Attentional biases toward social threat are a well-known phenomenon in social phobia. Recently, computer-delivered trainings have been developed to modify these patterns of attention and thereby reduce anxiety symptoms. Distribution of such attention trainings (ATs) via internet might be a promising approach in overcoming obstacles in health care utilization. However, there is no evidence supporting the effectiveness of internet-based ATs in clinical populations. The current trial examined effects of an internet-based AT on self-report measures, behavioral data and diagnostic status in individuals with social phobia (N = 56). Participants were randomly assigned to either AT using a modified dot probe paradigm or a control condition without attention modification. After training and at a 4-month follow-up, both groups showed small, albeit significant reductions in social anxiety and depression, but there was no evidence for superiority of the AT condition. The present findings question the effectiveness of internet-based ATs in social phobia. Future studies need to investigate effective variants of internet-based ATs before they can be widely applied.
