Abstract

In the context of a stated-binary-choice experiment, or Discrete Choice Experiment (DCE), straightliners are defined as survey participants whose responses are not related to the attributes of the two alternatives, and can therefore be perceived as choosing at random in every choice task that they face. Straightlining may be seen as an extreme form of attribute non-attendance which has been studied extensively in the choice modelling literature. We apply recently developed econometric methodology to the MAUCa Australian EORTC QLU-C10D DCE data in order: (1) to investigate the extent of straightlining response in this data source; (2) to estimate preference parameters in ways that adjust for the presence of straightliners; (3) to investigate ways in which straightlining depends on respondent characteristics. We find a prevalence of straightliners in the survey, and we allow for their presence in the estimation of the preference parameters that are of central interest. We verify that estimates are seriously biased if their presence is neglected. For each respondent in the survey, we obtain a posterior straightlining probability, which provides a simple means of identifying respondents who are most likely to be straightliners. We emphasise the usefulness of this as a justification for excluding identified straightliners from future surveys, leading to cost savings. Finally, we point out that the cross-country comparison of straightlining prevalence which is made possible by the MAUCa Consortium’s program of research, is likely to be of considerable interest to the choice modelling community in general. Following analysis of the Australian QLU-C10D DCE data, results will be shared with other MAUCa members, and the interest of the custodians of QLU-C10D DCE data from other countries will be sought, with the aim of conducting cross-country comparisons of the prevalence of straightliners.

Key words: Discrete choice experiment; Attribute non-attendance; straightlining.

About the presenter's visit

Peter Moffatt will be visiting the School of Economics on Wednesday 31 July 2019. While here he will be using room 520A Colin Clark building. If you would like to meet with him or have lunch or dinner with him please contact Alex Karakostas who will be his host while at The University of Queensland. Alex Karakostas can be contacted on a.karakostas@uq.edu.au

 

About Econometrics Colloquium Seminar Series

A seminar series designed specifically for econometricians to network and collaborate.

Subscribe to UQ Economics seminar updates

Venue

Level 6, Colin Clark building (#39)
The University of Queensland
St Lucia campus
Room: 
629