Trajectory Classes of Relational and Physical Bullying Victimization: Links with Peer and Teacher-student Relationships and Social-emotional Outcomes


Karlien Demol, KU Leuven, Belgium
Karine Verschueren, KU Leuven, Belgium
Isabel M. ten Bokkel, KU Leuven, Belgium
Fleur E. van Gils, KU Leuven, Belgium
Hilde Colpin, KU Leuven, Belgium


Bullying victimization, i.e., being the target of repeated and intentional aggressive acts by more powerful peers (Olweus, 2001), is a prevalent problem in upper elementary school. Increasing evidence suggests that interindividual differences in the severity of the associated outcomes result from differences in victimization experiences. A number of longitudinal studies examined heterogeneity in bullying victimization by empirically grouping together students who share a similar pattern of change. Most of these studies examined the development of victimization in general. Distinguishing between victimization forms is necessary to get a more complete picture of heterogeneity in victimization development and to shed light on the possible differential links with antecedents and outcomes.


This one-year three-wave study of the Teachers4Victims project identified patterns of development in physical and relational victimization together, and examined differences between these victimization trajectories regarding possible social antecedents (i.e., the quality of students’ relationships with peers and teachers) and social-emotional outcomes (i.e., students’ depressive symptoms and self-esteem).

A sample of 930 4th to 6th graders (55 classes, 53.1% girls, Mage = 10.55, SD = 0.90) completed self-reports about relational and physical victimization (W1-W3), depressive symptoms (W3), and self-esteem (W3). Peer nominations were used to measure the social antecedents (i.e., peer rejection and acceptance, teacher-student closeness and conflict) (W1).

Bivariate latent growth modeling was used to explore the extent to which development in relational victimization covaries with development in physical victimization and to explore both intraindividual development and interindividual differences in this development by identifying victimization trajectory classes. Both bivariate latent class growth analyses (LCGA) and bivariate growth mixture modeling (GMM) were used. Further, the modified 3-step BCH approach was used to analyze the differences between the trajectory classes concerning the antecedents and outcomes.


On average, students experienced less relational and physical victimization across the school year. Students experiencing higher victimization at the beginning of the school year reported steeper decreases. Further, students who experienced higher initial levels of relational victimization also experienced higher initial levels of physical victimization. Students who had steeper decreases in one form also had steeper decreases in the other form.

Robust evidence was found for two trajectories. Most students experienced low, decreasing relational and physical victimization. A smaller group experienced higher, generally stable victimization, more relational than physical. This class was more rejected, less accepted, reported more depressive symptoms and lower self-esteem. Teacher-student closeness and conflict were similar across the classes.


Despite the evidence for different victimization forms, the distinct trajectories showed a largely parallel development of physical and relational victimization. Low social status was found to be a risk factor for belonging to a trajectory that is characterized by stable levels of both relational and physical victimization, with higher levels of the relational form.

The findings indicate that, in order to reduce victimization and protect students’ well-being, it is necessary to start early in the school year with efforts to foster healthy relationships among students characterized by high levels of acceptance and low levels of rejection.


Link to paper > Contact > Website >
Back to Spotlights