Unfolding Conjoint Utilities
Ennis, D. M. (2020). IFPress, 23(2) 3-4.
Conjoint analysis is a popular method used in the study of items involving combinations of attributes at one or more levels which comprise decomposable stimuli. In contrast, unitary stimuli are items that cannot be readily decomposed into attributes experimentally (e.g., the components of a fine fragrance). The idea behind conjoint analysis is that there is a latent hedonic continuum and that there are part-worth utilities that correspond to parts of this continuum that can be estimated for each attribute level for groups of respondents or for individuals. Notwithstanding consistent violations of the assumptions underlying conjoint analysis, the method has proven to be of use in making marketing decisions. Conjoint analysis is limited to applications involving stimuli in which the variables and their levels are identified in advance. In this respect conjoint analysis has features in common with external preference mapping which also requires prior knowledge of the attributes of interest that may influence hedonics. The stimuli in a conjoint study are typically drawn from an orthogonal array and in this respect the method shares features in common with designed experiments involving factorial, central composite and fractional factorial designs.