The GLM Procedure
-
Overview
-
Getting Started
-
Syntax
-
DetailsStatistical Assumptions for Using PROC GLMSpecification of EffectsUsing PROC GLM InteractivelyParameterization of PROC GLM ModelsHypothesis Testing in PROC GLMEffect Size Measures for F Tests in GLMAbsorptionSpecification of ESTIMATE ExpressionsComparing GroupsMultivariate Analysis of VarianceRepeated Measures Analysis of VarianceRandom-Effects AnalysisMissing ValuesComputational ResourcesComputational MethodOutput Data SetsDisplayed OutputODS Table NamesODS Graphics
-
ExamplesRandomized Complete Blocks with Means Comparisons and ContrastsRegression with Mileage DataUnbalanced ANOVA for Two-Way Design with InteractionAnalysis of CovarianceThree-Way Analysis of Variance with ContrastsMultivariate Analysis of VarianceRepeated Measures Analysis of VarianceMixed Model Analysis of Variance with the RANDOM StatementAnalyzing a Doubly Multivariate Repeated Measures DesignTesting for Equal Group VariancesAnalysis of a Screening Design
- References
As described previously, PROC GLM can be used for many different analyses and has many special features not available in other SAS procedures. However, for some types of analyses, other procedures are available. As discussed in the sections PROC GLM for Unbalanced ANOVA and PROC GLM for Quadratic Least Squares Regression, sometimes these other procedures are more efficient than PROC GLM. The following procedures perform some of the same analyses as PROC GLM:
- ANOVA
-
performs analysis of variance for balanced designs. The ANOVA procedure is generally more efficient than PROC GLM for these designs.
- MIXED
-
fits mixed linear models by incorporating covariance structures in the model fitting process. Its RANDOM and REPEATED statements are similar to those in PROC GLM but offer different functionalities.
- NESTED
-
performs analysis of variance and estimates variance components for nested random models. The NESTED procedure is generally more efficient than PROC GLM for these models.
- NPAR1WAY
-
performs nonparametric one-way analysis of rank scores. This can also be done using the RANK procedure and PROC GLM.
- REG
-
performs simple linear regression. The REG procedure allows several MODEL statements and gives additional regression diagnostics, especially for detection of collinearity.
- RSREG
-
performs quadratic response surface regression, and canonical and ridge analysis. The RSREG procedure is generally recommended for data from a response surface experiment.
- TTEST
-
compares the means of two groups of observations. Also, tests for equality of variances for the two groups are available. The TTEST procedure is usually more efficient than PROC GLM for this type of data.
- VARCOMP