Measuring the effect of cognitive defusion using the Implicit Relational Assessment Procedure: An experimental analysis with a highly socially anxious sample

•Fusion interferes with behavioral fluency in both consistent and inconsistent trials.•Defusion group revealed significant decrease in consistent and inconsistent trials.•Control group revealed significant decrease only in inconsistent trials.

The current study tested the validity of the Implicit Relational Assessment Procedure (IRAP) as a tool for clinical assessment. Twenty-four students with high social anxiety were randomly assigned to either cognitive defusion or control conditions. Participants completed a self-report measure of the believability of anxiety-related thoughts and the Anxiety IRAP before and after the interventions. Significant decreases in response latency on the IRAP for both consistent and inconsistent trials were found only in participants in the cognitive defusion condition. We suggest that narrow and inflexible responding (i.e., fusion) interferes with behavioral fluency in both consistent and inconsistent blocks of the implicit measure, and thus a defusion intervention reduced response latencies in both types of blocks. In the control condition, however, we found a reduction in response latencies only for the inconsistent blocks, due to practice that occurred in the absence of a floor effect. The self-report believability measure, which produced effects for both the defusion and control conditions, showed a larger effect size for the control condition. Our results suggest that future studies that seek to analyze the impact of defusion interventions, and the psychological processes involved, employ a range of measures such as the IRAP, with a particular focus on separating out the effects of the therapeutic intervention from possible practice effects.