Abstract
This study empirically tests a dynamic choice model based upon a MULTILOGIT approach. The MULTILOGIT model offers tremendous forecasting potential to managers. The manager is able to examine if-and-when consumers will choose. In other words, the manager can generate the probability that an individual will never choose an alternative and also generate separate probabilities that the individual will choose one or two periods into the future.
The MULTILOGIT approach is utilized to predict attendance (choice) at home basketball games one and two weeks prior to actual choice. The parameter values associated with the independent variables provides some indication of the temporal dynamics of choice behavior. The influences on choice of attendance by girl/boy friend (DATE), scheduling constraints (TIME), a large crowd (CROWD), and parking constraints (PARK) change during each of the two weeks prior to actual choice.
Additional information
Notes on contributors
Michael D. Richard
Michael D. Richard is Assistant Professor of Marketing at Mississippi State University. Dr. Richard has published in such distinguished journals as the Journal of Services Marketing and the Journal of Consumer Marketing.
Arthur W. Allaway
Arthur W. Allaway is associate professof of Marketing at The University of Alabama. Dr. Allaway has published in such distinguished journals as the Journal of Marketing and the Journal of Retailing.
James A. Womack
James A. Womack is a graduate student at Auburn University. Mr. Womack has published in such distinguished journals as the Journal of Consumer Marketing.