Poster #12 - The automatic encoding of gender in transgender and cisgender children

• In Event: 3-003 - Poster Session 09
  In Poster Session: PS 09 Section - Attention, Learning, Memory

Integrative Statement

Due to restrictions on attention and memory, children cannot automatically encode (remember) all information when encountering a new person, yet they must encode some information in order to recognize people later. For example, when a child brand new to a school meets other children on their first day, how can they remember who they met to talk again? Research has suggested that gender may be one dimension that is encoded by cisgender children when they meet someone new (Weisman, Johnson, & Shutts, 2015; Shutts, Kenward, Falk, Ivegran, & Fawcett, 2017). In the present work we investigate if gender is also automatically encoded by transgender children. Transgender children are an important test case of the inevitability of gender encoding because they have been shown to stereotype and essentialize gender less (e.g., Olson & Enright, 2018) suggesting gender may be less central to their categorization. The current study investigated whether transgender children encode gender and if they do, whether the degree to which they encode gender differs from cisgender children. To test this question, we administered a face memory task to transgender participants (N=42), the siblings of transgender children (N=31), and unrelated cisgender controls (N=41) in which they heard about four children who went to the zoo and then were asked to remember the animals that each child saw (Weisman et al., 2015). Results suggested that the three groups did not significantly differ in the degree to which they encoded gender, F(1,109) = 2.74, p = .101, partial eta squared = .02. Critically, transgender participants encoded gender, t(40) = 3.79, p < .001, d = 0.59, as did their siblings, t(30) = 2.34, p = .026, d = 0.42. Unexpectedly, our control group did not show a significant gender encoding effect, t(38) = 0.41, p = .682, d = 0.07, despite several previous reports of significant gender encoding on this measure in cisgender children of the same age (Weisman et al., 2015; Shutts et al., 2017). To assess whether this unexpected lack of effect was unique to the location of testing, recruitment strategies employed for the control group, or if this was a spurious observation, we collected a second, larger sample of cisgender children (N=100) from the same subject pool. This group showed a strong, significant gender autoencoding effect, d = 0.37, similar in magnitude to the transgender children and the siblings in the original study. Therefore, across the four samples in this set of studies and the three past tests using this measure, the auto-encoding gender effect was observed 6 of 7 times it was tested, including in transgender children, their siblings, children attending a gender-neutral preschool, and in the U.S. and Sweden (Weisman et al., 2015; Shutts et al., 2017). The results are consistent with theories that suggest that gender is a fundamental social category that is automatically encoded.

Authors

• Jessica Glazier, University of Washington
  Presenting Author

• Selin Gülgöz, University of Washington
  Non-Presenting Author

• Kristina Olson, University of Washington
  Non-Presenting Author