I have a set of pearsons r effect sizes , converted to fishers z scores for meta analysis. I am wondering what data I need to calculate an inverse variance weight for each effect size and how I calculate these please? Also , is there an easy online calculator or spss method for doing this?



1 Answer 1


I you have the sample sizes, which I assume you do then the standard error is

$$ \frac{1}{\sqrt(N-3)} $$

So to get the sampling variance you just square that.

$$ \frac{1}{(N-3)} $$

  • $\begingroup$ Thanks mdewey. Am I correct in thinking that I need to use the above formulas for each individual effect size (not a total or average of all) while in the form of a fishers z score please? $\endgroup$
    – laboh
    Commented Aug 4, 2018 at 14:31
  • $\begingroup$ Yes, for each one. $\endgroup$
    – mdewey
    Commented Aug 4, 2018 at 15:24
  • $\begingroup$ Also , I have two different sample sizes related to two correlations ( both averaged in fishers z) in one article as it uses two studies within it. Is it acceptable to calculate an average of the two sample sizes for the standard error and inverse variance calculation please? $\endgroup$
    – laboh
    Commented Aug 4, 2018 at 19:09
  • $\begingroup$ Hi mdewey - when you say "So to get the sampling variance you just square that" - do you mean N as in sample size or the calculated standard error please? $\endgroup$
    – laboh
    Commented Aug 4, 2018 at 19:53
  • $\begingroup$ Yes, it is the sample size. $\endgroup$
    – mdewey
    Commented Aug 5, 2018 at 8:50

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.