Patterns of Proxy Usage in the 2001 National Household Travel Survey

• Activity Number: 165
• Type: Contributed
• Date/Time: Monday, August 9, 2004 : 2:00 PM to 3:50 PM
• Sponsor: Section on Survey Research Methods
• Author(s): Jonaki Bose*+ and Lee Giesbrecht
• Companies: Bureau of Transportation Statistics and Bureau of Transportation Statistics
• Keywords: NHTS ; travel data ; proxy responses

Abstract

The 2001 National Household Travel Survey is a nationally representative survey on the daily and long distance travel patterns of the nation. Data were collected from about 26,000 households about the travel behavior of all household residents. For adults, 18 and older, a proxy response was accepted if the person was unavailable to participate. Overall, data were collected from proxies for 22% of the adult respondents. This paper evaluates the quality of the data collected from proxy respondents, compares complete households with and without any proxy data, and examines the potential characteristics that might predict whether a person will have a proxy response. The ideal way to evaluate the quality of proxy data would be to have proxy- and self-reported data for each respondent. Despite the absence of such an experimental set-up, this study attempts to understand the phenomenon using survey data. It highlights some of the difficulties in deciphering whether differences in estimates between proxy and self respondents are due to true differences or measurement error, especially since the need for a proxy respondent may be directly linked to the topic of the survey.