Virtually Unexpected: No Role for Expectancy Violation in Virtual Reality Exposure for Public Speaking Anxiety

In the current study, we examined the role of expectancy violation and retrospective reasoning about the absence of feared outcomes in virtual reality exposure therapy (VRET). Participants fearful of public speaking were asked to give speeches in virtual reality. We asked each participant individually to report their expectancies about feared outcomes in public speaking situations and which of these could be tested in VRET. Each of the expectancies was categorized as being related to: (1) participants' own reactions, (2) (overt) reactions of the audience, or (3) (covert) negative evaluation. We examined whether the proportion of testable expectancies could predict treatment outcome and which type of expectancies were evaluated as being more testable in VRET. Additionally, we experimentally manipulated retrospective reasoning about whether or not expectancies related to the overt reactions of the audience could be violated by providing verbal information after VRET about whether or not the virtual audience was interactive. A reduction in public speaking anxiety was observed from pre- to post-VRET. Treatment effects were, however, not predicted by the individually reported proportions of testable expectancies. Participants evaluated expectancies about their own reactions as being more testable in VRET compared to expectancies about reactions of the audience or about being negatively evaluated. In addition, we did not find evidence that the experimental manipulation regarding whether or not the audience was interactive influenced treatment effects. In conclusion, the results of the current study suggest that the effects of VRET are not univocally explained by the mechanism of expectancy violation.