Abstract:Choice-based sampling has become widespread in transportation planning studies due largely to rising costs of surveys and the need to study infrequent travel choices. More recently, several panel surveys have been, or are being, conducted in the United States and Europe to measure changes in travel behavior, with a sample much smaller than repeated cross-sectional surveys would require. These two survey methods can be combined as a panel survey with choice-based sampling to economically produce data that facilitate the analysis of changes in infrequent choices. In choice-based sampling, where sample entities are selected based on the endogenous variable under study, the resulting sample will not be representative of the general population. The issue is further complicated with a panel survey because of selective attrition. The study is the first attempt to develop weights for a choice-based transportation panel sample. In this paper, weights are developed to correct for potential biases due to choice-based sampling and attrition. The weights are applied to a data set from the Puget Sound Region to produce population estimates of transitions in discrete travel choice.