post hoc Kruskal-Wallis

[janet]

Hi,
Can SPSS perform post hoc test with Kruskal-Wallis (non-parametric)? Or do you have to run a separate series of Mann-Whitney tests and make Bonferroni corrections?

Thanks,
Janet

[Bruce Weaver]

Hello Janet.  Pardon my impertinence, but why are you using the KW test?  Please tell us more about your data and the question you are addressing.

I ask because quite frequently people use it when they really want to use ANOVA, but fear it will not be robust given the circumstances.  But rank-based tests like KW and Mann-Whitney U also have assumptions, and in some situations, they are actually less robust than their parametric counterparts to violations of their assumtions.  So I'm wondering if the answer is that you don't really need to be using KW.  

HTH.

janet wrote

Hi,
Can SPSS perform post hoc test with Kruskal-Wallis (non-parametric)? Or do you have to run a separate series of Mann-Whitney tests and make Bonferroni corrections?

Thanks,
Janet

[Ryan]

http://www-01.ibm.com/support/docview.wss?uid=swg21477370

Ryan

On Thu, Feb 6, 2014 at 3:09 PM, janet <[hidden email]> wrote:

Hi,
Can SPSS perform post hoc test with Kruskal-Wallis (non-parametric)? Or do
you have to run a separate series of Mann-Whitney tests and make Bonferroni
corrections?

Thanks,
Janet



--
View this message in context: http://spssx-discussion.1045642.n5.nabble.com/post-hoc-Kruskal-Wallis-tp5724355.html
Sent from the SPSSX Discussion mailing list archive at Nabble.com.

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD