Is being good with faces one ability or several?
People who were good at matching unfamiliar faces also tended to be good at learning new faces and recognising famous ones as children, but this shared face ability explained only about a quarter of the differences between people.
Source
Individual differences in face identity processing
Study at a glance
- Design
- Cross-sectional — Individual-differences correlational design: Study 1 correlated three face-identity tests with a self-report ability questionnaire; Study 2 added two further face tasks and ten perceptual, executive and personality measures in an online sample
- N
- Study 1: 40 Aberdeen students recruited, 5 removed for software failure or a failed familiarity check, and 34 with questionnaire data; Study 2: 103 online panel members plus 4 students, aged 21-82
- Population
- UK university students (Study 1) and a broader online adult sample with mean age 53 (Study 2)
- Outcome
- Correlations among the Glasgow Face Matching Test, Cambridge Face Memory Test and an adapted Before They Were Famous task, and with self-ratings, face detection, Mooney faces, object and space perception, executive function, Big Five personality and empathy
Structured fields used in claim comparison tables when every cited study has a complete layer.
What they did
Across two studies, adults completed three face-identity tests: deciding whether two photos of strangers show the same person, learning and later recognising new faces, and recognising celebrities from photos taken before they were famous. In Study 1, 40 students also rated their own face-recognition abilities on six questions. In Study 2, a broader sample of 103 online panel members plus 4 students additionally completed face-detection and Mooney-face tasks plus measures of space and object perception, executive function, personality and empathy.
What they found
In both studies the three face-identity tests correlated positively with each other (face matching and face learning showed the strongest, moderate link), consistent with a general face ability. Self-ratings mostly failed to predict performance; only a question about recognising celebrities from childhood photos correlated with actual familiar-face recognition and face learning. No non-face task correlated with all three face-identity tests, but face matching correlated with detail-focused matching tasks, and object recognition from silhouettes correlated with face learning and familiar-face recognition. Personality and empathy showed no strong links.
The limits
What it doesn't show
Both samples were small for correlational research, which the authors acknowledge, so many individual correlations have wide confidence intervals and may not replicate. Many correlations were tested and the authors deliberately reported uncorrected p-values first, using a lenient false discovery rate as a check, so some isolated associations may be false positives. The familiar-face test hit a floor in Study 2 (20% recognised no celebrities as children), which weakened its correlations. Correlations cannot show that a single underlying 'face factor' actually exists, and it could account for at most about 25% of the variance.
Key terms
- General face factor (f)
- A proposed general ability underlying performance on many face tasks, by analogy with the general intelligence factor g.
- Unfamiliar face matching
- Deciding whether two simultaneously shown photos of strangers show the same person, measured here by the Glasgow Face Matching Test.
- Familiar face recognition
- Recognising people you already know; here tested by identifying celebrities from photos taken before they became famous.
- Individual-differences approach
- Studying how abilities vary across ordinary people and whether different abilities rise and fall together, rather than studying only extreme cases such as brain injury.
- Metacognitive insight
- How accurately people can judge their own abilities; here, whether self-ratings of face recognition matched test performance.
Flashcards
0 of 10 answers reviewed
Research intelligence for this paper
See its role on concept claims, tensions it is part of, placement history, and related discoveries.
Quiz yourself
Which test was used to measure familiar face recognition?
Common questions
If the face tests correlate, why not treat face recognition as one skill?
The shared variance was at most about 25%, so most differences between people on each test come from other, task-specific influences.
Are people good judges of their own face-recognition ability?
Mostly not for unfamiliar faces; only a question about recognising celebrities as children related to actual performance, suggesting some insight for familiar faces only.
Why might unfamiliar face matching correlate with non-face matching tasks?
Matching strangers' photos may rely on careful comparison of individual features, the same kind of detail-focused, controlled checking needed in object-matching and local-detail tasks.
More on Face processing