Date: Wed, 14 Sep 2005 15:58:41 -0400
Sender: "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From: Christina Mehta <Christina.Mehta@cancer.org>
Subject: family-wise error rate in post-hoc tests
Content-Type: text/plain; charset="US-ASCII"
First, I would like to thank all of the contributors to this listserv for
their helpful advice in solving SPSS problems and patience in explaining
things to newcomers such as myself. You are all providing a very valuable
service and it is very much appreciated. Second, I have a question
regarding the Tamhane's T2 as a post-hoc test in multiple linear
regression analysis. Does SPSS automatically set the family-wise error
rate at alpha=0.05? My guess is that the family-wise error rate is set
at alpha= 0.05 for all of the post-hoc options that are listed but I am
having difficulty confirming that when I reviewed the SPSS documentation.
Does anyone know the answer and if so is there a reference? Thanks again
for all of your help.
C. Christina Bahn Mehta, M.S.P.H.
Behavioral Research Center
American Cancer Society
1599 Clifton Road, Northeast
Atlanta, Georgia 30329
Phone: (404) 329-7756
Fax: (404) 929-6832