This paper analyzed repeated measures of drivers' opinions on the effectiveness of six standard railroad grade crossing warning devices and three modes of an in-vehicle receiver (IVR), which are an ITS concept. The warning devices were: advance warning sign, crossbuck sign, flashing lights, crossing gate, clanging bell, and train horn. The IVR operating modes were: audible, visual, and a combination of both. Drivers' opinions were collected using four opinion surveys over an 18-month period. Paired t-tests were used to show how drivers perceived the effectiveness of the standard warning devices compared to each other. An analysis of repeated measures using general linear models suggested that time lag between surveys could be important in the change of drivers' opinions for some devices. The sequence in which drivers were exposed to the IVR operation modes was found significant in the IVR assessment, indicating the existence of carryover effects. IVR with a combination of audible and visual warning messages was slightly better than IVR with only audible messages and IVR with only visual messages. The effectiveness of IVR in combination mode was found similar to train horn and clanging bell, but lower than flashing lights and crossing gates. IVR in combination mode was either similar or better than advance warning sign and crossbuck, depending on the sequence drivers previously experienced other IVR modes.
Abstract