On the malleability of automatic attentional biases: effects of feature-specific attention allocation.

In two experiments, we examined the extent to which automatic attentional biases, as indexed by performance in the emotional Stroop task (Experiment 1) and the dot-probe task (Experiment 2), are modulated by feature-specific attention allocation. In both experiments, participants were encouraged to attend to either affective stimulus information (affective groups) or non-affective semantic stimulus information (non-affective groups). In both experiments, an attentional bias towards negative stimuli was found in the affective groups but not in the non-affective groups. In Experiment 1, we also observed an attentional bias towards non-affective semantic stimulus information in the non-affective group but not in the affective groups. We argue that these effects are due to a modulation of automatic stimulus processing by feature-specific attention allocation. Our data demonstrate that automatic attentional biases toward negative stimuli are not unconditional but instead depend on the relevance of negative information. Moreover, the results of Experiment 1 suggest that attention is automatically biased towards non-affective stimulus dimensions that are currently relevant.