Increasing the transparency of stated choice studies for policy analysis: Designing experiments to produce raw response graphs

We believe a lack of transparency undermines both the credibility of, and interest in, stated choice studies among policy makers. Unlike articles reporting the results of contingent valuation studies, papers in the stated choice literature rarely present simple tabulations of raw response data (that is, a table or graph showing the percentage of respondents agreeing to purchase a good or service, or vote for a proposed management plan as a function of price). We describe an approach for adding “policy” tasks to a standard orthogonal-in-attribute-levels research design that allows the researcher to plot raw responses as a function of changes in only one characteristic of the offered good or service. We demonstrate this approach using data from a stated choice study of private demand for cholera and typhoid vaccines in Kolkata, India, carried out in the summer of 2004. © 2006 by the Association for Public Policy Analysis and Management.