Repeated Measures Analyses

IV
Mengling Liu

Mengling Liu

New York University, New York, NY, USA

Search for more papers by this author
Yongzhao Shao

Yongzhao Shao

New York University, New York, NY, USA

Search for more papers by this author
First published: 15 September 2008

Abstract

Repeated measures data arise when multiple measurements of response variables and/or covariates are obtained in sequence from each experimental unit and occur frequently in many disciplines. The repeated measures design provides opportunities for investigators to examine the pattern of response variable under different measurement occasions, to understand the relationship between repeated outcomes and covariates, and to discern the dependence pattern among the repeated measurements. A special feature of repeated measures data is that measurements from the same unit are usually positively correlated; thus, valid and effective analysis of repeated measures must deal with this within-subject correlation. In this article, we review some of the analysis methods for repeated measures, which include univariate summary methods, multivariate analysis of variance, parametric random-effects methods, marginal generalized estimation equation methods, and modern semiparametric and nonparametric approaches. Some simple methods for data exploration and graphical display are demonstrated using real data examples. We also briefly discuss related computational tools and statistical packages.

The full text of this article hosted at iucr.org is unavailable due to technical difficulties.