I want to visualize original confidence interval of individual studies in forest plot; however, weighting will be used to create pooled estimate.
I used weighted=FALSE. My code
res <- rma(lnhr, sei=se, data=dffc, method="REML", test="knha", weighted = FALSE)
res
par(mar = c(6, 6, 3, 3))
forest(res, addpred = FALSE, annotate= TRUE, at=seq(0,8,by=1),
order= dffc$exo,
showweights=TRUE,
ilab=cbind(expo, com), ilab.xpos=c(-3, -1.5), mlab = "", xlab= " ",
cex=1.1, xlim=c(8, 4), alim = c(-6, 6),
transf=transf.exp.int,
refline=1,
slab=paste(dffc$author)
)
My plot enter image description here
So, it is **unweighted pooled estimate; visualization of confindence interval of individual studies shows weighted confidence interval. **
However, I want opposite. Weighted pooled estimate and unweighted confidence interval of individual studies.