BackHome
Refresh: OffView Personal Schedule
Individual Submission Summary
Share...

Direct link:

Poster #208 - Perception of Facial Emotions of Varying Intensities in 3-Year-Olds

Thu, March 21, 9:30 to 10:45am, Baltimore Convention Center, Floor: Level 1, Exhibit Hall B

Integrative Statement

Facial emotion recognition emerges early in life and is considered a critical building block to social-emotional development. Notably, infants differentiate happy from angry and fearful expressions from 5-7 months and exhibit behavioral and neural biases towards these expressions (Leppanen & Nelson, 2009). However, little is known about facial emotion recognition between 2-4 years of age, particularly regarding subtle facial emotions. Here, we used an explicit behavioral task to investigate behavioral responses to facial emotions of varying intensities in 3-year-olds.
A relatively large sample of typically-developing 3-year-olds completed a behavioral task of facial emotion recognition (N = 208, mean age 38 months, range 36-45). Children were presented with happy, angry, and fearful faces at 7 levels of intensity (20%, 30%, 40%, 50%, 60%, 70% and 100%), and 1 neutral face, in random order; children sorted these faces into four distinct houses representing these four emotions (Gao & Maurer, 2009, 2010; Moulson et al., 2015). Critically, the task required an overt, explicit response but did not require verbal labelling.
Behavioral intensity thresholds (i.e., inflexion points of the observed psychometric curves) for facial emotion recognition were consistent across emotions at about 35% intensity, with no significant differences between expressions (nonlinear least-square models, non-parametric 95% confidence intervals obtained from N=10,000 bootstrap samples). Behavioral accuracy in response to maximal intensity expressions (i.e., the upper asymptote of the fitted psychometric curves) was highest for happy (89.85%) followed by angry (80.03%) and fear (61.30%) expressions. Accuracy for fear was significantly lower than for happy and angry, but only marginally lower for angry than for happy (non-parametric 95% confidence intervals obtained from N=10,000 bootstrap samples). High-intensity fear and angry expressions were significantly more confused with one another than with the happy or neutral expressions. Neutral faces were categorized as emotional (i.e., non-neutral) 62.02% of the time and were interpreted as happy more often than as angry (χ2[1] = 22.51, p < .001) or fearful (χ2[1] = 11.67, p < .001).
A subsample of these children (N = 125) also completed a functional near-infrared spectroscopy (fNIRS, 46-channels Hitachi) assessment during which hemodynamic responses to dynamic neutral, happy (100%), fearful (40%, 100%), and angry (40%, 100%) facial expressions were measured in bilateral fronto-temporal regions of the superficial cortex. Forty-seven of these children contributed valid fNIRS data, the ongoing analysis of which will extend the current behavioral findings. Briefly, fNIRS channels will be grouped into regions of interest defined anatomically based on their most probable cortical source (Perdue et al, in revision); average increases in oxy-hemoglobin concentration in each ROI will be derived for each condition, contrasted between conditions, and related to the behavioral performance.
To our knowledge, this is the first study investigating the recognition of facial emotions of varying intensities in preschoolers (2-4 year-old). We found no difference in perceptual thresholds between emotions, in contrast to existing reports at older ages. In addition, 3-year-olds exhibited a marked tendency to interpret neutral faces as positive, and were less accurate at recognizing high-intensity fearful faces compared to angry or happy faces.

Authors