Transforming the Telephone-Based National Household Travel Survey to the Internet: Application to University Students

摘要

The transformation of National Household Travel Survey (NHTS) andregional surveys to the internet is inevitable, partly because online surveys offeran efficient means of collecting data. Behavioral surveys are increasingly beingoffered as mixed mode, giving respondents the choice of filling out the surveyusing the internet, telephone, or other means. This study reports experience andlessons learned by transforming the computer-assisted telephone interviewmethodology used for NHTS to an Internet-based method for university students,who have ubiquitous access to the internet. The study reflects innovations insurvey research methodology, in the context of surveying students at largeuniversities in Virginia. This paper provides comparisons of two rounds ofbehavioral surveys conducted in 2009 and 2010, and develops statistical modelsthat quantify trip underreporting. In an attempt to closely mimic the NHTSinstrument, the internet instrument had a relatively high survey response burden.Trip underreporting was suspected in the first round of surveys, based on analysisof trip frequencies. This paper documents improvements in survey design thatwere intended to lower the survey response burden and reduced tripunderreporting. Statistical models are estimated to quantify how changes in theinstrument captured more trips. Results showed that survey design improvementscan encourage students to recall and report 15%-20% more trips. Discretionarytrips and shorter trips were more likely to be underreported, consistent withearlier studies. The results from this study are valuable for future regional andnational survey implementations.