Comparing alternative methods of collecting self-assessed overnight long-distance travel frequencies. (April 2020)

• Record Type: Journal Article
• Authors: Dowds, Jonathan; Aultman-Hall, Lisa; LaMondia, Jeffrey J.
• Is Part Of: Travel behaviour and society. Volume 19(2020)
• Page Start: 124
• Page End: 136
• Publication Date: 2020-04
• Subjects: Travel survey -- Long-distance travel -- Trip recall -- Overnight travel; Transportation -- Periodicals; Population geography -- Periodicals; 303.48305
• DOI: 10.1016/j.tbs.2019.12.004
• Languages: English
• ISSNs: 2214-367X

Abstract

Graphical abstract: Highlights: A one-year panel survey included two measures of overnight travel, a one-time self-assessed annual travel frequency estimate and record of overnight tours collected monthly. This study finds appreciable over and under estimation of self-reported personal, work, air and non-North American long-distance tours. We recommend targeted memorable trips be used as a relative measure of whether someone is a light or heavy long-distance traveler in routine travel surveys. We assess the aggregate sample-wide overnight trip rate across all people and tour types was accurate. Abstract: Due to the comparatively low frequency of long-distance and overnight travel, it can be challenging to measure using traditional travel surveys. In response to this dearth of data, several recent surveys have included self-reported frequency questions. However, the value and accuracy of these questions is unclear. This study leverages data from 628 panel members who completed a year-long, online survey in 2013–2014 by comparing their one-time self-assessment of typical overnight trip frequency to those reported in 12 subsequent monthly surveys. The self-assessed frequency of overnight tours, airplane tours and tours to non-North American destinations are consistent for only 68% and 70% of respondents for work and personal tours respectively. For most tour types, consistency is highest for low frequency (never and <1 per year) suggesting, unsurprisingly, that individuals are good at knowing they do not travel. Inconsistent estimators both over- and under-estimated trip-making suggest that not only is recall an issue but that prestige bias and the complexity and variability of long-distance travel are factors in recall as well. Few demographic factors were statistically significant in estimating whether participants were consistent reporters. Only 9% of participants consistently either over- or under-estimated both work and personal trips suggesting that there is simply a general inaccuracy in these survey questions for measuring long-distance travel. The aggregate trip rate across people and trip types was accurate suggesting this crude frequency measure may be acceptable for total frequency but not for understanding relative patterns or details in overnight long-distance travel.