01662716

Component

Transport studies have relied heavily on surveys to collect information regarding mode and life satisfaction to derive better public policies.Although researchers in other fields have subjected survey methods to meticulous critique, identified many potential biases and ways to alleviate them, such critiques are seldom undertaken within the transport field. One such bias that the authors focus on in this study is that of question order effects, whereby variation between responses can arise depending on what order respondents answer the questions and there has been no study, to the authors' knowledge, that examines how travel survey data can vary due to these effects. This study uses a transport and livability survey undertaken in Calgary, Canada to compare responses to satisfaction and convenience questions between two groups of respondents, divided at random from the beginning of the survey, who answer questions in differing orders.Using logit regression models and controlling for socio demographics and modes of commute (walk, bike, car, bus and light rail), the authors test three degrees of question order alterations. The authors' model shows that larger question order alterations produce greater, statistically significant differences in the likelihoods of respondents to report high life satisfaction scores, respondents who reported life satisfaction scores at the end of the survey were 78% more likely to report high life satisfaction score than those who provided this information at the beginning of the survey, without having yet answered any other questions. Additionally, the authors found that convenience of commute scores for the present season were 50% lower amongst respondents who reported these prior to those of a hypothetical season in future. Finally the authors found that these question order effects appear to impact all mode users to a comparable extent, while minor changes in satisfaction with a commute characteristics question order did not impact the answers. Findings from this study provide a better understanding of how survey question orders can influence responses across multiple questions, in this case regarding trip convenience, trip and life satisfaction, which can be of interest to transport researchers and professionals concerned with these topics among different.

This paper was sponsored by TRB committee ABJ40 Standing Committee on Travel Survey Methods.

18-00722

English

16p

2018

Transportation Research Board 97th Annual Meeting

Digital/other

Figures; References

Commuters; Methodology; Mode choice; Questionnaires; Surveys; Travel behavior

Planning and Forecasting; Transportation (General)

Transportation Research Board Annual Meeting 2018 Paper #18-00722

TRIS, TRB, ATRI

Jan 8 2018 10:11AM