Let Us Face It! a Meta-Analysis of Atypical Viewing Patterns in Individuals with ASD

Thursday, May 12, 2016: 11:30 AM-1:30 PM
Hall A (Baltimore Convention Center)
K. Evers1,2,3,4, J. Prinsen4,5, R. Van der Hallen1,4 and J. Wagemans1,4, (1)Laboratory of Experimental Psychology, KU Leuven, Leuven, Belgium, (2)Parenting and Special Education Research Unit, KU Leuven, Leuven, Belgium, (3)Department of Child Psychiatry, UPC-KU Leuven, Leuven, Belgium, (4)Leuven Autism Research (LAuRes), KU Leuven, Leuven, Belgium, (5)Research Group for Neuromotor Rehabilitation, KU Leuven, Leuven, Belgium
Background: Reduced eye-contact is one of the core symptoms of ASD, sometimes already present early in infancy. Some studies have indeed found that individuals with ASD made fewer fixations at faces, more specifically at eyes, and that they fixated mouths or the background more often than a control sample. Other studies, however, obtained no evidence for atypical viewing patterns in individuals with ASD. Differences are sometimes very subtle and the exact pattern of results appears strongly dependent on task and stimulus factors. 

Objectives: Instead of yet another empirical study evaluating the fixation patterns of individuals with ASD, the field is in need of a systematical and quantitative overview of all the evidence. We therefore wanted to examine and combine all available empirical data on this topic, by means of a meta-analysis, in which the effect size across different studies was calculated, evaluating the overall evidence for differences in social attention between individuals with and without ASD.

Methods: Our literature search yielded over 2,500 articles, of which all abstracts were further reviewed, applying a set of strict inclusion and exclusion criteria. The remaining set of 57 articles was included in our quantitative meta-analysis. Several moderator variables, such as participants’ age, gender, stimulus and task characteristics, and region of interest, were incorporated as moderator variables in our analysis. Hedges’ geffect sizes were examined.

Results: Overall, results provided evidence for a reduced saliency of faces in individuals with ASD (Hedges’ g effect size = -0.52, p < .0001), which was stronger for upper (g = -0.79, p < .0001) compared to lower (g = -0.32, p = .0341) face regions. These viewing pattern differences were partially overcome by using task instructions, compared with free-viewing tasks (F(1,208) = 5.96, p = .0154). Group differences appeared strongest in age groups between 12 and 25 years old, and for individuals with average IQ scores between 85 and 115.

Conclusions: This meta-analysis provided evidence for atypical viewing patterns in individuals with ASD, characterized by a reduced saliency of faces and – more specifically – all internal facial features, with moderate to strong effect sizes. The impact of several moderator variables will be discussed.