Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
5052786 | Economic Analysis and Policy | 2015 | 19 Pages |
Abstract
The order of a series of choice tasks presented to respondents in a discrete choice experiment (DCE) could affect the choice outcomes. This study explores the ordering effects in a DCE surveying preferences for improvements in cyclone warning services in Vietnam. Respondents' choices are analysed non-parametrically and parametrically to investigate the ordering effects in their preferences. Across the sequence of six choice questions, the stated demand of respondents is statistically significantly different at the first position from all other positions. Based on a parametric analysis using mixed logit models, we also find that the willingness-to-pay for a number of improvement programs estimated at the first position is relatively larger when compared with the other positions. The findings indicate that although DCEs can provide additional information on respondents' preferences when compared with survey methods using a single valuation question, the trade-off for more information is the ordering effects over a sequence of repeated questions.
Related Topics
Social Sciences and Humanities
Economics, Econometrics and Finance
Economics and Econometrics
Authors
Thanh Cong Nguyen, Jackie Robinson, Shinji Kaneko, Nguyen The Chinh,