Please use this identifier to cite or link to this item:
|Title:||The three legs of Dirichlet tables|
|Citation:||Proceedings of the Australian & New Zealand Marketing Academy Conference, 1-3 December, 2003, Adelaide, South Australia:pp.2553-2559|
|Publisher:||Australian and New Zealand Marketing Academy|
|Conference Name:||Australian & New Zealand Marketing Academy Conference (2003 : Adelaide, South Australia)|
|Carl Driesener, Cam Rungie, Cullen Habel and Jason Allsopp|
|Abstract:||The Dirichlet model is a widely applied model of purchase incidence and brand choice. Underlying the model, and indeed the many brand performance measures that it produces, are the A, K and S parameters. These parameters are usually estimated from panel data and then used to provide theoretical estimates for the previously observed values. Understanding the patterns in these parameters over different data sets provides greater understanding of the model. Also it would be useful to be able to estimate the parameters without using panel data so that the model may be used in categories where these data are not available. This paper details a number of parameter sets resulting from fitting the Dirichlet model to a various categories. The purpose of this is to begin the task of looking for patterns within these parameters that would allow a marketer to understand the model more and to estimate possible values of the parameters in the absence of panel data.|
|Appears in Collections:||Business School publications|
Files in This Item:
There are no files associated with this item.
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.