A SAS/IML program for implementing the modified Brown–Forsythe procedure in repeated measures designs

In this paper, we present a computer program written in version 9.1 of SAS’ interactive matrix language in order to implement a new approach for analyzing repeated measures data. Previous studies reported that the new procedure is as powerful as conventional solutions and generally more robust (i.e., insensitive) to violations of assumptions that underlie conventional solutions. The program also included a step-wise procedure based on the Bonferroni inequality to test comparisons among the repeated measurements. Both univariate and multivariate repeated measures data can be analyzed. Finally, the application of the SAS/IML program is illustrated with a numeric example.