dear allstat,
i have a question regarding transformation in meta analysis in R.
i have a number of studies that i would like to estimate a pooled proportion of a single event.
for example
study % of cases with Event
1 0.01
2 0.2
3 0.4
.
.
.
N 0.09
But as I can see from the output, in studies that the %event is 0% or 100% and in order to produce confidence intervals (and the forrest plot) then the value is transformed to a vary closed to 0% or 100% value.
Since i cannot find anything at the manual, could you please explain or send some references regarding this transformation?
Thank you in advance
Evie
You may leave the list at any time by sending the command
SIGNOFF allstat
to [log in to unmask], leaving the subject line blank.
|