Studying Gaze Behavior to Compare Three Different Hazard Perception Tasks

The aim of the current study was to compare the traditional, verbal, and motoric tasks regarding their contributions to hazard perception measurement.

Traditional hazard perception tasks require the participants to respond to filmed traffic conflicts in an imprecise way, such as by pressing a button. More sophisticated tasks include either verbal specification or motoric localization of the perceived hazards. The present study investigated the participants' gaze behavior when they were provided with an identical set of traffic animations but were instructed to perform one of three types of hazard perception tasks.

In an eye tracking study, 69 drivers were shown animated traffic scenarios and instructed to perform the traditional (press button), verbal, or speeded motoric localization hazard perception task. Eye tracking revealed whether and when the participant had fixated a certain hazard cue.

The participants in the traditional task group were slower to fixate emerging hazards, but quicker to respond to them than the participants of the verbal and the motoric groups. As a specific benefit, the verbal task differentiated between different types of failures.

Additional verbal or speeded motoric localization tasks seem to have increased the participants' alertness when watching the animations. The verbal task provides valuable additional information regarding the participants' performance. To approximate real-life hazard perception ability, it is recommended that researchers and practitioners use a combination of different hazard perception tasks for assessment and training.