This systematic review and meta-analysis examined the efficacy of Cognitive Bias Modification of Interpretation (CBM-I) to reduce hostile interpretation bias and aggressive behavior.
We searched PsycINFO, PubMed, Embase, Web of Science, and Cochrane Library databases up until January 1, 2023, and assessed risk of bias using Cochrane’s risk of bias tool (RoB 2). We used multi-level meta-analysis to synthesize effect sizes.
Fourteen articles including 16 randomized studies met the inclusion criteria, yielding data from 1449 participants. CBM-I had medium effects on hostile interpretation (d = −0.62, p < .001) and benign interpretation (d = 0.71, p < .001), and a small effect on aggressive behavior (d = −0.28, p = .017). However, effects on interpretation bias were significant only for measures that were similar to the training method, suggesting that CBM-I effects do not generalize beyond the training task. Moreover, effects on aggressive behavior were significant only for self-reported aggression, suggesting demand effects. Effects on non-self-reported aggression were small and even zero after excluding one study with an exceptionally large effect size.
There currently is insufficient evidence to use CBM-I as a treatment component to reduce aggressive behavior.