Statistical learning of across-trial regularities weakens emotion-induced blindness

Emotion-induced blindness (EIB) refers to a brief impairment in visual processing induced by the appearance of an emotional, but task-irrelevant, image. Recent research has shown that EIB is difficult to eliminate, although a few manipulations have succeeded in reducing its magnitude. However, whether EIB can be reduced or eliminated by statistical learning of across-trial regularities is not known. The present study examined this issue by introducing predictability in the timing of the targets and the emotionality and physical appearance of the distractors presented in a rapid stream of images. The results of Experiment 1 show that the typical EIB effect was eliminated when these regularities existed simultaneously, revealing an effect of statistical learning despite the absence of awareness of the regularities by the participants. Experiments 2–5 further analyzed the roles of specific regularities, and the results show that only the combined effects of predictable distractor emotionality and physical appearance can significantly weaken EIB, revealing the importance of those factors in the effect. The procedure represents one of the few successful attempts to weaken EIB, and provides important insights into the underlying mechanisms by linking EIB to other effects of statistical regularity on the allocation of attention.