2003 Volume 4 Pages 27-34
It is not easy for experimental biologists to select an appropriate statistical method, especially in cases in which a multiple comparison method is necessary. Since various types of multiple comparison procedures have been proposed, experimentalists have had to face this difficulty. We propose a new multiple comparison procedure which has a wide range of applicability and uses a resampling technique that is currently popular in biostatistical literature. This new procedure enables us to solve multiple comparison problems more easily than conventional methods such as the Dunnett method. The power of tests of the new procedure was compared with some conventional multiple comparison methods, and satisfactory results were obtained.