The Role of Face Masks in the Recognition of Emotions by Preschool Children

Abstract

Since the beginning of the COVID-19 pandemic, health policy requires staff working in preschool education to wear face masks. This has prompted worries about the ability of young children to recognize emotions and the possible impact on their development. Without face masks, preschoolers aged 36 to 72 months had a rate of correctly identified emotions on pictures of 11.8% to 13.1%.¹ Recent studies using photographs with digitally added face masks showed that participants had worse emotional recognition of the images with face masks; the first of these² tested preschoolers on a smartphone at home, and the second³ tested children aged 7 to 13 years. We therefore aimed to study the role of actual face masks on the recognition of joy, anger, and sadness in younger preschool children.

Full text

1
Title: The role of facemasks in the recognition of emotions by preschool children
Authors:
Juliane Schneider, MD, Developmental Unit, Neonatology, Department Woman-mother-child,
University hospital Lausanne, Switzerland
Vania Sandoz, MSc, Institute of Higher Education and Research in Healthcare, University of
Lausanne, Switzerland.
Lucile Equey, MD, Developmental Unit, Neonatology, Department Woman-mother-child, University
hospital Lausanne, Switzerland
Joanne Williams-Smith, MD, Developmental Unit, Neonatology, Department Woman-mother-child,
University hospital Lausanne, Switzerland
Antje Horsch, PhD, Prof., Institute of Higher Education and Research in Healthcare, University of
Lausanne, Switzerland.
Introduction: Since the beginning of the COVID-19 pandemic, health policy requires staff working in
preschool education to wear facemasks. This has prompted worries about the ability of young children
to recognize emotions, and the possible impact on their development. Without facemasks, pre-
schoolers aged 36- 72 months had correct responses rate between 11.8 % and 13.1%[1]. Recent studies
using photos with digitally added facemasks showed worse recognition with facemasks, the first tested
pre-schoolers on a smartphone at home [2] and the second tested 7- 13- year old children [3] . We
therefore aimed to study the role of actual facemasks on the recognition of joy, anger and sadness in
younger preschool children.
Methods: The primary outcome of this cross-sectional experimental study was the rate of correct
responses using pictures of adults displaying joy, anger or sadness. With 15 actors with and without a
surgical facemask, we created a dataset of 90 pictures displaying joy, anger or sadness (10 women, 5
men, based on demographic information of childminders in local public day-care centres) (Figure 1).
We built the experiment with E-Prime® [4]. The ethics committee for human research of the Canton
Vaud approved the study (study number: 2020-02687) and accepted that with the pandemic situation
consent could be waived. Parents of children attending public day-care centres received written, oral
and filmed information, with the possibility to opt out. Children aged 36 to 72 months without treated
neurodevelopmental impairment were eligible to participate. They sat in front of a computer, with a
known caretaker if they wanted, and a trained paediatrician showed randomly the 90 pictures. Children

2
could either name the emotion, point on a card showing emoticons of these three emotions, or choose
the response options “I don’t know” or “quit the experiment”. The statistical analysis included a
comparison of the correct response rate in the different conditions with χ2 tests and bias corrected
Cramer’s V to calculate effect sizes.
Results: Data was collected in nine public day-care centres. The sample consisted of 276 children
(girls: 48.9%, mean age=52.4 months, SD=9.6). The test lasted a median of 6.74 min per child (IQR
4.22-9.26). The rate of “I don’t know” responses was 3.1 % and 2.2 % children stopped the
experiment prematurely but their responses were included. The global correct response rate was
68.8%, 70.6% without facemask vs 66.9% with facemask (χ2(1) = 37. 783, p < .001, V = 0.0385,
95%CI [0.0266, 0.0515]), with a difference for joy (94.8 vs 87.3%, χ2(1) = 140.260, p < .001, V =
0.1301, 95%CI [0.1090, 0.1521] ), sadness (54.1 vs 48.9, χ2(1) = 21.937, p < .001, V = 0.0505, 95%CI
[0.0266, 0.0515]), but not anger (62.2 vs 64.6%, χ2(1) = 2.7094, p = .0997, V = 0.0147, 95%CI
[0.0000, 0.0399]). There was no difference between boys and girls. The rate of correct responses
increased with age (χ2(2) = 136.680, p < .001, V = 0.07363, 95%CI [0.0615, 0.0864]) (Figure 2A).
Finally, the analysis of the mistakes showed that up to 25 % pre-schoolers confused anger and sadness
and up to 21% answered joy for anger or sadness (Figure 2B).
Discussion: Actual facemasks, depicted on static pictures, were significantly associated with emotion
recognition of healthy preschool children, although differences were small and effects sizes were weak
(Cramer’s V ≤ 0.2). Joy was more recognised and mistakenly chosen for anger or sadness, probably
due to a positivity bias in children[6]. Overall, participants of our study, who had been exposed to
facemasks for nearly a year, recognized emotions on pictures better than reported in previous research,
even with facemasks [1, 6]. This study has several limitations including the generalizability of its
findings using static pictures instead of live actors, and the validity of the outcomes. Investigating the
role of facemasks in relation to other aspects of development and for children with developmental
issues remains important, particularly in the wake of a fourth wave of the COVID-19 pandemic.

3
Acknowledgements: No funding. Dr Bickle Graz had full access to all the data in the study and takes
responsibility for the integrity of the data and the accuracy of the data analysis.
References
1. Covic, A., N. von Steinbüchel, and C. Kiese-Himmel, Emotion Recognition in Kindergarten
Children. Folia Phoniatrica et Logopaedica, 2020. 72(4): p. 273-281.
2. Gori, M., L. Schiatti, and M.B. Amadeo, Masking Emotions: Face Masks Impair How We
Read Emotions. Front Psychol, 2021. 12: p. 669432.
3. Ruba, A.L. and S.D. Pollak, Children’s emotion inferences from masked faces: Implications
for social interactions during COVID-19. PLOS ONE, 2020. 15(12): p. e0243708.
4. Psychology Software Tools, I., E-prime 3.0. 2016.
5. Garcia, S.E. and E.C. Tully, Children's recognition of happy, sad, and angry facial
expressions across emotive intensities. J Exp Child Psychol, 2020. 197: p. 104881.
6. Nguyen, T.T. and N.L. Nelson, Winners and losers: Recognition of spontaneous emotional
expressions increases across childhood. Journal of Experimental Child Psychology, 2021.
209: p. 105184.
Figure 1. Examples of pictures of the same actor showing (A) joy, (B), sadness, and (C) anger without
and with facemasks.
A B C
Figure 2B Responses for joy, anger and sadness, without and with facemasks (n= 276)

4
Figure 2A Rate of emotion recognition by age, emotion, and presence of facemask (n = 276)
Joy w/o
mask
Joy with
mask
Anger w/o
mask
Anger with
mask
Sadness w/o
mask
Sadness with
mask
Joy 94.8 87.3 13.9 14.8 14.8 20.8
Anger 1.3 4.9 62.9 64.6 24.6 23.6
Sadness 1.3 4.2 16.0 15 54.1 48.9
Does not know 0.4 1.4 5.0 3.2 4.3 4.5
Quit 2.2 2.2 2.2 2.3 2.2 2.2
0
10
20
30
40
50
60
70
80
90
100
Percentage of responses
Emotion and condition
0
10
20
30
40
50
60
70
80
90
100
3 years old 4 years old 5 years old
Percentage of correct responses
age
Joy without mask
Joy with mask
All emotions
without mask
All emotions with
mask
Anger without
mask
Anger with mask
Sadness without
mask
Sadness with mask
***
**
**
** p<0.01
*** p<0.001