Driving Hazard Perception in Autism: An Eye Tracking Study

Background: Klin, Jones, Schultz and Volkmar (2003) identify driving as a ‘challenging task’ for individuals with ASD. However, an increasing number of individuals with ASD have been applying for driving licenses. Hazard perception is a component of driving that may pose particular problems for people with ASD, as many hazards involve the perception or anticipation of another person’s intentions and movements. In a previous study we found that participants with ASD were less accurate at detecting hazards that were social in nature i.e. contained a visible human figure (Sheppard, Ropar, Underwood & Van Loon, in press). Participants with ASD had longer response times for both kinds of hazard, but it was unclear whether this was due to slowness in perceiving the hazard or difficulties in making the response itself.

Objectives: This study used eye tracking to elucidate reasons for the slowness of those with ASD during a driving hazard perception task by determining when the hazardous object was first fixated by participants with and without ASD.

Methods: Eighteen males with HFA or AS, and 17 matched comparison participants viewed 20 video clips containing driving hazards. In half of the clips the cause of the hazard was a visible person (social), whilst in the other half the cause was a car (non-social). Participants were instructed to respond with a button-press as soon as they saw a hazard developing. They then identified the hazard verbally. Accuracy and timing of the responses was recorded. Participants’ eye movements were recorded using a tobii portable eye-tracker.

Results: Participants with ASD correctly responded to as many hazards as comparison participants. However, they were less likely to anticipate the driving hazards, by responding prior to their onset (U=91.5, N=35, p<.05). Also, number of anticipatory responses correlated negatively with AQ score (Kendall’s tau_b=-.45, N=35, p<.0005). Results of the eye tracking analyses will also be reported.  

Conclusions: The results here support the notion that people with ASD have difficulty perceiving and detecting driving hazards, and that this difficulty may relate to strength of autistic traits (self-reported). Implications of the results of the eye tracking analyses will also be discussed.