Eye Movement Patterns and Driving Performance

In the current study we examined the relationship between drivers’ eye movement patterns and driving performance in a dual-task driving paradigm. Drivers performed two tasks in a driving simulator. In a car following task, drivers were asked to maintain a constant headway from a leading vehicle as it varied its speed. In the second task (light detection task), drivers tried to detect changes in peripheral traffic lights. The performance in the car following task was measured with headway distance and RMS, and the performance in the light detection task was measured with response time and accuracy. We found that the frequency of fixations, fixation duration and spatial distribution of fixations were significantly correlated with drivers’ performance in the driving tasks. Specifically, driving performance improved with fewer eye movements, longer fixation durations and smaller spatial distribution of fixations.