Poster #49 - The Own-Age Face Bias is Present In Autism, but Contingent on Other-Age Faces

• In Event: 1-115 - Poster Session 03
  In Poster Session: PS 03 Section - Developmental Disabilities

Integrative Statement

A common interpretation of the face-processing deficit associated with autism spectrum disorder (ASD) is that it is resultant from a failure to develop normative levels of developmental expertise (Webb, Neuhaus, & Faja, 2017). This failure may be due, in part, to an inability to abstract prototypic representations from multiple exemplars (Best et al., 2005). The own-age bias posits that individuals demonstrate a processing advantage for faces of their own age, presumably due to more frequent and recent experiences (Rhodes & Anastasi, 2012). This effect is especially evident in domains of face recognition memory but less commonly investigated in social-emotional expertise (e.g., facial emotion recognition [FER]) where individuals with ASD have shown consistent deficits (Ebner & Johnson, 2009; Lozier, Vanmeter, & Marsh, 2014). We hypothesized that individuals with ASD would not exhibit an own-age bias in FER at the group level, but that a tendency toward this bias would be associated with reduced symptom severity on an individual level.

Eighty-six adolescents completed a standardized FER task (Table 1). Participants were classified into one of three groups: typically developing (TD), autism spectrum disorder (ASD), or autism using cut-offs determined by the Autism Diagnostic Observation Schedule-2 (Lord et al., 2012).

Regardless of group, adolescents made more FER errors when viewing adult faces compared to child faces (TD: t(33) = 5.33, p < .001; ASD: t(8) = 2.49, p = .04; autism: t(42) = 3.80, p < .001). The diagnostic groups did not differ from one another in the number of FER errors made for adult (F(2, 83) = 1.75, p = .18) or child (F(2, 83) = 2.27, p = .11) faces. Interestingly, diagnostic category was positively associated with FER errors for child (r(84) = .22, p = .04) but not adult (r(84) = .17, p = .11) faces. However, this relationship was moderated by adult FER errors (t(82) = -2.12, p = .04), such that the relationship was present when participants made few (b = .11, p = .02) and moderate (b = .08, p = .04) levels of adult FER errors, but not when they made many such errors (b = .05, p = .19; Figure 1).

Contrary to our first prediction, results indicated an intact other-age bias in adolescents with ASD. However, the number of FER errors adolescents made when viewing child faces differentially predicted ASD diagnosis, as per our second prediction, suggesting that the relative inability to accurately identify emotional facial expressions in peers is a greater predictor of ASD diagnosis than a broader deficit in FER. However, this was contingent upon the level of adult face FER performance such that the presence of an own-age bias was only related to a reduction in symptom severity if adolescents were, at the least, moderately adept at adult FER. This suggests that, for adolescents whom have developed a sufficient level of FER expertise, a priority for attending to, abstracting, and integrating peer facial representations is associated with lower ASD symptom severity.

Authors

• Kathryn M Hauschild, Stony Brook University
  Presenting Author

• Matthew Daniel Lerner, Stony Brook University
  Non-Presenting Author