Abstract
Purpose: Well-learned processes are considered to be automatic. The automaticity of emotional face processing was investigated using a Stroop analog photo/word task. Methods: Photos of male and female faces displaying different expressions (happy, sad, angry) were paired with words (happy, sad, angry) presented simultaneously for 147ms above and below a central fixation cross. Reaction time and accuracy were recorded. In the Face task, the emotional face served as target and the word was ignored. These conditions were reversed in the Word task. Each task comprised 144 trials with 72 congruent trials (expression and word match) and 72 incongruent trials (mismatch). Results: A ceiling effect was seen for the Word task. Responses were faster and more accurate in the Word than the Face task. For the Face task, happy faces had the quickest and most accurate responses. Sad faces had the slowest and least accurate responses when the poser was male, whereas female angry faces were responded to slowest and least accurately. A Stroop effect occurred, where mismatched emotional words interfered with the speed and accuracy of recognizing expressions, but facial expression had no significant effect on word responses. Responses on congruent angry trials were significantly faster with a male poser, and they were more accurate with a female poser. A gender specific effect was also evident in that male subjects' responses were faster on congruent sad trials, and slower on incongruent sad face/ happy word trials when the poser was female as opposed to male. Conclusions: 1) Poser gender influenced responses to sad and angry expressions differently. 2) A Stroop effect occurred in the Face task. 3) The Stroop effect varied as a function of the gender of the subject and poser. These results demonstrate that the gender of both the subject and poser may have an effect on the recognition of emotional expressions. In contrast, emotional word processing is unaffected by the faces.