Edited by Jeff Bennett
Chapter 13: Dealing with Scale and Scope Issues in Stated Preference Experiments
1 John Rolfe and Xuehong Wang 1 INTRODUCTION The application of stated preference techniques such as contingent valuation (CV) and choice modelling (CM) involves a constructed market setting where respondents are asked to identify how they would choose between a single trade-off (CV) or for several repeated trade-offs (CM). The analyst designing these applications typically has some discretion over the size and complexity of trade-offs that will be offered in the choice alternatives (Carson et al., 1994). In most cases the analyst balances the desire to make the choice task realistic against the desire to minimize choice complexity. Important decisions thus need to be made about the scope of the trade-off that will be presented to respondents, the actual quantities involved and the way in which those trade-offs are framed. As stated preference experiments can be replicated with the trade-offs constructed in different ways, it is relatively easy to test how changing the trade-offs can influence the choices made by respondents, and hence the subsequent values that are assigned. There was early evidence with the CV method that respondent choices were insensitive to changing the dimensions of the trade-off on offer, contrary to expectations from economic theory. This insensitivity to smaller and larger trade-offs became known as part-whole bias (Mitchell and Carson, 1989) or scope insensitivity (Arrow et al., 1993), and has generated heated and divergent views about the usefulness of the CV method (Schulze et al., 1998). An example of scope insensitivity was reported by Desvousges et al. (1993)...
You are not authenticated to view the full text of this chapter or article.