Y. Xu, A. Greenwood, G. Corso, M. Hunter, M.O. Rodgers

Pages: 113-126

Abstract
This paper presents an experiment in which participants were asked to identify freeway diverge locations under different work zone settings. A total of 130 participants were shown computer rendered images of freeway diverges and tasked with identifying the diverge location by clicking on the image with a mouse. The evaluation method employed the participant’s response time as one of the important surrogate safety measures, along with the response accuracy and the types of errors observed. The results show that response times are closely correlated with other measures in indicating participant performance, although some differences exist. Specifically, work zone delineation methods that show greater numbers of correct responses also tended to feature shorter response times. However, the use of correct response time reveals nuanced differences among channelizing devices that otherwise would not be recognized if the participant responses were only analyzed for accuracy. This was particularly true for a proposed linear channeling device where accuracy and response time demonstrated a wider divergence. Using a framework that simultaneously assessed response accuracy, error types, and response times, portable concrete barriers were found to be the most effective in assisting drivers with locating the freeway diverge in work zones. In contrast, drums were found least effective. The proposed linear channelizing device showed mixed results among the performance measures. The study method and results show the potential importance and relevance of response time as a possible surrogate safety measure in a comprehensive evaluation of innovative traffic control devices. The inclusion of response times not only reinforces, but also adds finer detail to methods that often only focus on response accuracy.

Keywords: traffic safety; freeway diverge; channelizing device; response time