Abstract
The paper outlines the usefulness of conjoint analysis for studying cultural and leisure preferences. The technique is advantageous because it permits one to make predictions about preferences, by using a factorial design, not only for choices that are explicity made but also for others only implied by the research design.
However, in analyzing the data available for this exercise, it became apparent that one must employ a minimum number of items for respondents to rank order. If there are two few, intercorrelations occur among items which can lead to misleading or invalid results.
KEYWORDS: