Laboratory and Closed-Course Measures of Driver Decision Making

Abstract

A technique for measuring two-choice reaction time using turn-signal activation in a stationary or moving vehicle was implemented on a closed driving course. Fifteen subjects, aged 26 to 45 completed five tasks, one in the laboratory, two in a stationary vehicle and two in a moving vehicle. The lab task included simple, choice and movement reaction times. Stationary-vehicle tasks included simple and choice reaction time and a memory/attention task implemented on a changeable message sign. Moving-vehicle tasks included timed lane selection in response to traffic signal changes and the memory/attention task. Data obtained in the vehicles were adjusted for differences in turn-signal activation characteristics through subtraction of individual vehicle delay constants. Lab and stationary-vehicle measures of choice reaction time correlated significantly with response times in driving-related tasks. Intercorrelations among driving-related tasks were strongest between moving-vehicle tasks, suggesting that controlling a moving vehicle exerted a stronger influence on individual differences in performance than did the cognitive requirements of the decision tasks. Tasks where stimuli appeared outside the vehicle at a distance were associated with higher variability than tasks using stimuli near to the subject. Correlations between replications of the same task were supportive of good reliability. The overall pattern of results suggests that drivers were more attentive during the first session, and less so during the second and third sessions. Performance in the first session may reflect drivers' capabilities, while subsequent performance may be more representative of the reduced attention given to everyday driving. Implications for the development and administration of the driving test battery are discussed.