Format

Send to

Choose Destination
PLoS One. 2017 Aug 14;12(8):e0178826. doi: 10.1371/journal.pone.0178826. eCollection 2017.

Do labeled versus unlabeled treatments of alternatives' names influence stated choice outputs? Results from a mode choice study.

Author information

1
Department of Industrial Engineering, Tsinghua University, Beijing 100084, China.
2
Research & Advanced Engineering, Ford Motor Company, 2101 Village Road MD-2149, Dearborn, MI 48121, United States of America.

Abstract

Discrete choice experiments have been widely applied to elicit behavioral preferences in the literature. In many of these experiments, the alternatives are named alternatives, meaning that they are naturally associated with specific names. For example, in a mode choice study, the alternatives can be associated with names such as car, taxi, bus, and subway. A fundamental issue that arises in stated choice experiments is whether to treat the alternatives' names as labels (that is, labeled treatment), or as attributes (that is, unlabeled treatment) in the design as well as the presentation phases of the choice sets. In this research, we investigate the impact of labeled versus unlabeled treatments of alternatives' names on the outcome of stated choice experiments, a question that has not been thoroughly investigated in the literature. Using results from a mode choice study, we find that the labeled or the unlabeled treatment of alternatives' names in either the design or the presentation phase of the choice experiment does not statistically affect the estimates of the coefficient parameters. We then proceed to measure the influence toward the willingness-to-pay (WTP) estimates. By using a random-effects model to relate the conditional WTP estimates to the socioeconomic characteristics of the individuals and the labeled versus unlabeled treatments of alternatives' names, we find that: a) Given the treatment of alternatives' names in the presentation phase, the treatment of alternatives' names in the design phase does not statistically affect the estimates of the WTP measures; and b) Given the treatment of alternatives' names in the design phase, the labeled treatment of alternatives' names in the presentation phase causes the corresponding WTP estimates to be slightly higher.

PMID:
28806764
PMCID:
PMC5555680
DOI:
10.1371/journal.pone.0178826
[Indexed for MEDLINE]
Free PMC Article

Supplemental Content

Full text links

Icon for Public Library of Science Icon for PubMed Central
Loading ...
Support Center