Answers are grounded in this paper and cite their sources.
Social evaluation increasingly unfolds in digital environments where visual cues can be computationally altered, yet it remains unclear whether facial self-resemblance increases positive evaluations of others in naturalistic social interactions. Here, we conducted a preregistered online experiment using a yoked-stimulus design—in which morphed faces were shown to both self-relevant and matched control viewers—to isolate the effects of facial self-resemblance from morphing artifacts. Sixty participants viewed eight two-person conversational videos in which speakers’ faces were morphed to resemble either the viewer or a different participant, yielding 480 evaluations of morphed speakers. We found that facial self-resemblance did not produce a consistent increase in positive evaluations across participants. Exploratory analyses further suggested substantial heterogeneity in responses: participants with higher self-esteem evaluated self-resembling speakers more positively, whereas participants with lower self-esteem showed the opposite tendency. These patterns were observed across both binary and continuous measures of evaluation and were absent in yoked control analyses designed to remove self-relevance. Together, our findings suggest that self-referential visual cues do not exert a fixed social effect. Instead, their consequences may depend on how individuals view themselves, suggesting that digitally altered appearance interacts with self-concept in shaping social evaluation.
No takes yet. Share an insight, caveat, or question.
Kasahara et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: