Dear Allstatlisters,
I've been asked by a colleague about how to make
a metaanalysis of proportions (prevalences to be exact),
in fact he has the P and V(P) estimates from each study.
Could he use the pool them just weighting by
the inverse of the variance, say Wolf's method?.
Thanks indeed,
Aurelio
[log in to unmask]
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|