R help - Jul 2011 - significant results with KW but not in post-hoc test

Dear all,
I did a Kruskall-Wallis test for a comparison of a variable of interest
between 10 sites and I get a significant result (p=0.0019).  however, when I
perform a post-hoc test using kruskalmc from the pgirmess package, I get no
difference between any of the paired comparisons.  I cannot find anything in
the internet, not in my stats books about how to explain this contradictory
result.  I have also seen that somebody else asked this question in this
forum but she got no answers.  Can anybody help me understand what is going
on?

Thanks, Silvia.

--
View this message in context:
http://r.789695.n4.nabble.com/significant-results-with-KW-but-not-in-post-hoc-test-tp3649545p3649545.html
Sent from the R help mailing list archive at Nabble.com.