Anxiety Sensitivity Shapes Whether Bystanders Step In During Bullying, VR Study Finds

A new experimental study suggests that how individuals interpret their own anxiety may be a critical factor in determining whether they intervene when witnessing bullying. Using a virtual reality (VR) simulation of social exclusion, researchers found that state anxiety can either promote or inhibit peer defending, depending on a bystander’s level of anxiety sensitivity.

The study, published in Behavioral Sciences, was conducted by psychologists at St. Francis Xavier University and involved 40 undergraduate students aged 18 to 25. Participants completed self-report measures of bullying victimization and anxiety sensitivity before taking part in a VR-based Cyberball task, a well-established paradigm used to simulate social exclusion. During the task, participants observed another player being excluded from a ball-tossing game and were given opportunities to respond with messages that could be coded as defending behavior .

The investigators were interested in whether previous experiences of being bullied predicted defending behavior, and whether this relationship operated through momentary, or state, anxiety after witnessing exclusion. They also examined anxiety sensitivity—a trait reflecting how strongly individuals fear the sensations and consequences of anxiety—as a potential moderator.

Contrary to expectations, prior bullying victimization was not significantly associated with changes in state anxiety or with defending behavior in the VR task. The hypothesized indirect pathway from victimization to defending through anxiety was not supported. However, a key finding emerged around anxiety sensitivity.

State anxiety alone showed a positive association with defending, but this relationship depended on anxiety sensitivity. For participants with lower anxiety sensitivity, increases in state anxiety after witnessing exclusion were linked to higher levels of defending. In contrast, for those with higher anxiety sensitivity, greater state anxiety predicted less defending. At average levels of anxiety sensitivity, state anxiety was not significantly related to intervention behavior.

“These findings suggest that anxiety is not uniformly helpful or harmful in motivating bystanders,” the authors noted. Instead, individuals who view anxiety as tolerable or manageable may channel it into action, whereas those who perceive anxiety symptoms as threatening may become inhibited and avoid intervening.

The study also highlights the value of VR methods for bullying research. By capturing real-time emotional responses and behaviors, the Cyberball-VR paradigm reduces reliance on retrospective self-report and allows researchers to observe how emotions unfold in the moment.

While the sample was relatively small and demographically homogeneous, limiting generalizability, the findings point to potential implications for bullying prevention programs. Interventions that broadly encourage all bystanders to step in may overlook important individual differences. Addressing anxiety sensitivity—a trait the authors note is modifiable—could help equip more individuals to act when they witness bullying or social exclusion.

Overall, the study adds nuance to the growing literature on bystander intervention, suggesting that the emotional processes underlying defending behavior are more complex than previously assumed.