Dear Allstaters,
I'm retrospectively analysing data for a new treatment of
a heterogenous group of patients. Treatment was applied at least
once, and repeated (at irregular intervals) until some criterion as
to treatment effect had been met.
I would like to find out whether any of the 9 groups of patients were
more likely to receive more treatment than others. Due to the
retrospective nature of the research, the size of the groups varies
from 4 to 20 members.
Would you agree that the median test is probably best for this? I
have also been suggested about Kruskall-Wallis, but the
data are not continuous. Any other ideas?
Paul
---------------------------
Paul H Artes, MCOptom
Research Optometrist
University of Manchester
[log in to unmask]
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|