Visual search of emotional faces: The role of affective content and featural distinctiveness

Abstract

We investigated the source of the visual search advantage of some emotional facial expressions. An emotional face target (happy, surprised, disgusted, fearful, angry, or sad) was presented in an array of neutral faces. A faster detection was found for happy targets, with angry and, especially, sad targets being detected more poorly. Physical image properties (e.g., luminance) were ruled out as a potential source of these differences in visual search. In contrast, the search advantage is partly due to the facilitated processing of affective content, as shown by an emotion identification task. Happy expressions were identified faster than the other expressions and were less likely to be confounded with neutral faces, whereas misjudgements occurred more often for angry and sad expressions. Nevertheless, the distinctiveness of some local features (e.g., teeth) that are consistently associated with emotional expressions plays the strongest role in the search advantage pattern. When the contribution of these features to visual search was factored out statistically, the advantage disappeared.

Acknowledgements

This research was supported by Grant SEJ2004–420/PSIC, from the Spanish Ministry of Education and Science.

We thank Margaret Dowens and Gernot Horstmann for their helpful comments on a previous draft of this article.

Notes

¹Kohler et al. (2004) assessed action units of four emotional face expressions (happy, fearful, angry, and sad), with 16 models for each emotion, and three certified FACS raters, using a similar procedure to ours. Although there was consistency between the Kohler and the current study regarding the association between some specific AUs and emotional expressions, there were also some discrepancies. These could be partly due to the fact that totally different stimuli were used in the Kohler et al. (Gur et al., 2002, stimuli) and the current study (KDEF database), and that Kohler et al. used only faces that showed extreme intensity emotions and each actor was assigned to only one emotion, whereas our images depicted a wider range of intensities and all actors showed all emotions. The discrepancies also call for some caution regarding the reliability of the assessment of facial features.