Relations Between Perceptions of Parental Messages Supporting Fighting and Nonviolence and Adolescents' Physical Aggression: Beliefs as Mediators

Abstract

The purpose of this study was to examine adolescents' beliefs about fighting as mediators of longitudinal relations between perceptions of parental support for fighting and nonviolence and changes in adolescents' physical aggression. Participants were 2,575 middle school students (M_age = 12.20, SD = 1.02; 52% female; 83% African American) from the southeastern U.S. attending schools in communities with high rates of violence. Participants completed four waves of assessments every 3 months (i.e., fall, winter, spring, and summer). Each belief subscale mediated relations between perceptions of parental support for fighting and nonviolence and changes in aggression. Parental support for nonviolence was negatively associated with beliefs supporting reactive aggression and positively associated with beliefs against fighting. Parental support for retaliation was positively associated with beliefs supporting reactive and proactive aggression, and negatively associated with beliefs against fighting. Parental support for fighting as sometimes necessary was positively associated with beliefs supporting reactive aggression and beliefs that fighting is sometimes necessary. Beliefs supporting reactive and proactive aggression and beliefs that fighting is sometimes necessary were positively associated with aggression, whereas beliefs against fighting was negatively associated with aggression. Parents' support for fighting and for nonviolence may directly and indirectly reduce adolescents' physical aggression by influencing beliefs about the appropriateness of using aggression for self-defense and to attain a goal. This highlights the importance of jointly investigating multiple types of parental messages and types of beliefs about fighting.

Keywords: Aggression; adolescents; beliefs about aggression; parental messages.

Figure 1

One-Sided Path Model Representing Each Belief Variable as a Mediator of Relations Between Parental Messages and Subsequent Changes in Physical Aggression Note. Parental messages variables and belief subscales were regressed on the covariates (i.e., sex, grade, and intervention phase), but were other treated as exogenous variables (i.e., they were allowed to correlate with each other across waves, and with physical aggression at the prior and current waves). Covariates were included in the model, but are not shown to reduce figure complexity. The figure also does not show covariances across residual variances for all variables within each wave, which were included in the model.