0
$\begingroup$

The my dataset is not normal distributed. In order to apply a non parametric analysis, exist only the Kruskal Wallis test or others? I need to find a test similar to ANOVA to apply a dataset not normal distributed. Thanks so much.

$\endgroup$
7
  • $\begingroup$ What are you trying to accomplish with this test? Difference of means? $\endgroup$
    – ilanman
    Jul 4 '15 at 14:53
  • $\begingroup$ @ilanman I need to see the difference of means (similar the ANOVA). In my case I didn't use the ANOVA analysis because the may dataset is not normale distribuite. $\endgroup$
    – Giorgia
    Jul 4 '15 at 14:58
  • $\begingroup$ In my experience the Mann Whitney test works well with non-normal data. It's very powerful (compared to t test) for large samples and requires no assumption on the distribution of the data. $\endgroup$
    – ilanman
    Jul 4 '15 at 15:06
  • $\begingroup$ @ilanman is possible do you write my the script? $\endgroup$
    – Giorgia
    Jul 4 '15 at 15:07
  • $\begingroup$ I won't do your work for you but I am happy to help steer you in the right direction. What does your data look like? $\endgroup$
    – ilanman
    Jul 4 '15 at 16:29
1
$\begingroup$

You can perform ANOVA using ranks of your variable instead of variable itself. This will lead to the same results with Kruskal-Wallis test.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.