The effect of a brief online self-compassion versus cognitive restructuring intervention on trait social anxiety.

We compared two brief online interventions of 2-week duration for individuals with elevated levels of social anxiety. Participants were randomized to self-compassion or cognitive restructuring conditions (N = 119) and assessed across five assessment points, including a 5-week follow-up. Mediators and moderators of outcome were also examined. Both interventions led to significant decreases in social anxiety (d's ranged from 0.26 to 0.58), which were maintained and improved at follow-up (d's from baseline ranged from 0.53 to 0.80). Of those who were above social anxiety cut-off at baseline (>75%), approximately 20% of participants in each group showed reliable and clinically significant changes in symptoms. No differences between the treatment conditions were found for social anxiety outcomes. Similarly, there were no measures that differentially mediated the effect of treatment condition on social anxiety. Furthermore, we did not find support for a theory-driven mediational model in which self-compassion reduced social anxiety through activation of the soothing system. Contrary to predictions and theory, neither self-criticism nor fear of self-compassion moderated the effect of the interventions. Although preliminary, the findings suggest that self-compassion techniques warrant further study as an additional means of reducing social anxiety.