Purchase this article with an account.
David De Vito, Cody Cushing, Hee Yeom Im, Reginald Adams, Jr., Kestutis Kveraga; Spatial and feature-based attention to emotional faces. Journal of Vision 2017;17(10):1290. doi: https://doi.org/10.1167/17.10.1290.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Anticipation enhances our ability to adaptively respond to positive and negative stimuli. However, research is quite mixed about how different types of expectations affect response efficiency. Specifically, we examined how cueing stimulus location (spatial attention) and/or emotion (feature-based attention) affects response efficiency on a trial-by-trial basis. Our task engaged participants (N=44) in speeded identification of emotion expressions (angry versus happy) that were presented peripherally (left versus right) for 250 ms. Prior to face onset, colored arrow cues, which were 95% predictive, were displayed for 1 s to inform subjects of either stimulus location or emotion, both, or neither (Uncued). Overall, responses to happy faces and expressions presented in the right visual field yielded shorter reaction times (RTs), but the RTs, accuracy, and response patterns were strongly modulated by cueing condition. Uncued trials elicited the longest RTs (average: 615 ms). Cueing location produced faster (average: 578 ms) and more accurate responses than uninformative cues (p< .001), while cueing emotion resulted in even shorter RTs (average: 546 ms) than cueing location (p< .001). Cueing both location and emotion (p< .001) evoked fastest responses (average: 481 ms), in a superadditive function reflecting the combined effects conferred by cueing emotion and location (p < .001) separately. Moreover, the face identity cues (gender and race) interacted with cueing. On Uncued trials or when only the location was cued, we found emotion x gender (faster for happy female and angry male), emotion x race (faster for angry black and happy white faces), as well as emotion x gender x race interactions, which were abolished by emotion cueing. In conclusion, we found that being able to anticipate facial emotion and location substantially speeds up recognition, while interactions with facial identity cues (race and gender) are abolished by emotion cueing.
Meeting abstract presented at VSS 2017
This PDF is available to Subscribers Only