(PS9-B45) The Impact of Inducing Self-compassion on Safety Behavior Use During a Social Interaction on Zoom

Safety behaviors are commonly used to attempt to decrease fear in social settings but often serve to maintain anxiety. Safety behaviors can be categorized into three subtypes; avoidance, impression management, and management of physical symptoms. Self-compassion has been beneficial for decreasing other aspects associated with social anxiety (e.g., anticipatory anxiety, post-event processing). The aim of the present study was to examine the impact of self-compassion on the use of safety behaviors. Given the current COVID-19 pandemic has led to the shift of many social interactions being online, the present study examined if self-compassion inductions can decrease safety behaviors used during Zoom conversations. Although safety behaviors are most common among those high in social anxiety, the present study included analyses across levels of social anxiety to allow for severity comparisons. Recruitment is still on-going, with final results expected in April 2022. Participants (currently N= 57 undergraduate students) join a zoom call where they are asked to have an open ended “get to know each other” conversation with a confederate. Prior to the conversation, participants are asked to recall a stressful social conversation from the past that still bothers them and then they are randomly assigned to either write about the situation in a self-compassionate way or not. Next, the participants complete the conversation with the confederate for 10 minutes. After the conversation, participants complete a second survey where their safety behavior use, self-compassion, distress, and state anxiety are self-reported. Preliminary analyses indicate that participants in the self-compassion condition are reporting lower distress scores prior to the conversation compared to controls, t(55) = -1.98, p= .05. Consistent with hypotheses, the self-compassion induction led to significantly lower use of management of physical symptoms safety behaviors compared to controls, t(37.28) = -2.06, p= .04. Moderation analyses show that this difference in management of physical symptoms was present for those with moderate and high levels of social anxiety, whereas those low in social anxiety showed no significant differences. Further, moderation analyses also supported that for those high in social anxiety, inducing self-compassion led to significantly less avoidance and overall safety behaviors. However, impression management did not differ between conditions, across all levels of social anxiety. No significant differences between conditions were found on post-conversation self-compassion or state anxiety scores. Overall, current analyses show that self-compassion may have beneficial effects for many aspects of social anxiety, including reduction of safety behaviors.