Introduction to Conjoint Analysis Conjoint analysis studies have become increasingly popular as a method of determining the ideal or most preferred configuration for a new product or service. Unfortunately, the logistics involved in most conjoint analysis studies are such that effectively gathering the data required for a valid conjoint study requires that at some point in the research process the respondent has the opportunity to directly view the different options that are being evaluated. In practical terms, this has implied that most conjoint studies are conducted in a face-to-face interview situation such as mall intercepts through mail surveys, or by using some variation on a multifaceted approach, such as a phone-mail-phone survey. In turn, this requirement for direct contact results in either increased costs in comparison to telephone surveys, or decreased external validity by virtue of the selection bias that characterizes samples gathered via mall intercepts or mail surveys. This manuscript reports on the development of procedures designed to facilitate collection of the data required for a conjoint via a single telephone survey. The approach we will describe is not a panacea for addressing all conjoint research questions; indeed, there are some significant limitations in the method, which we will describe. Finally, we will also present the results of some preliminary tests of the model that compare the results obtained in comparison to some classic conjoint data collection procedures. Throughout this manuscript, data analysis and data manipulation examples will be presented within the context of conducting a conjoint analysis using the SAS® system for all analytical procedures. |