1
$\begingroup$

I am using a Friedman's test with post-hoc analysis. I am using the function friedman.test.with.post.hoc available for R software.

Lets say we have 6 wines and we want an ordered list of the wines according to the judges' rates (best wine, ... ,worst wine).

So, we perform the Friedman test. Suppose $p < \alpha$, we reject the null hypothesis. In other words, there is a significant difference between the wines given the judges' rates. In order to identify which wines are different, we run the post-hoc analysis. By running the post-hoc test, we have that there is only significant difference between the pairs: Wine A/Wine C and Wine A/Wine D.

Given that, how do you conclude the order of the wines? Should I use any other tools?

Thanks.

Improve this question
$\endgroup$

1 Answer 1

Reset to default
1
$\begingroup$

No I don't think there is any magic statistical tricks to sprinkle on this to tell you the order between Wine C and D. Probably the difference is too small to detect for the given sample size. the remedy is "Get more data!" I hope you made the appropriate multiplicity adjustments for conducting several pairwise tests.

Improve this answer
$\endgroup$
1
  • $\begingroup$ Ok. Thanks. I wish getting more data was as easy as saying that :) I guess that, for now, I don't have enough evidence to say that there is different between the wines, even though the Friedman test says so. $\endgroup$
    – prksch
    Commented Jul 23, 2012 at 1:23

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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