Frequency choice model for non-grocery shopping trips, as an aspect of multidimensional travel behavior

Abstract

Trip generation model have generally received less attention than other aspects of travel decision making. This paper presents some explorations into structure of trip decisions for non-grocery shopping and a possibility of applying choice model for trip decisions. Firstly, we assume the nested-choice-structure of shopping behavior, which consists of three dimensional decisions with frequency choice at top level. Secondly, we examine two choice models to describe frequency choice; one is 'Ordered Logit Model' by Sheffi (1979), another is 'Repeated Logit Model' based upon binomial distribution. Using data of monthly shopping trips on holiday, we present the validity of this nested structure, and the effectiveness of two choice models in this structure, comparing with general linear model. And we propose 'Partially Constrained O.L. Model' that is a more effective version of 'O.L. Model'.