1
$\begingroup$

Every book/article claims this as if it's self-evident, but I have a hard time figuring out why. How do you get from the sum-of-squares style statistic of Friedman to the difference statistic involved in the sign test?

Improve this question
$\endgroup$
1
  • $\begingroup$ Perhaps a proper reference is due. In any case, what does the sign test do? What does it test? Does it only involve one sample and if so, what should the extension do for $k$ samples? Does Friedman do the job? $\endgroup$
    – User1865345
    Commented Sep 8, 2022 at 15:49

0

Reset to default

Your Answer

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