Methods: Using a replicated sequential single-case experimental design, we examined whether EMDR therapy for flashforwards reduces social anxiety and avoidance of fear-related situations, relative to baseline, in eight individuals with social anxiety disorder. Participants were recruited from a specialized treatment center after unsuccessful CBT or because of treatment complexity (e.g., comorbidity) and were randomly assigned to varying baseline durations before EMDR therapy. Outcome variables were daily social anxiety and avoidance. Randomization tests were conducted for each case separately to test whether EMDR therapy compared to a baseline phase resulted in less daily 1) social anxiety and 2) social avoidance for individual participants and for the total sample.

Results: The preregistered randomization tests did not show statistically significant between-phase differences in daily social anxiety, daily social avoidance, or weekly social anxiety severity. Exploratory, unplanned permutation distancing tests, conducted to examine the possible influence of serial dependency, indicated lower scores during the intervention phase across all three outcomes. At the individual level, these exploratory analyses indicated reductions in daily social anxiety for six participants, in daily social avoidance for three participants, and in weekly social anxiety severity for five participants. Aggregated effect-size estimates were medium for daily social anxiety, small for daily social avoidance, and small to medium for weekly social anxiety severity.

Limitations: Limitations include modest serial dependency, absence of an active treatment control condition, absence of long-term follow-up measurements, and inclusion of one participant without a clearly identifiable flashforward, although treatment adhered to the prespecified Stage 1 treatment protocol.

Conclusions: The preregistered analyses did not provide confirmatory evidence that EMDR therapy for flashforwards reduced social anxiety or avoidance. However, the descriptive findings, effect-size estimates, and exploratory analyses showed a pattern consistent with possible reductions in social anxiety, while findings for avoidance were weaker and less consistent. These results should be considered hypothesis-generating and require confirmation in adequately powered controlled studies.

Trial registration Pre-registration: https://osf.io/f9hw8/?view_only=dec61cab812d4093bf98842eec7cefd6(opens in new tab)