RESPONSE PATTERNS IN TRAVEL SURVEYS: THE VATS EXPERIENCE

This paper examines two key questions concerning mail-out, mail-back travel surveys involving a travel diary. The first is whether respondents to such surveys are likely to be dominated by travellers with less complex travel patterns. The second is whether repeat reminders, long perceived to be essential in increasing response rate, and the conduct of face-to-face interviews, deemed critical for correcting non-response bias, are effective in rectifying such problems of selectivity bias. Drawing on four years of VATS results, the household and travel characteristics of respondents are analysed with respect to the elapsed time in responding to the survey. The results show that follow-up reminders do play a significant role in raising response rates, especially in increasing the participation rate of families with children. The information obtained from late responses, however, suffers from under-reporting of out-of-home stops and a higher percentage of missing values. By contrast, non-response interviews prove to be a useful supplement to mail surveys, although missing data is also a cause for some concern. The study also reveals that non-respondents from similar family types have comparable trip rates to those of the early respondents. The findings confirm that travel surveys are susceptible to the risk of selectivity bias in the sample. They also challenge conventional thinking that non-respondents are more akin to late respondents and make fewer trips than respondents. To the extent that under-reporting of trips poses a more serious problem than total non-response in trip generation models, the value of repeat reminders is questioned. Non-response interviews are considered a favoured option when it comes to reducing non-response bias in self-administered, postal surveys. (a) For the covering entry of this conference, please see ITRD E200461.

Language

  • English

Media Info

  • Features: References;
  • Pagination: p. 591-608

Subject/Index Terms

Filing Info

  • Accession Number: 00796372
  • Record Type: Publication
  • Source Agency: ARRB
  • ISBN: 0-7307-2496-4
  • Files: ITRD, ATRI
  • Created Date: Aug 2 2000 12:00AM