Parents’ Differential Treatment of Siblings, and Child and Adolescent Problem Behavior: A Meta-Analysis

• In Event: 1-012 - The Impact of Siblings on Problem Behavior among Adolescents

Integrative Statement

Past research (Tamrouti-Makkink et al., 2004) has documented that parents’ differential treatment (PDT) of siblings has implications for problem behavior among children and adolescents (Scholte et al., 2007). The research in this area, however, has examined PDT in a variety of ways and domains. As one example, some studies examine absolute levels of PDT that focus on the amount of differential treatment exhibited by parents. Other studies examine relative levels of PDT that assess whether “favored” siblings have different outcomes than those who are “less favored.” These different approaches have important methodological and theoretical implications, but it is unclear how each approach may differ in its associations with problem behavior. Enough studies have been conducted that a thorough meta-analysis of the literature may help elucidate patterns linked with different approaches to assessing and measuring PDT.
Multiple online databases were searched for published articles, theses, and dissertations on topics relating to PDT. Search terms included non-shared environments, parental favoritism, parental differential treatment, differential parenting, within-family differences, siblings, and sibling favoritism. Top sibling scholars were also asked for unpublished studies or raw data. Studies were included if they employed quantitative data on some aspect of PDT and problem behavior in childhood or adolescence (aggression, delinquency, or externalizing behaviors). The search returned 242 effect sizes from 12 published articles, 4 theses or dissertations, and 4 raw data sets. The average sample size from these studies/data sets was 265.40 (SD = 215.61) and on average the participants were in late childhood (M age = 10.79, SD = 4.10; range 3.51-17.17).
Three-level multi-level meta-analysis models were examined in the package “metafor” in R. Bivariate correlations were used as the effect size outcome. Analyses were conducted separately for effect sizes for the absolute level of PDT and for relative PDT.
Results for relative PDT are presented in Table 1. The unconditional bivariate correlation between relative PDT and problem behavior was -.08. This suggests that across studies, children who are favored engage in less problem behavior. Results from the covariates suggested that this association was closer to zero when the PDT was reported on mother and father combined, rather than mother alone, or when responding on PDT regarding family resources.
Findings for absolute levels of PDT are presented in Table 2. The overall unconditional bivariate correlation between the absolute level of PDT and problem behavior was .11, suggesting that on average, at the bivariate level, when parents showed greater differences in treatment, their children engaged in more problem behavior. Results of the model with covariates suggested that the association was amplified by age, in samples with a greater proportion of younger siblings, and when parents reported on PDT. Findings also suggested that the correlation was closer to zero when the PDT was regarding a father, if it came from raw data, or a difference score was used to assess the PDT.
The discussion will focus on the theoretical and methodological implications of the overall effects of PDT and the more nuanced findings with various covariates/moderators.
 

Authors

• Alexander C. Jensen, Brigham Young University
  Presenting Author

• Savannah Kroff, Brigham Young University
  Non-Presenting Author

• Brittany Pollard, Brigham Young University
  Non-Presenting Author

• Veronica Hanna-Walker, Brigham Young University
  Non-Presenting Author

• McKell Jorgensen, Brigham Young University
  Non-Presenting Author