Theses Doctoral

Cognitive Style as a Mediator between Parental Psychological Maltreatment and Depression in Adolescent Boys

Melmed, Lisa Reingold

This study investigated cognitive style (CS) as a mediator between parental psychological maltreatment (PM) and depression (DEP) in a sample of adolescent boys. Rose and Abramson (1992) hypothesized that PM affects the development of DEP more than physical abuse (PA) or sexual abuse (SA) due to its unique characteristic of being a direct attack on an individual's self-worth (e.g., "You are stupid," "You never should have been born"). To assess PM in isolation, without confounding the data with other forms of maltreatment, PA and SA were measured and controlled in this study. In addition, this is the first time a comprehensive measure of PM (Comprehensive Assessment of Psychological Maltreatment - Child Version; Brassard et al., 2003-2011) has been used to assess the relationship between PM, CS, and DEP. A sample of 169 middle to upper-middle class ninth grade boys were administered five questionnaires: the Comprehensive Assessment of Psychological Maltreatment - Child Version, the Children's Cognitive Style Questionnaire (CCSQ; Abela, 1997), the Depression Subscale of the Behavior Assessment System for Children - Second Edition - Self-Report Adolescent (BASC-2; Reynolds and Kamphaus, 2004), and the Physical Assault Subscale and a single Sexual Abuse item from the Conflict Tactics Scale Parent-Child (CTSPC; Straus et al., 1998). On these measures, 10.4 percent of the participants reported significant PM, 8 percent reported negative CS, and 12.5 percent reported at risk or clinically significant levels of DEP. It was predicted that the relationship between PM and DEP would be partially mediated by CS, when controlling for PA and SA, as existing research on the development of depression indicates numerous contributing factors (e.g., biological predisposition, negative life events). This hypothesis was tested using Baron and Kenny's (1986) four-step procedure for determining mediation and Sobel's (1982) test of significance. The results supported the prediction of partial mediation: controlling for PA and SA, CS was found to be a significant mediator between PM and DEP. Specifically, CS mediated 11 percent of the total effect of PM on DEP. Controlling for PA, PM alone accounted for 28 percent of the variance in DEP. Adding CS to the model increased the total variance in DEP to 30 percent. The results of this study suggest that PM specifically affects CS and subsequently DEP, which until now has only been speculated by researchers. Conclusions about the directionality of the data are based on a theoretical understanding that DEP and CS do not cause PM to occur. Due to this study's cross-sectional design, causality cannot be determined. Thus, the conclusions of this study must be interpreted with caution.



  • thumnail for Melmed_columbia_0054D_10996.pdf Melmed_columbia_0054D_10996.pdf application/pdf 831 KB Download File

More About This Work

Academic Units
School Psychology
Thesis Advisors
Brassard, Marla Ruth
Ph.D., Columbia University
Published Here
October 11, 2012