Hi,
Please could someone put me out of my misery.
I have performed a couple of signed rank tests and am being continually
badgered by a colleague to produce confidence intervals for these results
on the grounds that
'the BMJ wont publish anything if the results don't all have confidence
intervals'.
I have seen another (BMJ) paper in which CIs were given for a non-paired
sample Wilcoxon test
but as my Mum always said 'If he went and jumped off a cliff would you do
it too?'
Any help would be gratefully recieved
(unless it shows that I'm wrong then I'll probably just sulk for a bit).
Jon
Research Statistician
Child Health
Univ. of Bristol
|