Abstract

The state-of-practice in the design and conduction of mobility surveys in metropolitan areas leaves much to be desired, even in most industrialized countries. It is argued that the state-of-the-art in this type of survey is a continuous (ongoing) method, beginning with a somewhat larger sample size in the first year and with smaller sample sizes in subsequent years. The paper discusses three issues related to this methodology. First, the sampling implications; in particular, sample sizes are considered based on stratifications by income, car ownership and household size, designed to estimate trip rates, zone-level car ownership and mode choice models for different strata. Second, the survey method implications, where state-of-the-art and -of-practice considers collecting data for each day of the week during the whole year. Third, the important issues of data correction and weighting, and the need for independent data and validation strategies. The paper concludes with a brief report on the main results of a 1-year pilot study of these recommendations as applied to Santiago, Chile, and with some preliminary data about the first wave of the large-scale mobility survey currently under way in that city that serves to illustrate the magnitude of the effort and to get a feeling about the quality of the approach. © 2004 Taylor and Francis Ltd.