Analysis of covariance by the SAS GLM procedure.

Analysis of covariance (ANCOVA) is an important statistical procedure for the analysis of biomedical data; the procedure is used for the comparison of means between two or more groups of subjects exposed to contrasting stimulus with adjustment for confounding bias. Although ANCOVA is available in all the major statistical packages (e.g. SAS, SPSSX, BMDP), not all the important statistical aspects of ANCOVA are explicitly considered by these packages. This article expands the analysis of a numeric example included in the SAS GLM procedure to cover several crucial statistical aspects relevant to ANCOVA, and to highlight the meaning of the statistical results obtained.