ABSTRACT
Best-worst scaling (BWS), or maximum difference scaling, is a survey method for eliciting individuals’ relative importance of items (Finn and Louviere, 1992). BWS was developed to expand the capability of discrete choice experiments (DCEs) (Flynn et al., 2010). Although BWS is categorized into three types based on the format of the choice sets (Flynn et al., 2010; Marley, 2010), this chapter only addresses one of these in detail, namely, object case BWS. The other two types, profile case BWS and multiprofile case BWS, are briefly outlined in the appendix to this chapter. (Hereafter, BWS refers to object case BWS.)
BWS requires respondents to select two items from a choice set containing three or more items. The choices reflect the extremes of a certain standard, such as “best” and “worst,” or “most important” and “least important.” The BWS users can decide what standards should be used. Items can also be selected relatively flexibly. In DCE studies, goods/services are treated as alternatives and the characteristics of the goods/services are set as attributes of the alternatives (see Chapter 3). Although the characteristics of the alternatives are set as items in BWS, the respondents’ views, experiences, and/or benefits related to the goods/services are also frequently set as items. For example,
about a guide” or “Tasted the wine previously,” as well as characteristics of the wine such as grape variety and origin. Cohen and Neira (2004) applied BWS to measure the relative importance consumers placed on 13 benefits of drinking coffee, such as “It makes me feel relaxed” and “It helps me keep warm.” Lusk and Briggeman (2009) used 11 food values, including price, taste, safety, and environmental impact as items in their BWS questions.
