Conjoint Analysis: An Application in Eliciting Patients’ Preferences

Conjoint analysis is a technique for establishing the relative importance of different attributes in the provision of a good or a service. In this study conjoint analysis was applied to characterize diabetic patients’ preferences for information during doctor-patient interactions. Patients’ utility function was further developed based on the random utility model that would account for inconsistencies that arises in patients’ choice behaviors. The unobserved portions of the utility function were specified as a combination of an IID (Independently Identically Distributed) distribution and another general distribution allows the model to be specified as mixed logit. The mixed logit approach provides an efficient estimate of correlation of the unobserved portions of patients’ utility function due to repeated choices made by the same respondent. Results from the analysis can be interpreted in terms of marginal rate of substitution (MRS) between attributes. Socio-economic characteristics of the patient were introduced into the model in the form of interaction terms explained how preferences varied across patients.