The Cognitive Costs of the Counter-Stereotypic: Gender, Emotion, and Social Presence

Abstract

We explored the concurrent and subsequent cognitive consequences of the experience of gender counter-stereotypic emotions. Participants experiencing gender counter-stereotypic emotions were expected to display less emotional expression and demonstrate poorer cognitive performance when in the public condition than when in the private condition. Seventy-one women and 66 men completed an anger- or sadness-inducing task privately or publicly. Participants completed two cognitive tasks: one during and one after the emotion-induction task. Participants exhibited poorer performance during and following gender counter-stereotypic emotions only in the public condition. Direct evidence for greater suppression of gender counter-stereotypic emotions in the public conditions was not obtained. These results suggest that the same public emotional events may be differentially cognitively depleting depending on one’s gender, potentially contributing to the perpetuation of stereotypes.

Keywords:

• cognitive depletion
• emotions
• gender
• social inhibition

Notes

1. 1. Notably, although the 35 non-compliant participants do not provide a good test of the current hypotheses because they did not experience the manipulated emotions, the results of the analyses remain largely the same when these participants are retained. The main effect of self-reported emotion type on the emotion manipulation check becomes marginal in the full sample. The interaction between self-reported emotion and emotion condition on the emotion manipulation check remains significant. However, the anger condition simple effect on the manipulation check becomes marginal as opposed to significant, further supporting the assertion that the non-compliant participants do not provide a good test of the current hypotheses. The interaction between gender and emotion condition on facial expressivity also becomes marginal in the full sample. The DRM performance effects (both omnibus and simple effect) remain significant in the full sample. Both the omnibus gender main effect and the predicted three-way interaction effects on anagram performance effects remain significant in the full sample, although the simple effects for the three-way interaction become trending. The unpredicted marginal gender by emotion condition interaction on anagram performance becomes non-significant in the full sample. No additional effects become marginal or significant when conducting these analyses on the full sample.

2. 2. Our measure of facial expressivity is further substantiated by supplemental analyses comparing the correlation between facial expressivity ratings and participants’ self-reported emotional experiences separately by gender, emotion condition, and social presence condition. In general, the correlation between facial expressivity ratings and participants’ self-reported emotional experiences in the counter-stereotypic emotion conditions were, as would be expected, stronger in the no social presence condition than in the social presence condition. For men, facial expressivity ratings of sadness were significantly correlated with self-reported sadness in the no social presence condition, r = .65, p < .01, but not in the social presence condition, r = .140, p = .579. For women, facial expressivity ratings of anger were not significantly correlated with self-reported anger in the no social presence condition, r = .130, p = .643, or in the social presence condition, r = –.125, p = .644, although these correlations differ in the predicted direction.

Additional information

Notes on contributors

Megan K. McCarty

Megan K. McCarty is affiliated with the Department of Psychological Sciences, Purdue University.

Janice R. Kelly

Janice R. Kelly is affiliated with the Department of Psychological Sciences, Purdue University.

Kipling D. Williams

Kipling D. Williams is affiliated with the Department of Psychological Sciences, Purdue University.