Similarities and differences in the perceptual structure of facial expressions of children and adults

Sad or fearful? The influence of body posture on adults’ and children’s perception of facial displays of emotion

“…This pattern of results is consistent with previously reported data from adults (Aviezer, Hassin, Bentin, et al, 2008) and is consistent with both circumplex models (Abelson & Sermat, 1962;Bullock & Russell, 1984;Schlosberg, 1952;Widen & Russell, 2008a) and the emotional seed model (Aviezer, Hassin, Bentin, et al, 2008) of emotion perception. This pattern of results extends previous research showing that, like adults (Carroll & Russell, 1997;Susskind et al, 2007), children's similarity judgments (Gao et al, 2010) and their errors in labeling and sorting facial expressions (Gao & Maurer, 2009;Russell & Widen, 2002;Wellman et al, 1995;Widen & Russell, 2008a) show systematic relationships among emotions, with some emotion pairs (e.g., fear and anger) being judged as more similar and more easily confused than others (e.g., fearful and happy). The current results suggest that when two emotions are similar (e.g., fearful and sad) context effects are large, but when two emotions are very dissimilar (e.g., happy and sad) context effects are absent.…”

“…Whereas fear and sadness both are negatively valenced emotions that differ only in arousal, happiness and sadness differ on both arousal and valence. Happy and sad facial expressions also share very few physical characteristics (Susskind et al, 2007), and multidimensional scaling indicates that children perceive happy and sad faces to be less similar than fearful and sad faces (Gao et al, 2010). Aviezer and colleagues (2008) reported that the influence of body posture is reduced when the emotion conveyed by the face is very different from that conveyed by the context.…”

“…A growing body of research suggests that children first perceive and understand emotions at the level of valence; by 2 years of age, they categorize ''happy'' separately from all negatively valenced emotions and only later discriminate among sadness, fear, and anger (i.e., negatively valenced emotions that differ in arousal) (Russell & Widen, 2002;Wellman, Harris, Banerjee, & Sinclair, 1995;Widen & Russell, 2008a). By 8 years of age, children sort prototypical expressions much like adults do (Kolb et al, 1992;Russell & Bullock, 1985), although the underlying dimensions are not fully adult-like until later (Gao et al, 2010). Because 8-year-olds appear to be sensitive to the underlying dimensions of valence and arousal, I predicted that their perception of facial expressions would be influenced by body posture-particularly when the emotion conveyed by the face and the body posture were similar in arousal and/or valence.…”

Wide eyes and drooping arms: Adult-like congruency effects emerge early in the development of sensitivity to emotional faces and body postures

“…The fact that 6-year-olds show congruency effects when tested with stimuli depicting sad/fear is consistent with evidence that children perceive these emotions to be relatively similar to each other (Gao et al, 2010;Russell & Bullock, 1985). Like adults, children's congruency effects may be influenced by the similarity of the emotions conveyed in the face and context.…”

“…That sensitivity emerges during the preschool years. When asked to sort a set of 20 facial expressions into a varying number (2-10) of piles in which everyone in the same pile felt alike, a measure of similarity suitable for testing 4-and 5-year-olds, preschoolers' representation of emotion was best characterized as a circumplex with two underlying axes, namely valence and arousal (Russell & Bullock, 1985; see also Russell & Widen, 2002;Wellman, Harris, Banerjee, & Sinclair, 1995;Widen & Russell, 2008a), like older children and adults (Gao, Maurer, & Nishimura, 2010), although sensitivity to arousal takes time to develop (Boyatzis et al, 1993;Camras & Allison, 1985;Widen & Russell, 2003).…”