Element category importance in conjoint analysis: Evidence for segment differences

Abstract

Conjoint measurement is a popular research tool that uses experimentally designed combinations of concept elements (phrases, pictures), obtains consumer reactions to these combinations, and then employs statistical modelling (eg regression analysis) to understand the drivers of concept acceptance. The elements used in conjoint analysis can be classified into categories or buckets of like-minded statements, such as benefits, brands, heritages, mode of action, etc. This paper introduces the concept of the relative importance index (RII), for categories comprising different numbers of elements (verbal or visual descriptors of a concept). This index is the proportion of the total sum of squares of individual utilities attributable to each category. The RII value can be further normalised by dividing by the number of elements within the category to estimate the importance of each category normalised for the number of concept elements. Categories exhibit different importance. This study presents empirical analyses of 12 studies, each with more than ten categories, and more than 100 concept elements. Segmentation of respondents based upon the pattern of element impacts often shows dramatic segment differences in this RII index, suggesting different product development and communication strategies.