Sibling relationship quality and problem behavior of children and adolescents: A meta-analysis

Sibling relationship quality and problem behavior of children and adolescents: A meta-analysis Introduction The sibling relationship is one of the most enduring and emotionally evocative relationships. Most sibling relationships are characterized by love and warmth as well as by conflict and rivalry. Until fairly recently, however, it was also one of the most neglected relationships in psychological practice and research. In the last decades, this has changed. Many studies have shown that siblings affect the psychosocial development of children. Generally, these studies have indicated that sibling warmth and support are protective factors in the development of internalizing as well as externalizing problems, whereas sibling conflict and hostility can be considered risk factors. However, a systematic overview of studies that have examined the link between sibling relationship quality and problem behavior is still lacking. Performing a meta-analysis enables us to statistically integrate existing research, and identify possible moderating factors. Concluding, the paper to be presented concerns a meta-analysis of the association between sibling relationship quality and child and adolescent internalizing and externalizing problem behavior. Sibling warmth, conflict and rivalry (or differential treatment) are used as indicators of sibling relationship quality. Additionally, anxiety and depression are used as indicators of internalizing problems, and delinquency and aggression as indicators of externalizing problems. Methods Several major databases have been examined: PubMED, ERIC and PsychInfo (1960-2010). Preliminary results indicate that over 40 studies met the inclusion criteria: (a) child age (0-18 years) and (b) statistical information permitting calculation of an effect size between sibling relationship quality and child problem behavior. Participants of the studies include at least 5,293 sibling pairs, mean age across studies ranges from 5.2 to 16.2 years. Each study is coded using a detailed coding system for sample, study and assessment characteristics. These variables will be analyzed as potential moderators. Examples of sample characteristics are: type of sample (community, high risk or offending), SES (lower/working class, middle class, or higher class), gender composition and age difference between siblings. Additionally, study characteristics are: study design (cross-sectional/ longitudinal), and sample size. Assessment characteristics are also coded, e.g., type of informant (child or parent), type of internalizing problems (anxiety or depression), and type of externalizing problems (delinquency or aggression). Results Preliminary analyses on 9 studies yielded modest but significant effect sizes for sibling warmth (r = -.14 for externalizing and r = -.21 for internalizing problems, p’s < .01) and conflict (r = .24 for externalizing and r = .21 for internalizing problems, p’s < .01). Significant Q statistics indicate that these studies were heterogeneous, allowing for moderator analyses. The effect size for rivalry was not significant for externalizing (r = .00, p > .05) nor for internalizing problems (r = -.02, p > .05), but Q statistics were significant, so we may examine potential moderating factors. Conclusions Our preliminary analyses show that externalizing and internalizing problems are associated with less sibling warmth and more sibling conflict. Additional studies will be coded, moderator analyses will be performed and final results and conclusions will be discussed.