CASEMAP: Computer-Assisted Self-Explication of Mutl-Attributed Preferences

Traditional conjoint analysis techniques do not take account of the fact that consumers often rule out from consideration products that have totally unacceptable attribute levels. This paper proposes a data collection and analysis technique that realistically models consumers as following a two-state decision process. In the first state, consumers rule out product options that have one or more totally unacceptable attribute levels. In the second state, product options that remain are traded-off against one another. The second state decision follows traditional preference models in that consumers chooose on the basis of a product’s overall utility, expressed as the sum of the utilities for the multiple attributes.To minimize the information overload on the respondents, the proposed method uses a computer-assisted, self-explicated approach to preference measurement. Compared to earlier self-explicated approaches, a conceptually more appropriate definition of attribute importance is used. The data are collected in a single computer-assisted telephone interview which improves the quality of the sample, minimizes interviewer and respondent errors, and permits a greater number of attributes to be included in the study compared to traditional conjiont analysis. The telephone interview also has the practical advantage that it is less expensive and reduces the time to collect and tabulate the data. In a practical application of the technique, the average price premium that consumers were willing to pay for a leading brand compared to the competitive brand was close to that obtained by classical conjoint analysis. In an empirical study in the context of predicting MBAs’ choices among job offers, the new approach yielded a slightly higher predictive validity than traditional conjoint analysis.