A meta-analysis of the magnitude of biased attention in depression

In this quantitative review, we examined the magnitude of attentional biases to negative stimuli in depression. Results from 29 empirical studies examining emotional Stroop or dot probe results in depressed participants (clinical depression, nonclinical dysphoria, and subjects undergoing depressive mood induction) were examined. Studies using the emotional Stroop task yielded marginally significant evidence of a difference between depressed and nondepressed samples, whereas those using the dot probe task showed significant differences between groups (d = 0.52). We found no evidence for significant moderation of these effects by age, sex, type of depressed sample, year of publication, stimulus presentation duration, or type of stimuli (verbal or nonverbal), although statistical power for these tests was limited. These results support the existence of biased attention to negative information in depression.