A method for construction of mixed models for categorical responses is described and its use is exemplified for three experimental designs: the repeated–measurements design, the change–over design and a design used for measurement of inter–observer agreement. The method may be viewed as an extension of the standard log–linear model methodology.

Log-linear mixed models for categorical data

. Estimation of success rates in the two–period change–over clinical trial with binary responses is considered under a mixed model for the design.

V. Fidler

Papers

Change–over trials with binary data: estimation