Date:         Fri, 30 Oct 1998 12:30:58 -0800
Reply-To:     Potato <potato@SEE-SIG.CUHK.EDU.HK>
Sender:       "SPSSX(r) Discussion" <SPSSX-L@UGA.CC.UGA.EDU>
From:         Potato <potato@SEE-SIG.CUHK.EDU.HK>
Organization: The Chinese University of Hong Kong
Subject:      Re: Post hoc tests for repeated measures
Content-Type: text/plain; charset=us-ascii

Stephen P. Baker wrote:
>
> The post-hoc tests all assume independent errors, and since repeated measures
>  are
> not independent, these tests are generally invalid in that setting.
 
Are there any literature related to this issue?  I searched the text at hand,
and
surprisingly found very little discussion on doing post-hoc tests for repeated
measures.  What confused me more is that Statistica 4.5 allows us to apply many
post
hoc tests on the cell means of any designs, including within-subject designs
and
mixed-designs.  Although using adjusted paired t-tests seem to be the "safest"
approach, the power might be way too low and doesn't use all information at
hand.  I'd
like to know more on this topic.  Many thanks in advance.
 
--
===================================================================
Shu-Fai Cheung (Potato)    email: potato@cuhk.edu.hk
Psychology Department      The Chinese University of Hong Kong
===================================================================