Poster #50 - Can Children Infer Stereotypes About a Group’s Competence by Observing how Group Members are Addressed?

• In Event: Poster Session 10

Abstract

Children endorse stereotypes about competence at an early age. Such stereotypes are harmful as they influence how people from different social groups are treated, shaping future life outcomes. Here, we investigated how competence stereotypes develop, focusing on the role of linguistic interactions. Specifically, we asked whether children and adults can infer a social group’s relative competence by observing whether group members are addressed with more complex versus simple language.
Participants observed a protagonist address members of two novel social groups (Figure 1). Across six trials, the protagonist consistently used low complexity (LC) language with one group, and high complexity (HC) language with the other. Complexity was manipulated via age of acquisition of words, speech rate, and pitch/intonation. After each trial, participants judged which group member is “smarter”; at the end of the study, participants rated how smart each group is, how smart a novel member from each group is (both on a 1-4 scale), and why.
Results suggest that adults (n = 32, Mage = 26.7) and 8-9-year-olds (n = 32, Mage = 9.0), but not 5-6-year-olds (n = 32, Mage = 5.7) inferred that the group who was spoken to with HC language is smarter. Across six trials, 8-9-year-olds (M = 0.61, p = .009) and adults (M = 0.86, p < .001) selected the HC character significantly above chance, but 5-6-year-olds’ choices did not differ from chance, (M = 0.52, p = .90; Figure 2). When asked how smart each group is, 8-9-year-olds rated the HC group (M = 3.53) significantly higher than the LC group (M = 2.78, p < .001), as did adults (HC: M = 3.5. LC: M = 2.78; p < .001); However, 5-6-year-olds rated the HC group (M = 2.94) and LC group (M = 3.06) similarly, p = .71. Finally, when asked how smart novel group members are, adults rated the HC character (M = 3.53) significantly higher than the LC character (M = 2.69; p < .001), but 5-6- and 8-9-year-olds rated the HC character (M = 2.91) and the LC character (M = 3.17) similarly, p = .10. Analyses of open-ended justifications suggest that participants who mentioned the way that the protagonist spoke to each group were significantly more likely to select the HC character as being smarter (DV1), rated the HC group as significantly smarter (DV2), and these adults (but not children) rated the HC novel character as significantly smarter (DV3).
Together, these data suggest that by at least third grade, children are sensitive to the relative complexity with which speakers address members of different social groups, and infer the relative competence of these groups on this basis. This raises the possibility that this is one mechanism underlying the development and perpetuation of stereotypes. Younger children may not have made such inferences due to several cognitive and meta-linguistic factors (e.g., not recognizing that the different groups received language of different complexity). We discuss future directions, including ongoing work exploring other dimensions of linguistic interactions that may convey stereotypes.

Authors

• Roya Baharloo, University of California - Berkeley
  Presenting Author

• Aneesa Conine-Nakano

• Mahesh Srinivasan, University of California - Berkeley