1
$\begingroup$

I am conducting meta-analytic SEM and interested in looking at one outcome. A previous meta I was going to use as part of my data reports meta-analytic correlations for this outcome, however it is pulled into two sub-categories. It is more common in recent literature to report one outcome. Can I just take an average of these two meta-analytic correlations or do I need to use a different approach in combining?

Thank you for your help.

Improve this question
$\endgroup$
4
  • $\begingroup$ Dear Curious, welcome to a universe of other curious people! Perhaps you could add some details to help clarify the question. As I'm sure you know, one can average the numbers of red jelly beans in two jars, but one can't average the proportions of red jelly beans in two jars. Which kind of problem are you dealing with? $\endgroup$ Commented Mar 21, 2019 at 16:31
  • $\begingroup$ I think you are going to have to provide more detail here for anyone to try a response. What do you mean by a meta-analytic SEM? What are the meta-analytic correlations reported between? When you say average the correlations do you mean take their arithmetic mean, transform first, or meta-analyse them? $\endgroup$ Commented Mar 21, 2019 at 16:32
  • $\begingroup$ Explore the mvmeta R or Stata package. It should be able to pool correlated outcomes. $\endgroup$ Commented Mar 22, 2019 at 14:51
  • $\begingroup$ Welcome C M-A! If I understand your question correctly (I agree w/ @mdewey that more detail--and perhaps an example--could be helpful), you may find my response on this thread addresses your question about averaging the two correlations: stats.stackexchange.com/questions/327186/… $\endgroup$ Commented Mar 23, 2019 at 6:41

0

Reset to default

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.