COMPLEX POSTAL SURVEYS IN TRANSPORT PLANNING: METHODOLOGY AND RESPONSE RATES

The paper compares and contrasts two travel surveys using postal survey methods to elicit complex data. One group of surveys were part of a study of long distance travel in the Amsterdam-Groningen corridor in Holland. A sample of journeys over 30 km by train, bus and car was required and, because such travel represents only a small part of all travel, a choice-based sample was decided upon. The only way to collect the data economically was by post and yet the volume and complexity of the data required from each individual was very great. A multi-stage survey was undertaken with travellers by the three modes completing a short factual questionnaire. This questionnaire was processed quickly, typically within one day, to enable a sample to be selected for a follow-up questionnaire. The method allows flexibility in sample selection at the second stage - including the possibility of basing it on various criteria derived from answers to the first stage. This also enables the researcher to devise different second-stage questionnaires for, say, different journey purposes or market segments. In Holland 12 different questionnaires were used. The other study, in the West Midlands, adopted choice-based sampling for bus and rail travellers, with home interviews for car users. In the context of urban and sub-urban trips, especially at peak periods, travel times and other conditions on public transport dictated as short a first stage questionnaire as possible. The resulting data provided a base matrix of movement in the corridors of interest. Specific localities, and hence sub-sets of respondents were then selected for second-stage postal surveying to obtain more detailed data on which to calibrate travel demand models. The paper will describe the theory and methodology of this technique, analyse the response rates achieved, comment on bias and how it can be controlled and draw general lessons from the studies considered, for use in other applications. (Author/TRRL)