Modeling consumer choices of health plans: a comparison of two techniques.

This paper has two objectives. First, we will describe how conjoint measurement, a multivariate marketing research technique, can be applied in health care marketing. Second, we will compare the validity of results from two conjoint measurement techniques--the full profile approach and the tradeoff approach. A convenience sample of 97 university students was used in the study. Fifty-two students supplied data by using the full profile approach. Each respondent provided a complete rank order of 26 profile cards which included the following ambulatory health service attributes: charge for routine visit, travel time, office hours, length of time needed to make an appointment, waiting time in physician's office, practice arrangement/freedom of physician choice, parking arrangements and type of hospital. A fractional factorial design was used to determine different attribute levels (e.g. charge for routine office visit could be set at $10, $20 or $30) for each card. Forty-five students performed ranking tasks for the trade-off approach to conjoint measurement. These respondents ranked 28 grids which represent all combinations of factors taken two at a time. From the data collected in the ranking tasks, utilities or part-worth values for each level of each attribute were estimated by using dummy variable regression. Relative importance of ambulatory service attributes was inferred from the range of utility values of the attributes. Three measures of validity were assessed--adherence of estimated utility scores to monotonic assumptions, plausability of importance rankings and comparative validity. The results from the full-profile approach satisfied all three criteria. In contrast, the tradeoff approach results satisfied the first two criteria, but its comparative validity was only marginal. Valid conjoint data can be used for: simulations of market responses to different health services configurations; market segmentation studies; and development of promotional efforts.