Being Bullied by Same- versus Other-Sex Peers: Does It Matter for Adolescent Victims?

Abstract

The negative consequences of peer victimization on psychosocial adjustment are well documented. The consequences, however, may depend on who the bullies are. In this study, we examined the consequences of same- versus other-sex victimization. The sample consisted of 4,941 Finnish adolescents (ages 14–15; 47.7% boys). We used structural equation modeling to examine both concurrent and longitudinal associations of same- and other-sex victimization with depression, negative perception of peers, and social self-esteem. Both same- and other-sex victimization were related to psychosocial adjustment. Concurrently, the victimization experiences with same-sex peers in particular were associated with generalized cognitions about peers, whereas being bullied by other-sex peers was related to adolescents’ social self-esteem more strongly than victimization by same-sex peers. The longitudinal associations, in turn, showed that only being bullied by boys had carry-over effects on girls’ adjustment. Other-sex victimization can have serious consequences especially on girls’ psychosocial adjustment.

[Supplementary material is available for this article. Go to the publisher's online edition of the Journal of Clinical Child & Adolescent Psychology for the following supplemental resource: Complementary information on model fit indices and the nested model chi-square difference tests.]

Acknowledgments

This study is part of the KiVa project funded by the Finnish Ministry of Education and Culture, for developing an antibullying intervention program for the comprehensive schools. The project is coled by Professor Christina Salmivalli (Department of Psychology), and Ph.D. Elisa Poskiparta (Centre of Learning Research) at the University of Turku. This research was supported by two grants from the Academy of Finland (134843, 135577) to Christina Salmivalli and by a grant from NSF to Todd Little (NSF 1053160, Wei Wu, co-PI). We are grateful to all the children, their parents, and the teachers who made the study possible, and all the colleagues working for realizing the KiVa project. We also thank the KiVa seminar group for valuable comments on earlier versions of the article, and Aaron Boulton and Ihno Lee for their advice with the analyses.

Notes

¹Different parcel allocation methods can lead to changes in the parameter estimates and fit statistics (e.g., Sterba & MacCallum, 2010). In our analyses, random assignment to parcels changed the model fit indices slightly, but the results and practical conclusions made were essentially the same. We also attempted to use the original items instead of parcels; this, however, led to estimation problems.

Note: The nested model chi-square differences are obtained comparing the model to the preceding model. CFI = comparative fit index; TLI = Tucker–Lewis index; RMSEA = root mean square error of approximation; CI = confidence interval; SRMR = standardized root mean square redidual.

^a Adjustment constructs (depression, negative perception of peers, and social self-esteem) included in the same model.

^b All study constructs included in the model.

^c Separate models for depression, negative perception of peers, and social self-esteem.

²A large number of additional models were estimated to examine statistical significance of (a) the parameter estimates, (b) the differences between boys and girls, (c) the differences between same- and other-sex victimization, and (d) the differences in the direction of the effects in the longitudinal analyses. The exact model fit indices including the nested model chi-square differences tests can be found as supplementary material in the publisher's online edition of the Journal of Clinical Child & Adolescent Psychology.

Note: Underscored values mean that boys (n = 2,356) below and girls above (n = 2,585) diagonal differ from each other statistically significantly in the respective parameter estimate based on the nested model chi-square difference tests (p < .01).

^a M age = 14.5.

^b M age = 15.

^† p < .05. **p < .01. ***p < .001. The parameter estimate differs from zero based on nested model chi-square difference test.

Note: The sex difference was examined first using chi-square differences tests (Step 1, Δdfs = 1). The differences between same- and other-sex victimization (Step 2, Δdfs = 1) were examined allowing boys and girls to differ if a sex difference was found in Step 1 (p < .01).

^aSame-sex and other-sex victimization predicting adjustment differ statistically significantly (p < .01) in the concurrent analyses.

^bSame-sex and other-sex victimization predicting adjustment differ statistically significantly (p < .01) in the longitudinal analyses.

†p < .05. **p < .01. ***p < .001. The parameter estimate differs from zero based on nested model chi-square difference test.