Poster #80 - 'Infants' emotion recognition during the covid-19 pandemic: do masks differentially hinder emotion recognition?

• In Event: Poster Session 19

Abstract

Infants' emotion recognition abilities support socio-emotional and regulatory skills development (Leppänen & Hietanen, 2001; Nowicki & Duke, 1992). The widespread use of facemasks might hinder 'infants' ability to recognize emotions from facial expressions. We tested this question using two conditions of a cross-modal emotion recognition task(based on Walker-Andrews, 1986).In both conditions, infants view two facial expressions side-by-side, accompanied by a vocal expression corresponding to one of the two emotions shown (see figure 1). In one of the conditions, a facemask partially occludes the facial expression.
We hypothesize that: (H1) looking time to the congruent expression (index of emotion recognition) will decrease in the facemask condition; (H2) In the facemask condition, emotion recognition of sadness will be affected more than happiness.
Data was collected online via Zoom from October 2020 to January 2021 with a sample of 127 infants (age: M=18.98 ±2.63 months, 49 % female). The infants were recorded during watching the task. For each video, the 'infants' direction of gaze (right/left/center/outside the screen) was coded continuously. Coders were blind to task conditions. Kappa interrater reliability is 0.78.
To test our hypothesis, we conducted a mixed effect regression model. Our dependent variable is the proportion of looking time to the congruent side (out of the looking time to both sides of the screen). The independent variables are condition (mask \ no mask) and emotion (happy\ sad), and age and sex as control. We tested if the proportion of looking time is significantly greater than 0.5.
We found a main effect of congruency; infants look longer to the congruent expression (t(103)= 5.95,p < .000). Additionally, gender and age did not predict looking time proportion.
We found a significant effect of condition ( t (105) = -2.95,p = .003). A follow-up analysis reveals that the proportion in the mask condition is not significantly different from 0.5, but the proportion in the no-mask condition is. (Figure 2).
We did not find an interaction between emotion and condition ( t (128)= -0.384,p = .7), but the main effect of emotion was significant(t (128)= 2.79,p = .005). The proportion of looking time in the happy emotion is larger than 0.5 but not in the sad emotion (Figure 2).
The main goal of the current study was to examine the influence of face mask wearing on emotion recognition. We found that occludes in the lower part of the face reduce the ability to recognize emotion across modalities in toddlers.

Authors

• Liat Israeli-Ran, Ben Gurion University of the Negev, Israel
  Presenting Author

• Florina Uzefovsky, Ben Gurion University in the Negev