Emotional facial expressions differentially influence predictions and performance for face recognition

Abstract

This study examined how participants' predictions of future memory performance are influenced by emotional facial expressions. Participants made judgements of learning (JOLs) predicting the likelihood that they would correctly identify a face displaying a happy, angry, or neutral emotional expression in a future two-alternative forced-choice recognition test of identity (i.e., recognition that a person's face was seen before). JOLs were higher for studied faces with happy and angry emotional expressions than for neutral faces. However, neutral test faces with studied neutral expressions had significantly higher identity recognition rates than neutral test faces studied with happy or angry expressions. Thus, these data are the first to demonstrate that people believe happy and angry emotional expressions will lead to better identity recognition in the future relative to neutral expressions. This occurred despite the fact that neutral expressions elicited better identity recognition than happy and angry expressions. These findings contribute to the growing literature examining the interaction of cognition and emotion.

Keywords:

• Cognition
• Emotion
• Memory
• Metamemory
• Emotional facial expressions
• Face recognition

Acknowledgments

This work was authored as part of the Contributor's official duties as an employee of the United States Government and is therefore a work of the United States Government. In accordance with 17 U.S.C. 105 no copyright protection is available for such works under U.S. law.

Notes

¹The scale was 50–100 because two-alternative forced-choice testing entails that 50% is equivalent to chance performance.

²All participants completed two short study/test practice blocks prior to completing the main portion of the experiment. Each study/test practice block consisted of two studied faces facing straight ahead with a happy and neutral emotional expression and two test face pairs at 45 degree angles with neutral expressions utilising the same method outlined above.

³A second experiment using an implicit encoding condition of gender identification (i.e., Is this person male/female?) was conducted on an additional 114 CSU students (M _age=18.77, SD=1.84; 75% female, 91% right-handed). Whether the encoding task was explicit or implicit did not impact the overall patterns found, as no interactions were found. A 2 Condition (implicit, explicit)×3 Emotion (happy, angry, neutral) mixed-model ANOVA produced no significant interactions for judgements of learning (JOLs); F(2, 225) = 1.56, p=.21, identification rates, F(2, 225) = 0.12, p=.89, or retrospective confidence ratings, F(2, 225) = 1.01, p=.36. This demonstrated that the influence of happy, angry and neutral emotional expressions on JOLs, identity recognition, and retrospective confidence judgements persists even when attention is not explicitly directed at the studied expression within the context of this experiment.