PROGRESS IN GEOGRAPHY ›› 2010, Vol. 29 ›› Issue (12): 1470-1478.doi: 10.11820/dlkxjz.2010.12.002

• Original Articles • Previous Articles     Next Articles

A Review on the Models in Research of Consumer Behavior in Commercial Space

ZHU Wei, WANG De   

  1. Tongji University, Shanghai 200092, China
  • Received:2010-06-01 Revised:2010-07-01 Online:2010-12-25 Published:2010-12-25


This paper reviews the main models used in the research of consumer behavior in commercial space. It takes a historical perspective, divides the development of the models into the stages of aggregate models and individual models, and classifies the model application into macro, meso, and micro scales. For the aggregate models, the paper firstly introduces gravity models based on spatial interaction theory, including the basic model form, constrained forms, and the competing destination model. The second part for the aggregate models section introduces Markov Chain models for describing dynamic consumer behavior, with the emphasis on the development and applications from static transition probabilities to varying transition probabilities. The individual model section introduces discrete choice models based on random utility theory, with the emphasis on the widely applied multinomial logit and nested logit models. This is followed by an introduction to multi-agent technology as a simulation tool. The review includes the fundamentals of underlying theories of the models, related literatures and model features. It is considered that the fitness between the model and the nature of the research is important for model selection. Aggregate models have the advantage of grasping the overall regularities, but are limited in exploring highly heterogeneous behavior. The advantage of individual models is the flexibility to represent heterogeneous behavior, while the idea of bottom-up simulation to form aggregate behavior requires deeper understandings of inter-individual interactions.

Key words: commercial space, consumer behavior, model, review