Proxy Respondents in Household Travel Surveys

Abstract

Most current transportation planning models use household travel/activity data that are collected through household/activity surveys. These surveys collect travel and activity information from each person in the sampled households. Because data from complete households are desired, travel/activity information about a house member who is unable or unwilling to cooperate is sometimes accepted from another person, i.e., a proxy response. A problem with proxy responses in surveys is that there may be differences in the types of people who are likely to respond for themselves and those who have some other person respond for them. If proxy respondents answer differently from self respondents, and the proportion of proxy respondents is large, or concentrated in particular subgroups of respondents, then the travel estimates based on these data may be biased. Knowledge of the accuracy of proxy records is important because, if records obtained by proxy differ significantly from records that are self reported, transportation planners need to be aware of, and adjust for these differences when using these data in modeling and planning applications.