Log-linear models provide a natural analog to the classical least square regression for categorical frequency data. These models are also extended for the data having ordered category. In this article ordinal log-linear models and RC-association (log- bilinear) model are compared. Also presented an idea for the interpretation of the parameters on the basis of odds-ratio. These models were applied to a data to illustrate the computation and comparability.
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
Copyright (c) 1995 Array