Impact of Autism Label on Face Recognition

Background: Labeling an individual can influence the inferences others make about them. For example, using a face-inversion paradigm, Civile et al. (2018) found the difference between memory for upright vs. inverted faces was larger for faces labeled “regular people” than those labeled “diagnosed with autism.” Civile et al. suggested this could mean participants treated faces labeled as “diagnosed with autism” in a more objectified manner than faces labeled as “regular people.”

Objective: In two pre-registered studies, we attempted to replicate and extend Civile et al.’s (2018) findings. Study 1 was a close replication using the face inversion paradigm. Study 2 was a conceptual replication using an in-group vs. out-group recognition memory paradigm (Bernstein et al., 2006).

Method: In Study 1, we recruited a sample of 140 adults via MTurk. Participants saw 64 faces at encoding: Half were upright and half were inverted. As in Civile et al.’s study, half of participants were informed the faces they would see were of “regular people” and half were informed the faces were of “people diagnosed with autism.” At test, participants saw the 64 old faces (in the same orientation at encoding) and 64 new faces (half upright and half inverted), and made old/new judgments.

In Study 2, 147 undergraduates in one of three conditions saw 40 faces at encoding. In “identity-first” and “person-first” conditions, half of the faces appeared on blue backgrounds and were labeled either “autistic person” or “person with autism,” and half appeared on orange backgrounds and were labeled either “non-autistic person” or “person without autism.” In a control condition, faces appeared on blue and orange backgrounds, but were not given a label. At test, all participants saw the 40 old faces on the same color backgrounds as at encoding and 40 new faces (half on blue and half on orange). No faces were labeled at test.