An Empirical Assessment of the Role of Driver Motivation and Emotion State, and Driving Conditions in Perceived Safety Margins

Abstract

Models of motivation in driver behavior have been developed to predict driver performance under different conditions. Among existing models, Näätänen and Summala’s multi-dimensional threshold model (1976) was selected as basis for the current study. Näätänen and Summala proposed that driver behavior is modified not only according the degree of difficulty of traffic situations, but also based on driver risk tolerance in specific tasks. They also observed that risk-taking is based on driver motives and emotional states. According to this model, various factors may influence driver behavior. For instance, traffic patterns (other driver behavior) and driver motivation to comply with social norms, extreme emotions triggered by special circumstances (emergencies), and long-term emotional tendencies. The latter factor has been assessed using Driver Stress Inventories (Matthews, Desmond, Joyner, Carcary, & Gilliland, 1996). Näätänen and Summala’s model also indentified measures of change in driver risk-taking decisions, or safety margins, as being predictors of driver performance. The objectives of this study were to: 1) provide empirical evidence of the influence of motivation and emotional factors, as identified by Näätänen and Summala, in driver risk-taking behavior; and 2) identify any additional variables that might mediate the effects of motivational factors on behavior, including roadway environment complexity. The study examined the following specific factors: 1) traffic patterns, including traffic jam, school zone, normal traffic flow and speeding conditions to assess the influence of social norms on driver behavior; 2) driver payment systems, including time-based and performance-based compensation to assess the influence of extreme emotions on performance; and 3) environment complexity, including rural and city conditions to assess the influence of changes in task difficulty on behavior. Response measures included safety margins and speed measures. Safety margin measures consisted of: 1) spatial variables, including headway distance (HW) and lateral distance (DH); and 2) time variables, including time headway (THW), time to collision (TTC) and time to line crossing (TTLC). Speed measures consisted of average speed, maximum speed and the percentage of time spent speeding. Ten participants drove a virtual car in a high-fidelity simulator and performed daily driving tasks (e.g., lane maintenance, lead-car following, passing, negotiating intersections, etc.). A split-plot experiment design was used with the whole-plot (trial) factors including environment complexity and the payment system. Traffic pattern was manipulated as the split-plot factor with each of the four patterns occurring during a single segment in each trial. Participants complete eight test trials, including two replications all combinations of complexity and payment system. Participants were also required to finish a DSI prior to simulated-driving tasks. The experiment results revealed the effect of the motivational factor/payment system. More risky driving behavior was associated with the performance-based payment system compared to the time-based system. The influence of environment complexity was also observed. Smaller safety margins appeared in the rural environment as compared to city. The effects of traffic pattern were significant across all response measures except TTLC: traffic jams led to minimum safety margins; speeding segments produced the highest driving speeds and largest safety margins; school zones were associated with conservative behavior, including lower speeds and larger safety margins. The traffic pattern also interacted with the roadway complexity condition in terms of THW. Drivers were influenced more by the behaviors of other drivers in the city versus rural setting. Correlation analyses showed significant linear associations between long-term emotional tendencies and safety margin and speed measures. In summary, the current research contributed to the further development of motivational models of driving behavior by providing reliable evidence of factors that are significantly influential in perceived safety margins and performance. The study also identified additional (lateral vs. longitudinal) measures for sensitively specifying safety margins. Future research should investigate a broader range of emotional factors that may be related with safety margins. The role of long-term emotional tendencies in driver performance should also be studied under a broader range of conditions, including roadway hazard exposure. In addition to this, future work might examine more diverse driving populations (beyond college students) to obtain a more comprehensive understanding of motivational/emotional factors in driving performance.