|Date: ||Mon, 21 Aug 2000 04:04:26 -0700|
|Reply-To: ||Philip Whittall <philip.whittallNOphSPAM@UNILEVER.COM.INVALID>|
|Sender: ||"SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>|
|From: ||Philip Whittall <philip.whittallNOphSPAM@UNILEVER.COM.INVALID>|
|Organization: ||http://www.remarq.com: The World's Usenet/Discussions Start Here|
|Subject: ||Re: proc ttest p-values interpretation|
I think the important point is that the
p-values for the t-test are valid only if
the variances of the 2 populations are equal.
You should check the p-values under this
assumption and the p-values under the assumption
that they are different - if you're lucky they'll
be nearly equal. If the test for equality of the
variances is significant then take the p-value
for unequal variances.
* Sent from AltaVista http://www.altavista.com Where you can also find related Web Pages, Images, Audios, Videos, News, and Shopping. Smart is Beautiful