and between (τ2) study variability (Borenstein, Hedges, Higgins, & Rothstein, 2009). Unweighted and weighted averages, and indices of heterogeneity, are presented in Table 1. We found a small association between internalizing and income, r+ = -.18, 95% CI [-.31, -.04], and between internalizing and education, r+ = -.12, 95% CI [-.15, -.09]. We found a smaller average association with internalizing and overall socioeconomic status, but the confidence interval included zero, r+ = -.02, 95% CI [-.14, .10]. In relation to externalizing, we found smaller associations with income, r+ = -.02, 95% CI [-.15, .10], education, r+ = -.03, 95% CI [-.16, .10], and overall socioeconomic status, r+ = -.05, 95% CI [-.11, .01], although these confidence intervals included zero. Only the sex composition of the meta-analytic sample moderated the latter association such that predominantly male samples (i.e., greater than 50%) had a larger average association between overall socioeconomic status and externalizing, r+ = -.07, 95% CI [-.11, -.04], compared to predominantly female samples, r+ = .01, 95% CI [-.04, .06]. Moreover, among meta-analyses including only children and adolescents, overall socioeconomic status was positively related to internalizing, r+ = .04, 95% CI [.03, .06]. Our findings suggest that it is necessary to examine mechanisms by which each individual component of socioeconomic status relates to the development of behavior problems. Increasing attention on specific components of socioeconomic status is likely to help elucidate how access to particular resources are implicated in the development of behavior problems at different points in the life course across demographic groups.

Authors

• Kevin M. Korous, Arizona State University
  Presenting Author

• José M. Causadias, Arizona State University
  Non-Presenting Author

• Robert H Bradley, Arizona State University
  Non-Presenting Author

• Suniya S Luthar, Arizona State University
  Non-Presenting Author