Dissociable Effects of Faces on Attention During Constrained and Unconstrained Viewing
It has long been proposed that some stimulus classes are so biologically important that they are automatically prioritized by the attention system, irrespective of context. However, issues of ecological validity undermine laboratory-based experiments that attempt to establish the existence of context-independent attentional biases. Here we measured attention to faces and facial expressions of emotion while participants were sitting in a waiting room before the experiment, and again in the same individuals in a laboratory-based reaction-time (RT) task. A robust bias towards images of faces was observed in the waiting room, but not in the RT task. Conversely, a robust attentional bias towards emotional faces was observed in the RT task, but not in the waiting room. Despite large individual differences in attentional biases towards face and facial emotions, measures of bias in a given individual in one setting did not predict their bias in another. We conclude that attentional capture by faces and facial emotions is highly sensitive to context.