A frequently-used source of data is the on-board bus survey, in which bus riders are asked to participate in some form of survey. While such surveys are administered with considerable frequency, there has been little written about good designs for these surveys, and little consistent attempt made to develop improved designs. This paper addresses the overall design of on-board bus surveys, the sampling to be used, coverage of the sample, design of instruments and supporting materials, and the potential scope of the surveys for data collection. Alternative survey designs are considered for activities that range from fare-sampling to an in-depth survey that can support many aspects of long-range transportation planning and modelling. Among the specific topics covered are the use of stratified, multistage samples; the use of drivers or on-board surveyors for the survey administration; the development of survey schedules for covering the sample; tracking nonresponse through the survey process; and the use of two-part survey forms for partial return on the bus and partial return by mail. The various elements of design are described in case studies conducted by the author. (Author/publisher) for the covering abstract of the conference see IRRD 286978.
Abstract