Estimating temporal trends in preferences measured by graded paired comparisons

A sizable marketing research literature on preference measurement, prediction, and explanation has thus far been more concerned with point estimates of preference than with consistent changes over time. Such temporal trends can be estimated using data collected from graded paired comparisons. An appropriate technique for deriving individual and group trend parameters is first described and then illustrated using data from an experiment dealing with aesthetic preferences. Suggestions are offered concerning its advantages and limitations for applications in marketing.