Determinants of Attrition Rates in Two-Wave, Two-Day Household Activity Diary: Probit Analysis

Abstract

Transportation researchers often have used panel data to examine the dynamics of travel behavior. Although panel data are probably the best means by which to better understand choice dynamics, they present various challenges to the analyst, with attrition or dropout one of the most critical. Presented are the results of an analysis of the determinants of attrition in a two-wave panel, conducted in the Netherlands. Panelists reported their activities for a consecutive 2 days with 1 year between. Attrition rates between the two waves were predicted as a function of sociodemographics; number of activities, trips and tours; and the amount of reporting error in the first wave, by using a probit model. The results of the study indicate that none of these variables was significant, suggesting that no systematic attrition occurred between the two waves.