Citation

Heinen E, Maat K. Transp. Res. Rec. 2012; 2314: 72-80.

Copyright

(Copyright © 2012, Transportation Research Board, National Research Council, National Academy of Sciences USA, Publisher SAGE Publishing)

DOI

10.3141/2314-10

PMID

unavailable

Abstract

Mode alternation in the Netherlands was compared, and data from a longitudinal survey were compared with a single-moment survey that focused on bicycle commuting to evaluate its reliability. Travel data are usually collected at a single moment in time. The use of single-moment survey data to investigate variable behavior raises questions about the reliability of such data. Repeated measures, resulting in longitudinal data, may be the solution because in that case individuals would have to report only 1 day at a time, a relatively easy task. Collection of longitudinal data, however, is more expensive and takes more time. Analyses showed that mode alternation occurred frequently, especially for car and bicycle commuters. Single-moment surveys appeared to be unable to measure this accurately. Many respondents inaccurately reported being part-time cyclists although they always cycled to work or, conversely, claimed to be full-time cyclists although they varied modes. Moreover, many did not accurately report cycling frequency in a single-moment survey. However, it is difficult to determine the characteristics that predict the probability of an accurate report. In addition, the degree of accuracy is not readily linked to individual characteristics. Thus, it appears that the error in single-moment surveys cannot be easily corrected. Even repetitive behavior is subject to much variation, so surveys should include specific questions that would increase insight into this phenomenon and transport models should include mode variation. For this innovation in transport models to occur, longitudinal data must be collected and analyzed.