Re: Nemenyi's multiple comparison test:
It's analogous to Tukey's test, using rank sums in place of means
and using sqrt(n*(nk)(nk+1)/12) as the estimate of standard error (SE),
where n is the size of each sample and k is the number of samples (means).
Similarly to the Tukey test, you compare
(rank sum A - rank sum B)/SE to the studentized range for k.
It should be equivalent to the Dunn/Miller test described in
Hollander and Wolfe's Nonparametric Statistical Methods (Wiley, 1973,
p. 124), which uses mean ranks and standard error sqrt(k*(nk+1)/12).
Zar describes this version as well.
The original reference for Nemenyi is his doctoral dissertation,
which might be a bit hard to track down. Hollander and Wolfe
should be at least as easy to find as Zar (which now has a 3rd edition
from 1996, by the way).
Hope this helps,
>>Kathy Godfrey
[log in to unmask]
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|