A comparative study of four methods for analysing repeated measures data

Abstract

This paper compares four alternative statistical methods for the analysis of repeated measures data. The data set involves fatty acid composition of red blood cell membrane examined after acute MI and is characterized by a moderate sample size, unbalanced repeated measures and high within-individual correlations. Two methods are based on individual curve fitting and two on the random effects model. Results yielded by the four methods differed. The advantages and disadvantages of the methods are discussed. Assumptions required for the valid application of these methods are tested.