I need to test if there is a difference between 6 conditions for 1 parameter. The values are non-parametric, c

could I do a Kruskal-Wallis one way analysis of variance?

I need to test if there is a difference between 6 conditions for 1 parameter. The values are non-parametric, c
yeh, throw it into Minitab and do an ANOVA
Reply:No, not as stated. First that test is for samples from a population. Conditions would imply that this may be


a more clinical style of experiment, and would require a non-parametric sequential selection. If so, start by testing the hypothesis that the treatment does not work on the patient.


You will find the result faster with fewer runs. If all values are non-parametric, you have to watch those anova tests that say they are non-parametric because they rely on normal distributions, but they do it more subtle like testing the median and comparing it to normal distribution. That is the Kruskal Wallis and it is not 'pure' non-parametric. While I can't rule it out from what you said, you would have to know more about the situation being tested or else look for a pure non-parametric test if you have to use the question as stated.

potential breakup song