Abstract
Although the eye-tracking technique has been widely used to passively study emotion recognition, no studies have utilised this technique to actively manipulate eye-gaze strategies during the recognition facial emotions. The present study aims to fill this gap by employing a gaze-contingent paradigm. Observers were asked to determine the emotion displayed by centrally presented upright or inverted faces. Under the window condition, only a single fixated facial feature was available at a time, only allowing for foveal processing. Under the mask condition, the fixated facial feature was masked while the rest of the face remained visible, thereby disrupting foveal processing but allowing for extrafoveal processing. These conditions were compared with a full-view condition. The results revealed that while both foveal and extrafoveal information typically contribute to emotion identification, at a standard conversation distance, the latter alone generally suffices for efficient emotion identification.