Re: Kruskal-Wallis critical values
- From: "\"Luis A. Afonso\"" <licas_@xxxxxxxxxxx>
- Date: Thu, 22 Feb 2007 19:04:27 EST
Kruskal-Wallis critical values (C.V.)
__k = number of samples
__j = number of values per sample
_________k=4__________j=4
______C.V. _95%___99%___
___________7.21___ 9.25___
What about TIES?
I performed two experiences
a) r(1) = r(2) = 1.5, r(15) = r(16) = 15.5
___________7.21____9.21
Any difference for the 95% conf. level, a slight one for 99%.
b) r(1) = r(2) = r(3) = r(4) = 2.5
___________7.15____9.15
For a tie of 4 items the differences on C.V. are substantial.
_______licas (Luis A. Afonso)
.
- Prev by Date: Re: Kruskal-Wallis critical values
- Next by Date: Re: Do people actually believe in Dr. Doofus?
- Previous by thread: Re: Kruskal-Wallis critical values
- Next by thread: Re: Kruskal-Wallis critical values
- Index(es):