The Impact of Traffic Images on Mode Choice in Stated-Preference Surveys

Abstract

A difficult aspect of using stated-preference choice experiments to predict travel behavior is properly presenting attributes and characteristics of hypothetical trips to respondents. With the growing in number of transportation choices recently, the task of concisely and accurately communicating trip attributes in the stated-preference setting become increasingly more important. Recent attempts to introduce innovative strategies to the stated-preference setting have yielded techniques to more efficiently summarize trip attributes to respondents. One technique is to use images of traffic conditions as a supplemental means of summarizing average trip speed, travel time reliability, or degree of congestion. However, little research has been performed testing the effect that the use of traffic images has on models of mode choice built from this kind of stated-preference data. In this research, a stated-preference setting was developed in which the influence that images of traffic conditions was measured. Pictures of traffic conditions that correlated to average trip speed were either shown or withheld depending on random assignment to a survey population from Austin, Texas. From the significant differences in respondent preferences across mode choice, a mixed-logit model was built to describe the respondent's choice behavior. Overall model parameters discovered no evidence to support the assertion that traffic image presentation has a statistically significant effect on mode choice with respect to Value of Travel Time Savings, or Value of Travel Time Reliability.