Message-ID: <b8668b8b-1b66-f6af-7868-8f34c1e02eff@math.uni-giessen.de>
Date: 2023-08-30T07:46:26Z
From: Gerrit Eichner
Subject: Request for advice: Multivariate Wilcoxon Test
In-Reply-To: <CAEatWUoXXqLwb3Wi3C=DvN-FcEXs8T3cOLmyMBu4TZgV9nBDMA@mail.gmail.com>
Hello, Faith,
this is a rather not R-related, but statistical question for which you
should probably seek (professional?) advice. Nevertheless, as unspecific
as you formulated it, you may want to look at Friedman's test which
allows to use a (second) blocking variable, or at a regression approach.
If you are really kean on a nonparametric _multivariate_ method the
references below *may* be helpful. The first is really mathematical, the
second more applied.
?Hth? --? Gerrit
@misc{puri1971nonparametric,
address = {New York {[u.a.]},
author = {Puri, Madan Lal},
isbn = {0471702404},
series = {Wiley series in probability and mathematical statistics},
title = {Nonparametric methods in multivariate analysis},
year = 1971
}
@misc{pesarin2010permutation,
address = {Chichester},
author = {Pesarin, Fortunato},
isbn = {0470516410},
series = {Wiley series in probability and statistics},
title = {Permutation tests for complex data : theory, applications and
software},
year = 2010
}
Am 29.08.2023 um 20:56 schrieb Ebhodaghe Faith:
> Hello,
>
> I performed an unpaired Wilcoxon test on a continuous measure with
> non-parametric distribution and detected a significant difference between
> two levels of an independent variable. However, I suspect that another
> independent variable could be confounding this outcome. Could you kindly
> advise on how I can control for this possible confounder using a
> multivariate Wilcoxon test? I'm not quite sure how to go about this.
>
> Thank you,
> Faith
>
> [[alternative HTML version deleted]]
>
> _______________________________________________
> R-sig-mixed-models at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models
[[alternative HTML version deleted]]