Abstract
Visual attention is biased towards stimuli portraying emotion. Recently, Gupta, Hur, and Lavie (2016; Emotion) demonstrated that positive valence stimuli have a unique effect on attention; whereas high perceptual loads can prevent negative task-irrelevant stimuli from interfering with task performance, they do not prevent interference from positive task-irrelevant stimuli. Across two experiments, we examined the generalizability of this finding by testing whether positive task-irrelevant stimuli capture attention even when inconsistent with an observer’s attentional control set (ACS). Participants in Experiment 1 completed a colour-based go/no-go task, thereby inducing an ACS for the go colour. Go/no-go targets were presented to the left or right of fixation, and shortly before each target a schematic drawing of an emotionally neutral face appeared at fixation gazing to the left or right. As is typically found, although gaze direction was non-predictive of the peripheral target’s location, participants responded more quickly to go targets appearing in the gazed-at location than the non-gazed-at location, suggesting these gaze cues captured and directed attention. Capture only occurred for faces drawn in the go colour, however, establishing that attentional capture by neutral gaze cues is contingent on ACSs. To examine the interaction between emotional valence and ACSs, Experiment 2 used the same design, but introduced an emotional manipulation: the expression on the face was either happy or sad. While sad faces replicated the results from Experiment 1, happy faces elicited a cueing effect regardless of the colour they were drawn in, indicating that happy faces are capable of capturing attention even when the face colour is inconsistent with the ACS. In addition to generalizing the conclusions of Gupta et al., the present results suggest that stimuli portraying positive emotions may be of sufficient behavioural relevance that “positivity” is a feature included in all ACSs by default.
Acknowledgement: This work was supported in part by NSERC grants to Naseem Al-Aidroos and Lindsay Plater.