An interactive computer program for randomization analysis of response curves with facilities for multiple comparisons.

An interactive Fortran program, MUCRA, is presented. The program can perform randomization analysis of a completely randomized or randomized-blocks design extended to growth and response curves. A single-step Scheffé-type procedure as well as the Peritz's closed step-down procedure have been implemented which control the familywise type I error-rate. In general, MUCRA is suitable as a computer tool for a distribution-free analysis of variance with repeated measures. The use of MUCRA is demonstrated by analyzing the effects oxprenolol and atenolol have on exercise heart rate. Oxprenolol is a non-selective beta-blocker with moderate intrinsic sympathomimetic activity (ISA), given by the Oros delivery system. Atenolol is a beta 1-selective blocker without ISA. A randomized placebo-controlled crossover design was used to compare the effects of the beta 1-blockers on heart rate during a progressive maximal exercise test on a bicycle ergometer. Application of the Scheffé-type procedure showed that the two drugs significantly (alpha = .05) reduce the heart rate during the exercise test at the three prechosen times (2, 5, and 24 hr) after intake. The reduction from atenolol is more pronounced than from oxprenolol Oros at 2 and 5 hr.