An Empirical Analysis for a Case-based Decision to Watch Japanese TV dramas

This article empirically analyzes consumer behavior of watching TV dramas via case-based decision theory. This theory models consumer decision under uncertainty in product quality, based on subjective evaluations of previous purchases for similar goods. Our empirical analysis is concerned with getsuku , the Japanese TV dramas broadcasted at 9pm Monday by the Fuji Television Network, whose quality necessarily has uncertainty because no program is completely equivalent to any other. Owing to the regularity of the schedule and the long-sustained popularity of the program that enables us to be able to collect consumer data easily, we conduct a web survey of individual audiences on subjective evaluations of previously watched dramas. Our empirical analysis demonstrates better performance by the case-based models regarding both statistical model selection and one-step-ahead prediction than traditional utility-based models. We also reveal that the good performance of the case-based model in our analysis depends on the availability of individual subjective evaluations and that it is difficult to replace the individual-specific information by demographic information and aggregate data.