Purchase this article with an account.
Timothy D. Sweeny, Marcia Grabowecky, Ken A. Paller, Satoru Suzuki; Discriminating fleeting facial expressions using featural and configural information. Journal of Vision 2009;9(8):498. doi: https://doi.org/10.1167/9.8.498.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Humans are adept at discriminating fleeting emotional expressions. We investigated how the type of expression and duration of presentation influenced discrimination accuracy. Observers viewed two sequentially presented facial expressions, one neutral and the other emotional (fearful, angry, or happy), in a two-interval forced-choice task with stimulus duration varied across trials (10, 20, 30, 40, or 50 ms). All faces were masked by a face with a surprised expression. On each trial, observers attempted to select the face with the emotional expression, and to name the expression. Discrimination against neutral was above chance at all durations, and more accurate for happy than for angry or fearful expressions. Emotional expressions that displayed teeth yielded the highest accuracy. To evaluate discrimination among emotional expressions during expression naming, we calculated d' using ‘hits’, ‘misses’, ‘false alarms’, and ‘correct rejections’ specific to each expression pair (e.g., responding ‘angry’ or ‘happy’ on angry and happy trials). Discrimination between angry and happy expressions was better than discrimination between fearful and happy expressions, and both pairs were discriminated above chance even when presented for only 20 ms. In contrast, discrimination between fearful and angry expressions was near chance at all durations. Discrimination between these expressions was also near chance in an experiment in which emotional faces were all negative in valence (fearful, angry, or disgusted), suggesting that discrimination between fleeting fearful and angry expressions is difficult, and not merely a deficit incurred from simultaneously discriminating expressions with negative and positive valence. With inverted faces, only discrimination between angry and happy expressions was impaired, which suggests a contribution of configural processing in this particular discrimination. Together, these results demonstrate that surprisingly brief presentations are sufficient for discriminating emotional expressions from neutral expressions. However, discriminating emotional expressions is difficult and depends on information from individual features and from their configurations.
This PDF is available to Subscribers Only