Solutions to some problems in the implementation of conjoint analysis

Abstract

Methodological problems encountered in implementing conjoint analysis include (1) the impractically large set of multiattribute choice alternatives created by the factorial combination of more than a few attributes, (2) the hypothetical nature of the alternatives in the choice set, and (3) the assumption that each individual’s preferences can be described by the same composition rule. The techniques of tailoring, belief matching, and axiom testing are suggested as solutions to these problems, and their use is demonstrated in a conjoint analysis study of individuals’ contraceptive preferences. It is noted that tailoring and belief matching can also be used as methodological enhancements in functional measurement studies.

Preparation of this paper was supported in part by Grant HD-14403 from the National Institute of Child Health and Human Development. The assistance and suggestions of Sarah Gaskill, Alice Healy, Lou McClelland, Michael O’Reilly, and several reviewers are greatly appreciated.