- noun a research method aimed at discovering the best combination of features for a product or service, e.g. price and size
- (written as Conjoint Analysis)A range of techniques used to elicit individuals' preferences and willingness to pay for health states, health services or particular features of services. Within this range are such methods as card sorts, choice modelling, discrete choice analysis, hierarchical choice, pairwise comparisons, stated preference analysis and trade-off matrices. A variety of hypothetical questions or vignettes is formulated, each varying in its mix of attributes (which have to be considered 'conjointly' by the subject) and subjects are asked to give discrete answer (yes/no or 'I prefer option A', etc.). Regression analysis is used to estimate the relative strength of preferences for attributes. In a study of the quality of a fertility service, for example, the attributes to do with quality might include attitudes of staff to the patient (uncaring/unsympathetic; caring/sympathetic), continuity of contact with staff (see same staff; see many different staff), time on waiting list for first IVF attempts (1, 3, 6, 18, 36 months), cost to patient per attempt ($0, $750, $1500, $2500, $3000), chances of taking home a baby (5, 10, 15, 25, 35 per cent), follow up support (yes; no). The resultant scores are usually treated as utilities. Willingness to pay estimates are also sometimes made. The inventors were Luce and Tukey (1964).