Hello all, a fast question. If I have two samples and one is non-parametric while the others follow the normal distribution. I will perform Kruskal-Wallis but then in order to investigate among pairs I can use t-test for the normally distributed data, but how can I compare the non-parametric dataset with the normally distributed?
![[Live] : ONE 168 วันนี้!! "โจนาธาน vs ซุปเปอร์เล็ก"](http://i.ytimg.com/vi/0jIN65KgJsI/mqdefault.jpg)
As always, simple and straightforward, most importantly, free. Thanks a lot!
You're welcome, thanks for watching -
thank you!
but what should i do if the distribution shape is not similar?
And if the distribution shape isn't similar? Which test should you use then?
beautifully explainedm thank you
hello, may we just conduct a KS test to see if these two independent samples have the same distribution in order to test the assumptions? thanks!
thank you so much! so well-explained!
You're welcome!
Hello all, a fast question. If I have two samples and one is non-parametric while the others follow the normal distribution. I will perform Kruskal-Wallis but then in order to investigate among pairs I can use t-test for the normally distributed data, but how can I compare the non-parametric dataset with the normally distributed?
Hi there.It is a very informative video.
What to do if the distribution shape isn't similar?
If the distribution shape is not similar, the test results represent the test of the mean rank instead of median.
Maybe a stupid question but what does that mean for the results?