On the Analysis of Repeated-Measurements Experiments

Publisher Website
Google Scholar
Abstract
An example is given involving repeated measurements on the same individuals over time. The assumptions for and techniques of the usual univariate analysis of variance procedure for such a set-up are given. The assumption of equal variances and covariances, the so-called symmetry assumption, is tested and it is concluded that this assumption is not fulfilled for these data. The multivariate procedures are then given for the situations where a valid univariate analysis is not justified. It is noted that asymptotically, the univariate and multivariate tests are identical. The conclusions are that for this example, essentially the same inferences are made from the univariate and multivariate analyses.

This publication has 1 reference indexed in Scilit: