Using Capacity Analysis to Evaluate the Effect of MP3 Player Usage on Driver Behavior

Abstract

Objective: To develop measures of driver interaction with in-vehicle devices through capacity analysis. Background: Common measures of driver behavior focus on the mean values of measures that do not consider temporal distribution of attention. Capacity analysis can quantify the temporal profile of a drivers' capacity and the resulting task efficiency. Methods: Fifty participants performed search tasks on either an MP3 player or its aftermarket controller in a medium-fidelity driving simulator. A capacity analysis used a hazard function to describe tasks according to their probability of completion. Results: Capacity coefficients, calculated as the ratio of two integrated hazard functions, were not constant over time indicating that the probability of task completion is dynamic and must be described with more than a measure of the central tendency of distribution, such as mean task completion time. Conclusion: Along with other commonly used driving performance metrics, capacity analysis can reveal task completion dynamics that are often neglected with reaction time measures, providing insight into how drivers direct their attention away from the road to complete secondary tasks.