Assessing the Allocation of Visual Attention In Adults with Autism Using a Change Detection Paradigm

Background:  Change detection, the ability to detect changes in the environment, can be affected by the type of information that is prioritized. Among typically developing (TD) adults, change detection is more efficient when change occurs in foreground (i.e. the front plane) compared to background elements (i.e. the ground) of geometric visual displays. One question is whether adults with autism present the same spontaneously-biased attention allocation to foreground information. Another question is how perceptual characteristics of the visual scene, such as configurality (generally not prioritized among autistic individuals) and dynamicity (used as an external attentional cue) interact with higher-level attentional processes solicited by a change detection task.

Objectives:  The objectives of this study were to assess (1) whether the typically-manifested attention bias for foreground information is displayed by individuals with autism, and (2) to what extent, if any, manipulating the configurality or dynamicity of foreground information differentially influences visual attention allocation among autistic as compared to TD adults.

Methods:  Fifteen adults with autism (18-35 years) with average IQs and 15 gender- age- and IQ-matched TD adults were administered an adapted version of the change blindness paradigm from Mazza et al. (2005). The background consisted of 20 columns (alternating between light and dark grey) comprising ten vertically-oriented rectangles (1.81 ^o x 1.24 ^o).  The foreground consisted of 6 horizontally-oriented rectangles (3 light grey and 3 dark grey) arranged in either a circular (configural) or random (non-configural) manner. On a given trial, either a foreground change (horizontal rectangles changed shades of grey), a background change (vertical rectangles changed shades), or no change occurred between two successively presented displays (500 ms in duration). In the static condition, no motion occurred within the rectangles. In the dynamic condition, either the foreground or background rectangles contained dynamic noise. The participants indicated whether or not they perceived a change between the two successively presented displays.

Results:  For the static condition, both groups were more efficient at detecting foreground relative to background change. Moreover, the TD adults were more likely to detect the background changes than the participants with autism. In addition, the TD participants were better able to use dynamic information as a cue to detect changes (either background or foreground). However, only the participants with autism were influenced by the configurality of the elements, as they detected more changes when the foreground rectangles were randomly organized compared to when they were presented configurally.

Conclusions:  Different patterns of prioritization of visual attention were found between adults with autism and TD adults on a change detection paradigm when perceptual attributes such as the configurality of the foreground elements were manipulated. As compared to the TD participants, the participants with autism processed information more locally and were less likely to orient their attention to the dynamic elements of a scene. This suggests that atypical perceptual information processing in autism may influence higher-level cognitive functions, such as visual attention.