Emotion as Information: Children Infer Competence From Others’ Emotional Reactions to Performance

• In Event: Learning About Others From Others: Implications for Stereotype Formation and Intervention

Abstract

Children are motivated to learn not just about the physical world, but also about the social world: the self, other individuals, and social groups. Indeed, children receive much information about the self and others in their social interactions and communications with others. While much work has focused on how (biased) beliefs about social groups may be transmitted through language, less is known about how such beliefs may also be learned through non-verbal information, such as observations of others’ emotional expressions. Building on recent work suggesting that children have an intuitive understanding of how others’ emotional expressions relate to their mental states, here we ask whether children can draw inferences about competence from others’ emotional reactions to their performance outcomes.

Across three experiments, we ask whether expressions of surprise inform children’s evaluations of others’ relative competence. Participants saw trials where two students both succeeded (success trials) or failed (fail trials) at different activities (sports, academics); a teacher showed surprise to one student (surprise student) and a neutral expression to the other (no-surprise student). Participants were asked which student was better at the activity. Our key predictions were that when both students succeeded, participants would select the no-surprise student as better, but when both students failed, participants would choose the surprise student.

In Experiment 1 (Adults; N=67, 8 trials/subject), adults confirmed our predictions: 92.2% chose the no-surprise student in success trials (Z = 7.45, p < .001, Exact Wilcoxon-Pratt Signed-Rank Test), and 93.0% chose the surprise student in fail trials (Z = 7.51, p < .001). In Experiment 2 (N=28, 4-9 year-olds, 8 trials/subject), we found that older children, but not younger children showed the predicted pattern of responses. For the older half of children, 98.2% chose no-surprise student in success trials (Z = 3.64, p < .001), and 76.8% chose surprise student in fail trials (Z = 2.16, p = .039); younger children were at chance in both trial types. In Experiment 3, we run a pre-registered study with a larger sample of older children (Current N=78, 6-8 year-olds; pre-registered N=90, 4 trials/subject). So far, we have replicated the older children’s results from Experiment 2: 68.1% have chosen the no-surprise student in success trials (Z = 3.394, p < .001) and 85.6% have chosen the surprise student in fail trials (Z = 6.671, p < .001).

These results provide initial evidence that school-aged children can use others’ emotional expressions, specifically surprise, to evaluate an individual’s competence. Next steps involve investigating whether children may also use such expressions to develop beliefs about a social group’s competencies and generalize those to a novel member of that group. Though parents and teachers may be learning to communicate in unbiased ways, their non-verbal communication (even non-valenced expressions of surprise) may nonetheless reveal their biases towards individuals and social groups. Our work is one of the first steps towards characterizing the cognitive capacities underlying how biases may be transmitted from emotional expressions and non-verbal communications broadly.

Authors

• Mika Asaba, Yale University
  Presenting Author

• Yang Wu, Stanford University
  Non-Presenting Author

• Brandon Carrillo, Yale University
  Non-Presenting Author

• Hyowon Gweon, Stanford University
  Non-Presenting Author