Category: Adult Anxiety - Social

PS9- #A18 - Effect of Safety Behaviors on Postevent Processing in a Speech Context

Saturday, Nov 18
9:45 AM – 10:45 AM
Location: Indigo Ballroom CDGH, Level 2, Indigo Level

Keywords: Social Anxiety | Adult Anxiety


  Safety behavior use and post-event processing are implicated as two important maintaining mechanisms in the theoretical cognitive models of social anxiety. The relationship between social anxiety and safety behavior use has preliminarily been found to be mediated by anxiety control, but has not yet been examined in an experimental study. The potential mediation between social anxiety and post-event processing with perceived control has also not been explored. Furthermore, there is still controversy in the literature about a direct link between post-event processing and safety behavior use, and the types of safety behaviors that best predict post-event processing currently is uncertain. Finally, post-event processing and anticipatory anxiety have been linked theoretically but not yet established experimentally.


The current study examined safety behavior use during an impromptu speech situation, and looked at the subsequent effect on post-event processing of the event. The participants were first given a baseline battery of questionnaires. After that, they were told to give a 3 minute speech in front of a research assistant “videoing” the speech. They were then given a few more questionnaires and were excused. One week later, participants returned and were first given questionnaires for post-event processing and state anxiety. Then they were told they would have to give another speech. They rated their anticipatory anxiety for the subsequent speech task and then were told they did not have to give the speech.   


The results confirmed that greater safety behavior use led to greater reported post-event processing at both times following the speech situation. Safety behavior use was a significant predictor of post-event processing, and among safety behaviors, the restricting and active subscales on the SAFE were the only significant predictors of post-event processing immediately after the speech, and the restricting subscale was the only significant predictor for post-event processing one week after the speech. Also, both social anxiety and post-event processing significantly predicted anticipatory anxiety levels for a further speech task. Finally, perceived control mediated the relationship between social anxiety levels and safety behavior use, such that the higher social anxiety score one endorsed, the less perceived control reported and more safety behavior use in the speech. A similar mediation was found for the relationship between social anxiety and post-event processing. 


The study provides support for the current models of social anxiety and the direct association between safety behavior use and post-event processing, as well as the association between post-event processing and anticipatory anxiety. There is also support for specific types of safety behaviors having a more predictive effect on post-event processing than others. This is the first time the mediation model between social anxiety and safety behavior use with perceived control has been tested experimentally. It is also the first time the model has been applied to social anxiety and post-event processing. Implications of the study include a greater emphasis on anxiety control for people with social anxiety.

Nicolette D. Carnahan

American University
Washington, District Of Columbia

Michele M. Carter

American University