The functions so far agree with an example problem from Conover 1999, but it would be good to use simulations to test 1) the randomization version more stringently, and 2) the multiple-comparison procedure. Ideally it would be good to look at effects of varying the number of blocks and treatments.
The functions so far agree with an example problem from Conover 1999, but it would be good to use simulations to test 1) the randomization version more stringently, and 2) the multiple-comparison procedure. Ideally it would be good to look at effects of varying the number of blocks and treatments.