Do face masks make it harder for children to recognise faces?
Masks cut children's ability to recognise newly learned faces by about a fifth and made them process faces less as a whole, an effect somewhat larger than in adults on a difficulty-matched test.
Source
Face masks disrupt holistic processing and face perception in school-age children
Study at a glance
- Design
- Human experiment — Online experiment: mask vs no mask between subjects, upright vs inverted faces within subjects, on the CFMT-K; children compared with adults on the difficulty-matched adult CFMT and on the same CFMT-K
- N
- N=72 · 72 children aged 6-14 (37 masked, 35 unmasked); comparison groups of 495 adults on the adult CFMT and 72 adults on the CFMT-K
- Population
- School-age children in Canada recruited by snowball sampling, plus adult online participants recruited via Prolific
- Outcome
- Percentage correct on the Cambridge Face Memory Test (Kids and adult versions) for upright and inverted faces; face inversion effect as an index of holistic processing
Structured fields used in claim comparison tables when every cited study has a complete layer.
What they did
Children completed the Cambridge Face Memory Test for Kids at home online, learning children's faces and then picking them out from new viewpoints, lighting and added noise. They were randomly assigned to see all faces with a mask or without, and each child did the test once with upright and once with upside-down faces, since a large drop for inverted faces signals holistic (whole-face) processing. Their results were compared with 495 adults who took the difficulty-matched adult version, and with 72 adults who took the same children's version.
What they found
Masks lowered children's accuracy for upright faces by about 20%. Turning faces upside down hurt performance by 23% for unmasked faces but only 15% for masked faces, suggesting masked faces are processed less holistically; masked inverted faces were still well above chance, so this was not a floor effect. The mask cost was about the same for younger and older children, even though older children performed better overall and showed a bigger inversion effect. Compared with adults on the difficulty-matched test, children's mask cost was larger (20.1% vs 13.6%), but when adults took the children's test they were so accurate that no difference in mask cost could be detected; girls outperformed boys.
The limits
What it doesn't show
Children did the test unsupervised at home, and parents were encouraged to help younger children read instructions, so testing conditions varied. The child-adult difference in mask cost was small and appeared only on the difficulty-matched comparison, not when adults took the same child test, where a ceiling effect clouds interpretation. The CFMT-K uses only male Caucasian faces, which limits claims about sex differences and other-race faces. Masks were added digitally to static photos, so the study says nothing directly about recognising real masked teachers and classmates or about social consequences.
Key terms
- Holistic face processing
- Perceiving a face as an integrated whole, including the spacing between features, rather than as separate parts.
- Face inversion effect
- The large drop in face recognition when faces are upside down, used as an index of how much holistic processing is normally used.
- Cambridge Face Memory Test (CFMT / CFMT-K)
- A standard test in which people learn several unfamiliar faces and then pick them out from three options under increasingly difficult conditions; the Kids version uses children's faces and is easier.
- Floor effect
- When performance is so close to the lowest possible (or chance) level that further drops cannot be detected.
- Ceiling effect
- When performance is so close to perfect that differences between conditions or groups are hidden.
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
What does a smaller face inversion effect for masked faces suggest?
Common questions
Why test upside-down faces at all?
Inversion disrupts holistic processing, so the size of the drop tells you how much people normally rely on it; a smaller drop for masked faces implies masked faces were not being processed holistically in the first place.
Are children really more affected by masks than adults?
Possibly, but modestly: the difference appeared when each group took a difficulty-matched version, not when adults took the same children's test, where adults were near ceiling.
Does the mask effect shrink as children get older?
No; although older children recognised faces better overall, the cost of masks was similar for younger and older children.
More on Face processing