Tuesday, January 03, 2006

st: RE: Nonparametric Multiple Comparison Testing

I have comments on various levels, but no solution that you will like.

What is best in such cases is to be clear that the appropriate procedures do exist -- on which I have no idea -- and to use the appropriate mix of -search-, -findit- and Google to see what there is in Stataland. Did you do this first before posting?

From one statistical-theological point of view, you are already in a grievous state of sin. By throwing both parametric and non-parametric procedures at a dataset, you are already guilty of data snooping. Wanting to apply a Bonferroni correction at one level, but regarding the application of multiple procedures as fair strategy, is from that point of view dubious methodology. I point this out without worrying much about it myself, as checking whether different procedures agree seems sensible to me, whatever the purists say.

Nick n.j.cox@durham.ac.uk

Chris Simpkins > Is it possible to perform nonparametric multiple comparison > testing in Stata? I am attempting to compare the > distributions of a continuous variable (AGE) between five > groups (group5) with unequal variances (based upon Bartlett's > test using the oneway command) and I don't see any option for > multiple comparison testing with the kwallis command. I > searched through the Statalist archives and Stata's help but > haven't been able to turn up any information on this. > > Here are the commands that I used: > > oneway AGE group5, bonferroni > > kwallis AGE, by(group5) >

* * For searches and help try: * http://www.stata.com/support/faqs/res/findit.html * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/


Links to this post:

Create a Link

<< Home

This page is powered by Blogger. Isn't yours?