For repeated measures analysis, PROC MIXED is generally the recommended procedure. However, repeated measures analysis can also be done in PROC GLM using either a univariate, or a multivariate, approach. If you use PROC GLM for a repeated measures analysis and you observe differences in the results of these two approaches, the two most common causes are missing data and failure to define the correct univariate model with appropriate TEST or RANDOM statements. For more information, see "Repeated Measures Analysis of Variance" in the Details section of the GLM documentation. This includes information on organization of data, model specification, and hypothesis testing in repeated measures analyses. See also the example titled "Repeated Measures Analysis of Variance" in the GLM chapter. Additional examples and discussion of repeated measures analysis can be found in SAS for Linear Models, Fourth Edition.

Operating System and Release Information