Aggression and Victimization in a Late Elementary School Sample: Identification and Relation to Friendship
Open Access
- Author:
- Temkin, Deborah A
- Graduate Program:
- Human Development and Family Studies
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- May 07, 2012
- Committee Members:
- Scott David Gest, Dissertation Advisor/Co-Advisor
Scott David Gest, Committee Chair/Co-Chair
Dana Lynn Mitra, Committee Member
Mark T Greenberg, Committee Member
Meg Leavy Small, Committee Member
D Wayne Osgood, Committee Member - Keywords:
- aggression
victimization
LCA
friendship
bullying - Abstract:
- Although research on peer aggression and victimization has greatly expanded in the last several decades, it is often limited by its tendency to only use single-informants with a-priori determined classifications, to limit findings to either aggression or victimization but not both, and to explore relations to friendship primarily among early and late adolescents but not younger children. This dissertation aimed to (1) utilize latent class analysis to aggregate self-, teacher-, and peer-reported aggression and victimization into statistically derived and validated classifications of aggressors and victims; and (2) explore friendship dynamics among both LCA and cut-score identified aggressors and victims to test existing theory regarding friendship relations and aggression and victimization with an early elementary school population and expand theory to those who are both aggressive and are victimized. Data were from 538 youth from 28 third- and fifth-grade classrooms in a highly urban state and a more rural state, assessed at three time-points over the course of a school year: early (September/October), winter (January/February) and spring (May/June). Latent class analyses revealed a five-class solution for aggression and victimization, with expected classes for uninvolved, aggressors, victims, and aggressive victims and an additional class for moderate aggressors. The LCA successfully aggregated the only moderately concordant data for self-ratings and peer-nominations into a valid aggregated class. The moderately aggressive class had moderate levels of all behaviors and contained more girls and more individuals from the more urban state. Reasons for this finding are explored. Models testing friendship relations with aggression and victimization were mixed and often inconsistent with previous findings and sometimes inconsistent between LCA and cut-score identified classes. Having non-victimized friends was protective against victimization only for cut-score identified victims during the period between the early and winter assessments. Friendship was strongly protective for aggressive victims. Having victimized friends or aggressive friends did not increase risk for becoming victimized or engaging in aggression behavior. Being victimized and being aggressive both reduced numbers of incoming friendship ties. Reasons for differences between these findings with younger youth and previous findings with older youth, differences between cut-score and LCA identified classes, and differences between earlier and later waves are all explored. Overall, this dissertation demonstrates the need to closely consider how aggressors and victims are identified and the need to expand research to test existing theories on friendship, aggression, and victimization on younger children and on those who are both aggressive and are victimized.