SURVEILLANCE AND MONITORING OF A BUS SYSTEM.

Most transit operators occasionally conduct an on-board survey of riders. Based on experiences in Washtenaw County (Ann Arbor) Michigan, Dade County (Miami) Florida, and Honolulu, Hawaii, this paper examines three aspects of such surveys. First, a survey instrument is described that permits considerably more information to be collected than is possible from the traditional postcard type of on-board survey. Descriptions of the types of data needed to be collected on the participatory self-administered survey of riders for both systemwide surveillance and individual route monitoring are provided. Second, procedures are described for reducing nonresponse bias for collecting at least some information from a subgroup of riders who would otherwise be nonrespondents. Third, sampling strategies (including the necessary sample sizes) are described both for systemwide surveillance and individual route monitoring.