Abstract
Methods to assess the impact of in-vehicle information and guidance on drivers' route choice are reviewed. The authors then describe a method using stated preference techniques. The content of questionnaires is outlined and the results analysed. The analysis considers the accuracy of perceived journey times and compliance as a function of implied diversion time. A disaggregate model is then applied to the data. The results show the effect of different messages. The `10 minute delay' message was more effective than `roadworks', `5 minute delay' or `some delay likely' messages. The use of messages by traffic system managers and the limitations of the questionnaires are discussed.