I would like to ask a question regarding the output of the rma.mv function.

I had raw correlations which I transformed into Fisher's Z correlations using the 'escalc' function. Then, I performed a multilevel meta-regression with the 'rma.mv' function.

Now I would like to ask: is the "estimate" output that I get (i.e., the summary effect) given in terms of the raw correlation or the fisher Z transformation? I believe that since I entered the fisher-Z correlations, I should get the output in Fisher Z as well, but I would nevertheless like to verify it.

I attach the code and the output for convenience.

Thank you all for your help

  • CODE:

D <- escalc(measure="ZCOR", ri=r, ni=N, data=D, vtype="AV")

metRes1 <- rma.mv(yi=yi, V=vi, random = ~ factor(id) | Paper_num, data=D)


Model Results:

estimate se zval pval ci.lb ci.ub 0.3813 0.0671 5.6818 <.0001 0.2497 0.5128 ***


1 Answer 1


You can easily check this your self. Make up a small dummy data-set with large in magnitude r values. I suggest about 0.8 to 0.9. Then run rma.mv and check the estimate. If it is greater than 1 then it must be a z.

  • $\begingroup$ Thank you so much for the great idea! For those interested, I checked this and the results are indeed given in terms of Fisher Z (and not correlations). $\endgroup$
    – Lior
    Dec 17, 2019 at 15:04

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.