In Phase II of this investigation, we used a fully interactive PC-based STISIM driving simulator, to conduct two experiments which were similar to experiments in Phase I. The participants were 120 licensed drivers from three age groups-18-24, 32-47, and 55-65 years old-who drove, in free flow traffic, for approximately 20 miles on a four-lane freeway before encountering target messages on Changeable Message Signs (CMSs). The Phase II CMS messages were clearer and less complex than those used in Phase I. In the first experiment the target message was changed to "Road Closed/Crash Ahead/Use Thompson Exit:" 93.3% of the participants took the exit-as compared to 55.8% in Phase I. In the second experiment, the message was changed to "Abducted Child/Tune To/Radio 88.5 FM:" 71.7% of the participants (71.7%) could remember enough information to enable them tune to 88.5 FM-as compared to 8.3% who could recall some vehicle information and at least five license plate letters and numbers from the Phase I CMS ("AMBER Alert/Red Ford Truck/ MN Lic# SLM 509"). Some participants reduced speed on approaching the CMSs, suggesting similar reductions could occur in real world driving in free flow conditions. However, when traffic is congested speeds are typically slower, and drivers are less likely to reduce speed still further to read CMSs.
We conducted a survey which showed drivers think it is very useful to have information about traffic problems and roadway maintenance schedules on CMSs. Also, we analyzed real-world traffic speed data obtained when CMS messages were deployed. Finally, we conducted observations at Mn/DOT's Regional Transportation Management Center (RTMC), focusing on the decision-making processes involved when traffic-related CMS messages are deployed.