The Internal Validity of the Trade‐Off Method of Conjoint Analysis

ABSTRACT The discovery of consumers' preferences for multiattribute products/services has recently become possible with the development of conjoint measurement techniques. Conjoint analysis uses an individual's preference ranking or rating of some deliberately manipulated constructs of a concept to determine his/her numerically valued preferences for levels of those attributes that comprise the concept. Generally speaking, there are two methods of data collection for conjoint analysis: the full profile and the trade‐off methods. This paper calls into question the internal validity of the preferences obtained via the trade‐off method. An empirical analysis, which is based on the subjects' preferences for selecting a hospital, clearly shows that the estimates derived from the trade‐off method may not reflect consumers' true preferences.