Poster #170 - Children’s Automatic Categorization of Language Group and Preference Information

• In Event: 3-003 - Poster Session 09
  In Poster Session: PS 09 Section - Race, Ethnicity, Culture, Context

Integrative Statement

Past research has demonstrated that children have a tendency to automatically encode category membership even when it is not relevant to an interaction (e.g., Weisman et al., 2015). However, no work has been done to determine if children’s automatic encoding is restricted to certain dimensions. One possibility is that children’s automatic encoding is dependent on the kind of information they receive. To test this question, we adapted the ‘Who Said What’ paradigm (Taylor et al., 1978), a task which measures participants’ automatic categorization of others, for use with children.
We evaluated children’s automatic categorization in two scenarios which we believe conveyed different kinds of information. In the first, we tested children’s consideration of spoken language, a psychologically salient social category for young children (e.g., Kinzler et al., 2007); in the second, we tested children’s consideration of others’ stated preferences, typically thought to be an individuating feature (e.g., Neuberg & Fiske, 1987). We hypothesized that children would automatically categorize people according to their language, but not their preferences.
During the familiarization portion of each block in the task, participants (N=72, 4–6 years, 36 females) saw photographs of four unfamiliar children of the same gender. Each child was depicted with a different image (a book with a picture of an animal on it) and a different voice. The content of the voice recordings was dependent on random assignment to one of two between-subjects conditions: language or preference. In the language condition, participants heard two voice recordings in English and two voice recordings in Spanish; each recording indicated that the child liked books. In the preference condition, participants heard all-English recordings, two of which expressed liking of books and two of which expressed disliking of books. During the test portion of each block, participants were tasked with remembering which child went with each book.
Participants’ performance was assessed by calculating a mean-error difference score (e.g., see Kurzban et al., 2001). This score is calculated from participants’ errors in the task, distinguishing between errors that occur within a category (e.g., misattributing one Spanish speaker’s book to another) from errors that occur across categories (e.g., misattributing a Spanish speaker’s book to an English speaker), and correcting for the fact that some errors are more likely due to chance alone. Positive scores indicate automatic encoding, while scores of zero indicate equivalent errors between and across categories. Preliminary analyses reveal that participants automatically encoded both preference (M=2.67, SD=2.37, t(35)=6.75, p<.001) and language (M=2.64, SD=3.10, t(32)=5.11 , p<.001).
The present findings indicate that automatic categorization is applied broadly during childhood, being used in the encoding of both spoken language and preference information. The latter result is surprising given that preferences are typically considered an individuating feature. Plans for future work include exploring whether context manipulations (e.g., prospect of future interactions), as opposed to manipulating information kind, influences automatic categorization. Further, we plan to test whether prior familiarity with targets interacts with automatic categorization.

Authors

• Nathan M Vasquez, University of Wisconsin-Madison
  Presenting Author

• Chuck Kalish, Society for Research in Child Development
  Non-Presenting Author

• Kristin Shutts, University of Wisconsin-Madison
  Non-Presenting Author