Individual Submission Summary
Share...

Direct link:

Poster #133 - Multigroup Comparison of Adolescent School Misbehavior and Social Cognitive Judgements

Thu, March 21, 4:00 to 5:15pm, Baltimore Convention Center, Floor: Level 1, Exhibit Hall B

Integrative Statement

Most research on adolescent classroom misbehavior views students as passive beings whose misbehavior depends on outside forces such as school conditions (Nie & Lau, 2009) or classroom management (Emmer & Gerwels, 2002). Viewing adolescents as active agents, the present study drew on social domain theory (Turiel, 1983, 2002) to examine how misbehaving and nonmisbehaving students coordinate social cognitive beliefs when thinking about misbehavior. Whereas previous domain research describes age-related universal development in moral thinking, this research focused on individual differences in the relationships between moral, conventional, prudential, and personal reasoning about acts of misbehavior and engaging in the acts.
During their 7th grade, a sample of 89 students (M = 13.07 years; 46 girls) was asked to consider misconduct across 11 classroom misbehaviors (e.g., teasing others, passing notes, interrupting teacher, etc.). Student engagement in misbehavior was assessed using self-report (1 = never engage, 3 = sometimes engage; 5 = always engage) and through teacher subjective ratings of students as members of the “misbehaving” (n=42) or “nonmisbehaving” (n=47) groups. In addition, four separate sorting tasks examined whether participants considered domain criterion to be relevant when thinking about a behavior (e.g., for the moral domain: “Could this act harm other people?”). From these tasks, we created variables assessing the degree to which each issue involved moral, conventional, personal, and prudential concerns (0 = domain not relevant, 1 = domain is relevant).
Multigroup path analysis was utilized to test the moderating role of misbehavior grouping in a model linking social cognitive beliefs to engagement in school misbehavior (Figure 1). Based on previous research (Daddis & Brunell, 2015) it was hypothesized that misbehavior grouping would moderate the effect of moral and personal reasoning on engagement in misbehavior.
A series of multigroup path analyses using AMOS was performed to test model invariance across behavior groups. A test of the unconstrained model resulted in a reasonably good fit, χ2 = 12.46, 9 df, p = .19; CFI = .94, RMSEA = .06; SRMR = .02. Compared to this baseline model, analysis of the fully constrained model revealed a significant decrease in fit, Δχ2 = 16.92, 4 df, p < .001, indicating differences between groups. Next, each path was separately constrained and compared to the baseline model. As hypothesized, there was a significant decrease in fit when moral path was constrained, Δχ2 = 14.34, 1 df, p < .001, indicating that, for the misbehaving group (but not the nonmisbehaving group), belief that the act involved harm to another was related to decreased misbehavior, β = -.61, p <.001. No other moderations were revealed. Beliefs that the act involved conventional concerns was related to a decrease in misbehavior, for both Misbehaving (β = -.37, p <.01) and Nonmisbehaving groups (β = -.33, p <.05).
Results indicated that individual differences exist in the ways that moral and non-moral social cognitive beliefs are coordinated when considering everyday moral acts. The discussion focuses on how to apply these findings in a school setting.

Author