Visual Attention during Narration in Autism: A Cross-Cultural Study

Background: Differences in visual attention have been documented in ASD, and are thought to importantly relate to clinical-behavioral features (Frazier et al., 2017). However, most of this work has been conducted in Western countries, where attentional bias toward salient information (particularly social stimuli) has been observed. Cross-cultural investigation is important to evaluate potential environmental/cultural influence on visual attention in ASD. In East Asian cultures, attentional biases toward contextual or less focal information (e.g., making judgements about a character based on the contextual setting) is commonly observed (Nisbett et al., 2001; Kitayama et al., 2003; Blais et al., 2008), similar to the attentional biases reported in ASD in Western cultures. This study investigated cross-cultural differences in visual attentional during narration in ASD across Western (US) and East Asian (Hong Kong; HK) cultures, with the prediction that the ASD-HK group would exhibit similar social attention patterns to Control-HK and ASD-US groups, but both ASD groups would differ significantly from control-US individuals.

Objectives: To characterize the cultural influence on visual attention patterns in ASD.

Methods: US participants included 27 individuals with ASD and 82 controls. Age- and IQ-comparable samples were included from HK (n=37 ASD, n=61 control). Gaze was tracked while participants completed two narrative tasks differing in structure and content (a structured wordless picture-book (PB), and a less structured more emotionally evocative task involving six images selected from the Thematic Apperception Test (TAT) (Murray, 1943)). Proportions of fixations to face, bodies, and setting were recorded, and in the TAT, analyzed across scenes differing in dominance of animate (i.e., face-dominant or body-dominant) vs. contextual scene (i.e., setting) information.

Results: The HK ASD and control groups made more setting fixations compared to both US groups in the structured PB task (ps<.01). In the TAT, during setting-dominant images, a hierarchical pattern emerged with the control-US group making the fewest setting fixations compared to control-HK, followed by ASD-US, followed by ASD-HK, who made the greatest number of fixations towards the setting (ps<.05). During face-dominant images, the ASD-HK group made fewer face fixations than the ASD-US group (p<.05). Finally, in a body-dominant image depicting interacting characters, both control groups made more body fixations than the ASD-HK group (ps<.05), and both US groups made more face fixations than the ASD-HK group (ps<.05).

Conclusions: Cultural differences in gaze emerged and varied by context. During the more structured task, both HK groups differed in setting fixations from the US groups, but only individuals with ASD (regardless of culture) fixated more towards the setting during the less structured, more emotionally evocative TAT task. The hierarchical pattern that emerged in fixation to setting across cultures by groups suggests an interplay between cultural influence and ASD-status on visual attention that is impacted by narrative structure and emotional content. Further, attention towards faces was reduced across both HK groups, suggests a cultural influence observed particularly in images depicting human interactions characters.