Publication Detail

Survey of Route Choice Behavior: Empirical Results from Southern California and Their Implications for Advanced Traveler Information Systems


Research Report

Suggested Citation:
Abdel-Aty, Mohamed A., Kenneth M. Vaughn, Ryuichi Kitamura, Paul P. Jovanis (1993) Survey of Route Choice Behavior: Empirical Results from Southern California and Their Implications for Advanced Traveler Information Systems. Institute of Transportation Studies, University of California, Davis, Research Report UCD-ITS-RR-93-12

A computer-aided telephone interview (CATI) survey was designed to gain a basic understanding of drivers' route choice behavior, to collect detailed information about their commute routes, and to understand their perceptions about their commute. There is still little understanding of commuters' route choice, which is essential for the development of ATIS systems. The survey was undertaken in May and early June of 1992, targeting Los Angeles area morning commuters. The survey employed a Mitofsky-Waksberg cluster sampling design, and yielded 944 completed surveys. A comparison between the survey data and the 1990 census, and the distribution of the main socio-demographic characteristics of the sample, showed that the sample well represents the population.

Cross tabulations were performed on the data to detect any specific trends in the commuters' behavior, and to investigate interrelationships among the different variables in an attempt to develop an understanding of factors that may influence route choice behavior. Two different modelling attempts were also made. The first was to model whether an individual will follow the same route to work every day or not, using a binary logit model formulation. The second was to model the frequency of route changes per month based on pre-trip or en-route traffic reports, using Poisson regression.

The descriptive statistics and the models performed is a basis for later network based analysis. The results indicated that a large percentage of the commuters perceive little variation in traffic conditions from day to day, and that traffic conditions on their usual route are reasonable. The majority also indicated that traffic is moderate on the freeway segments they use. Respondents who perceived variable and/or bad traffic conditions were more likely to listen to traffic reports. More females used pre-trip traffic information than males, while more males sought en-route information. Only 15.5% of the sample reported that they do not follow the same route to work every day, indicating a very promising potential benefit from an information system that would make more people familiar with alternative routes. Freeways were less often used as part of secondary routes, possibly as an alternative for commuters to avoid incident congestion on freeways. High income, high education level, departure time, and traffic variation on the usual route, were among the factors that contributed to commuters not following the same route to work every day. The logit model results was very much along the same line, and predicted 80.8 percent correct that a respondent will follow or not the same route to work every day. The Poisson models illustrated the importance of traffic information on the flexibility in changing routes.