Analysis of Repeated Measurements of Animals

Abstract

An example of non-random repeated information of lactating rats is analyzed to clarify statistical procedures that avoid misleading inferences from such data. Recognition of split-plot structure and examination of heterogeneous correlation of data are emphasized. Methods are outlined for testing the variance-covariance structure, and multivariate procedures are discussed for testing hypotheses about effects of repeated factors (e.g., time) and for comparing profiles of responses (over time) for two treatment groups.