Pairwise comparisons for parallel profile models with mixed effects

Abstract

There are several linear and nonlinear models for analyzing repeated measurements.
The mean response for an individual depends on the regression parameters speciåc
to that individual. One of the simple form is the sum of vectors of åxed parameters and
random eãects. When the models with mixed eãects for several groups are parallel, pairwise
comparisons of level diãerences are considered. For the comparisons, simultaneous
conådence intervals are given approximately.