Attention and interpretation cognitive bias change: A systematic review and meta-analysis of bias modification paradigms

This systematic review and meta-analysis examines the effect of Cognitive Bias Modification for attention (CBM-A) and interpretation (CBM-I) on reducing the targeted biases and investigates moderators of each approach. PsycINFO, PsychArticles, and PubMED databases were searched for randomized-controlled studies published before March 2020 with pre- and post-CBM cognitive bias outcome measures, resulting in 91 CBM-A (n = 5914 individuals) and 70 CBM-I samples (n = 4802 individuals). Random-effects models and Hedge's g calculation showed significant medium overall effects of bias reduction with moderate to high heterogeneity (CBM-A g = 0.49 [0.36, 0.64], I2 = 85.19%; CBM-I g = 0.58 [0.48, 0.68], I2 = 70.92%). Effect sizes did not differ between approaches and remained significant after trim-and-fill adjustment for possible publication bias. Moderator variables were investigated with meta-regression and subgroup analyses. Participant age, symptom type, control condition and number of trials moderated CBM-A; student and clinical status moderated CBM-I effect size. Results support attention and interpretation modification in controlled laboratory and variable (online) training settings for non-clinical and clinical samples across various symptom types (anxiety, depression, substance use, eating disorders). Further empirical evidence is necessary to determine optimal sample and methodological combinations most strongly associated with adaptive behavioral outcomes.