Python - Mann Whitney U (exact) test
Вставка
- Опубліковано 5 лют 2025
- Instructional video on performing a Mann-Whitney U test with Python. Includes both the exact as the normal approximation.
This test is often used if you want to compare the results of an ordinal variable between two categories.
Companion website: PeterStatistic...
Jupyter notebook from video: bit.ly/3jG2TEe
Data file used: bit.ly/2WbhfSA
Excellent tutorial. Thanks for finally addressing the issue with Scipy not calculating the r-value (effect size)
Glad it was helpful!
This is extremely useful! Thank you!
Glad it was helpful!
Nice tutorial, you could consider do more videos for performing statistical test in Python. It is very difficult to find something with clear, step by step explanations for Python .
Thanks for the suggestion. I actually already have uploaded some videos on how to do a few tests with Python. Was there any particular test you were interested in?
@@stikpet Hi, actually is a few: some examples for anova with Post- Hoc test, Kruskal - Wallis, Chi square test of associations. I have not looked up Chi square or Kruskal Wallis yet but watching Anova tutorial gave me some headache.
I am very new to Python ( two months learning) , generally to programming.
@@antygona-iq8ew Here we go....
Kruskal Wallis H test: ua-cam.com/video/Af1YuaFylqM/v-deo.html
Post-hoc Dunn test: ua-cam.com/video/oDOlY3Hmatk/v-deo.html
And a regular 'one-way ANOVA': ua-cam.com/video/_X45N7ERtY4/v-deo.html
And post hoc for that: ua-cam.com/video/BOGCsSzQNR4/v-deo.html
On my website PeterStatistics.com you can find quite a lot although it is differently structured than most statistical websites. The Kruskal Wallis is what I recommend for a nominal-ordinal situation, so if you click on 'two variables - unpaired' and then nominal vs ordinal a menu on the left appears. If you then click on 'Test' I explain a bit about the test and there is then always mentioned 'click here to see...' if you click on that and then Python you get to see the video. Similar for nominal vs. scale for the one-way ANOVA.
@@stikpet Thanks a million!!! I will check everything out.