Sensitivity to genuine versus posed emotion specified in facial displays

Abstract

Two experiments were performed to investigate whether social perceivers were sensitive to the veracity of sad and fear facial displays as well as happiness. In Experiment 1, participants were asked to consider in blocks whether targets were happy or not, sad or not, fearful or not. Triads of photographs (neutral, posed, genuine) were displayed and results showed participants were sensitive to whether each emotion was present and distinguished posed from genuine displays. This sensitivity was emotion specific. In Experiment 2, participants completed a priming task to eliminate instructions to judge target displays. Neutral, posed and genuine displays from a single target were used as primes in a word valance identification task. The results revealed faster responding to positive words following genuine than posed happiness and faster responding to negative words following genuine than posed fear. Together the two experiments demonstrated perceiver sensitivity to negative emotion in an explicit and implicit context.

Keywords:

• Emotion
• Facial expression
• Sensitivity
• Spontaneous

Notes

¹A sex categorisation task (Walton, 2004) with the same methodology was used to control for possible impairments in face perception that might influence emotion recognition[0]. The task employed a neutral display from eight targets. The mean accuracy rate was 99%, showing that participants were able to detect information relevant to sex identification from facial displays and consequently no participant was excluded based on difficulties perceiving this information.

²A measure of response bias (B′) was also calculated to confirm that participants adopted a more stringent response criterion in the feel than in the show condition. Response bias was compared to 0 using single-sample t-tests. A response bias was found in the show but not in the feel condition, therefore, participants did not demonstrate a proclivity to respond with one response over the other in the feel condition. The formula used to calculate sensitivity takes response bias into account and therefore it is not considered further.

³It is accepted that the meaningful range of A′ is from 0 to 1.00. Higher scores are indicative of higher sensitivity. A sensitivity score of 0.5 is indicative of chance-level responding.

⁴Although there was a main effect of handedness, F(1, 20)=5.4, p<.05, on sensitivity scores with left-handed individuals being more sensitive than right-handed individuals (Ms=0.82 vs. 0.75), there were no interaction effects between handedness and any of the key IVs (emotion; condition) and the same pattern of results was seen for both right- and left-handed individuals. Hence handedness was not considered further.