Skip to main content
. 2022 Jul 16;7:63. doi: 10.1186/s41235-022-00403-8

Table 3.

Emotion-specific effects of face masks on emotion recognition accuracy in the young and old cohort

Accuracy
Anger Fear Sadness Contempt
Δ hu Cohen's d BF10 % Δ hu Cohen's d BF10 % Δ hu Cohen's d BF10 % Δ hu Cohen's d BF10 %
Young cohort
w/o mask 0.25 1.7 > 100 0.29 2.0 > 100 0.30 2.4 > 100 0.25 1.9 > 100
With mask 0.17 1.7 > 100 0.22 1.7 > 100 0.17 1.7 > 100 0.13 1.3 > 100
Mask effect 0.08 0.45 22 31% 0.07 0.43 14 25% 0.14 0.86 > 100 45% 0.12 0.85 > 100 48%
Old cohort
w/o mask 0.11 1.1 > 100 0.12 1.0 > 100 0.17 1.3 > 100 0.15 1.4 > 100
With mask 0.10 0.96 > 100 0.10 0.96 > 100 0.10 1.0 > 100 0.07 0.82 > 100
Mask effect 0.01 0.07 0.25 8% 0.02 0.17 0.55 20% 0.08 0.51 53 44% 0.07 0.57 > 100 50%
Age effect
Mean 0.11 1.2 > 100 0.15 1.5 > 100 0.10 1.2 > 100 0.08 0.92 > 100
Mask effect 0.07 − 0.46 0.07 0.05 − 0.31 0.09 0.06 − 0.39 0.08 0.05 − 0.34 0.09

All comparisons are one-sided and signs indicate the direction of an effect relative to the prediction (see text). BF10 > 3 (at least moderate evidence for H1) are bold and BF10 < .33 (at least moderate evidence against H1) are bold and in italics. Δ hu, unbiased hit rate minus chance; %, relative decline due to face masks (relative to faces without masks)