A weekend shopping activity participation model dependent on weekday shopping behavior

摘要

The pattern of shopping activities in daily life varies across individuals; some may make go shopping every day, while others may do so only once at weekends, by adopting a once-a-week shopping schedule. It is believed that shopping behavior on one day affects the generation of shopping activities on other days. Focusing on weekend shopping activities, this paper analyzed the characteristics of individual shopping behavior using 1-week activity diary data from Utsu-nomiya, Japan; it proposed a simultaneous-equations (SE) model integrating a weekday shopping frequency model of Ordered Probit (OP) type with a weekend shopping participation model of Binary Probit (BP) type. The SE model, based on utility maximizing principles, was developed by assuming that the error terms of the latent preference function in the OP model and the utility function in BP model follow a bivariate normal distribution, and also by incorporating the effect of weekday shopping frequency in the model. The validity of the SE model was statistically tested using the activity data, and the effects of weekday shopping frequency on weekend shopping activity were discussed based on the model. As a result, it was shown that personal attributes do not affect weekend shopping patterns directly, but rather indirectly through the weekday shopping behavior. Besides, the effects are greater on Saturday than on Sunday.