Using partial profile choice experiments to handle large numbers of attributes
Keith Chrzan
Maritz Research
Background
Conjoint analysis has proven to be a versatile tool for applied researchers (Green & Srinivasan 1978; Wittink & Cattin 1989). Exposing respondents to hypothetical, experimentally designed products, researchers can quantify the value, or utility, that customers derive from the attribute levels that comprise the products; these utilities enable researchers to predict consumer response to new combinations of attribute levels. Conjoint tasks are ‘full profile’ (FP) when the hypothetical products, or ‘profiles’, specify levels for every attribute of interest. Conjoint stimuli eliciting choice responses are...