I'd probably go for the REGWQ, assuming you want to test all pairwise comparisons (that is, use a shotgun rather than a rifle). REGWQ will hold familwise alpha at its nominal level but give you more power than Tukey.
Karl W. -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Sent: Sunday, March 30, 2003 3:51 PM To: [EMAIL PROTECTED] Subject: Post Hoc Test I am comparing 4 class means. Each having different sizes. I have ran a F-test and there is a significant difference between the means. I want to run a Post Hoc test to find out which ones. Do I use Tukey? Do you need equal class sizes? Scheffe? . . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . ================================================================= . . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
