Date:         Tue, 6 Sep 2005 09:59:11 -0400
Reply-To:     qq.consulting@sympatico.ca
Sender:       "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From:         Susan Elgie <qq.consulting@sympatico.ca>
Subject:      Re: Post-hoc Test for Kruskal-Wallis
Comments: To: "Frank J. Gallo" <fjgallo@verizon.net>
Content-Type: text/plain; charset=iso-8859-1

Hi Frank,
I think most people use a series of Mann-Whitney tests in this situation.  If you do this, you will want to set a more conservative level of alpha, using perhaps the Bonferroni correction, to evaluate the results.  To obtain the Bonferroni correction, divide the alpha you are using by the number of tests you are doing, and use the result to decide if your results are significant.

Best,
Susan
__________________________
Susan Elgie
QQ Consulting
Toronto, Canada
416-531-9548
qq.consulting@sympatico.ca
>
> From: "Frank J. Gallo" <fjgallo@verizon.net>
> Date: 2005/09/06 Tue AM 09:40:14 EST
> To: SPSSX-L@LISTSERV.UGA.EDU
> Subject: Post-hoc Test for Kruskal-Wallis
>
> Hi All,
>
>
>
> Can anyone suggest a post-hoc comparison test, using SPSS, as a follow-up to
> the Kruskall-Wallis Test? I am looking for a pairwise ranking post-hoc test
> that controls the significance level for the entire family of comparisons.
> Does SPSS offer this feature? I am using version 10.1.
>
>
>
> Thanks,
>
> Frank
>