iHEA Webinar - June 1, 2020: An Introduction to the Construction of Discrete Choice Experiments

Discrete choice experiments (DCEs) are a popular stated preference tool in health-related areas. In a typical choice experiment respondents are presented with a scenario in which a choice must be made, and a set of possible options (a choice set). The options are described by a number of attributes, each of which takes a specific level in each option. Respondents are assumed to make utility-maximising decisions, the utility is assumed to have a deterministic component which is a function of the attributes and levels, and the goal of the DCE is to model how the attribute levels affect respondent choices.
The design problem for DCEs is simply stated: which options should be presented together in a choice set, and which choice sets should be presented together to make a choice experiment.
After discussing the basic features of a DCE, this talk will outline how a set of possible DCEs can be compared and ordered from best to worst. We will then go on to discuss how to construct generator-developed designs, a robust approach to obtaining DCEs. Various algorithmic constructions for DCEs have been proposed and these will be briefly outlined.