Modality-specific attentional biases to facial vs. vocal emotions

Most research on attentional biases towards emotional information has been conducted in the visual modality. There are theoretical debates about whether attentional processes generalize across nonverbal modalities (e.g., facial vs. vocal), but no research has directly compared how emotional faces and voices capture attentional resources. The current study compared response time-based measures of attentional bias to emotional expressions in facial and vocal dot-probe tasks. Participants (18-63 years old; 52% men, 45% women, 3% non-binary) located cues that replaced one of two simultaneously presented stimuli (pairings of angry-neutral, happy-neutral, or angry-happy faces/voices). Participants showed greater difficulty disengaging from non-targets (i.e., neutral stimuli, or happy stimuli when paired with angry) in the vocal than the facial modality. Attention shifted from greater avoidance of vocal emotional targets to greater vigilance towards them across trials, but there was no change in attentional biases towards facial stimuli across the task. Differential attentional biases to emotional versus neutral voices may thus be larger in magnitude and more dynamic than with faces. Taken together, these findings provide emerging evidence of distinct patterns of attentional allocation between nonverbal modalities in the dot-probe. Future work should examine whether stimulus or task features contribute to modality-specific attentional biases to emotion.