Wilcoxon (and many other permutation tests) work for identifying whether
two dependent distributions are identical if the data is given as
matched pairs.
The test will not be able to give confidence intervals for how similar
they are unless you assume a location shift model.
If you are interested in confidence intervals for the difference in
their means, check out my tests (and software) that are exact. R package
called npExact.
https://homepage.univie.ac.at/karl.schlag/statistics.php
Similarly, if you are interested in confidence intervals for the
difference in their medians, check out the exact tests i have explained
in (Schlag, 2015).
https://homepage.univie.ac.at/karl.schlag/research/statistics/exact%20np%201.pdf
Note that exact means that they are correct in the finite sample, and do
not rely on asymptotic approximations as does the t test. One day i do
believe people will start using correct (ie exact) tests.
greetings, karl
On 27.9.18 11:59, Michail Tsagris wrote:
> Hello.
>
> Does anyone know of a test for comparing 2 distributions from dependent samples?
> Kolmogorov-Smirnov and Anderson-Darling work with independent samples. Is there a modification for dependent samples?
> Any suggestion could prove useful.
>
> With kind regards,
> Michail
>
> You may leave the list at any time by sending the command
>
> SIGNOFF allstat
>
> to [log in to unmask], leaving the subject line blank.
--
---------------------------------------------------------------------
Karl Schlag
Professor
Department of Economics email: [log in to unmask]
University of Vienna landline: +43-1-4277-37437
Oskar-Morgenstern-Platz 1 mobile: +43-664-60277-37437
1090 Vienna, Austria fax: +43-1-4277-9374
room: 5.641 (fifth floor)
http://homepage.univie.ac.at/karl.schlag/
You may leave the list at any time by sending the command
SIGNOFF allstat
to [log in to unmask], leaving the subject line blank.
|