The analysis of variance arises directly from the randomization theory
of analysing experimental data. The Normal linear model analysis is
usually used as an approximation to the exact randomisation
analysis. If the responses are counts or proportions a generalized
linear model (GLM) is often used instead. It will be shown that GLMs
do not in general arise as natural approximations to the randomization
model. Instead the natural approximation is a generalized linear mixed
model (GLMM). Examples will be given to illustrate that it can make an
important difference in practice.


When: Nov 14, at 3-4 pm
Where: Room B705, Department of Statistics, Stockholm University


All welcome