Regression Modeling of Ordinal Data with Nonzero Baselines (1996)

Summary:

This paper develops a regression model for ordinal data with non-zero control response proba­bilities. The model is especially useful in dose-response studies where the spontaneous or natural response rate is nonnegligible and the dosage is logarithmic. The model generalizes Abbott's for­mula, which has been commonly used to model binary data with non-zero background observations. We describe a biologically plausible latent structure and develop an EM algorithm for fitting the model. The EM algorithm can be implemented using standard software for ordinal regression. Analysis of historical data on the severity of virus-induced deformities in Chicken Embryos illus­trates the methodology. 

Keywords:

Nonzero baseline model, Ordinal data, latent structure, EM-algorithm 

Author: 
Minge XieDouglas G. Simpson
Publication Date: 
Friday, March 1, 1996
File Attachment: 
PDF icon tr43.pdf
Report Number: 
43