Schedule of Events | Symposia

Face perception differentially contributions to individual differences in face recognition ability depending on how it is measured

Poster Session F - Tuesday, April 1, 2025, 8:00 – 10:00 am EDT, Back Bay Ballroom/Republic Ballroom

Tayan Zhang^1,2 (zhangtayan@gmail.com), Alison Campbell^1,2,5, Jeremy Wilmer⁴, Laura Germine³, Joseph DeGutis^2,3; ¹Boston University, ²Boston Attention and Learning Laboratory, ³Harvard Medical School, ⁴Wellesley College, ⁵Boston University Chobanian and Avedisian School of Medicine

Face recognition ability is a stable, domain-specific cognitive ability that is highly heritable. Researchers have identified several potential mechanisms that contribute to face recognition ability, including holistic processing and associative memory. However, the involvement of these potential mechanisms may rely on how face recognition ability is assessed. To test this, we examined the contribution of face perception (measured using the Cambridge Face Perception Test-CFPT) to two popular methods of assessing face recognition ability: 1) Cambridge Face Memory Test (CFMT), a test that requires learning and recognizing six unfamiliar faces, and 2) Famous Face Memory Test (FFMT), a test that requires providing identifying information for highly-familiar famous faces. To ensure the replicability of the results, we recruited two large web-based US samples (Sample 1: N=632; Sample 2: N=649) that each completed the CFPT, CFMT, and a different version of the FFMT. First, we found that the CFMT and FFMT were strongly correlated in both samples (r=.55/.55). Importantly, in both samples, the correlation between CFPT and CFMT (r=.56/.55) was significantly stronger than for the CFPT and FFMT (r=.35/.34). Thus, when accounting for perceptual ability, there remains a larger amount of unexplained variance for familiar face recognition. This suggests that, while learning unfamiliar faces in the CFMT depends largely on perceptual ability, recognizing famous faces in the FFMT appears to depend more heavily on extra-perceptual processes such as making and retrieving semantic associations. Mechanistic models of individual differences in face recognition should consider the specific methods used to assess this ability.

Topic Area: PERCEPTION & ACTION: Vision