Meta-Analysis Resources

Tools for Those Who Summarize the Evidence Base

Resources and networking for those who conduct or interpret meta-analyses related to any phenomenon that is gauged in multiple studies.

Hi, Srikanth.

For categorical moderators, it is possible to model it with a multiple-group approach. The answer is both no and yes. "No" means that it is not yet available in the metaSEM package. "Yes" means that it can be done in OpenMx. Suzanne Jak ( and I have a paper under review on this topic. We have some R code to do the analysis. She is the correspondent author of the paper. You may email her for a copy of the paper.

For continuous moderators, it is much more complicated. One suggestion is to categorize the continuous moderators into categorical moderators. However, this may lose some information. S.F. Cheung and I discus how to handle heterogeneity that may also be related to your case.

Cheung, M. W.­L., & Cheung, S.­F. (in press). Random­ effects models for meta-­analytic structural equation modeling: Review, issues, and illustrations. Research Synthesis Methods.

Hope it helps.


> Hi,
> I recently bought the book on "Meta-Analysis: A structural equation modeling approach"

by Dr. Cheung.
> I am planning run a path model, with indirect effects, similar to the path model of

"cognitive ability, job knowledge, work sample and supervisor rating", in Figure 7.4 page 266.
> I want to know whether your R packages have the capability to analyze

categorical\continuous moderators for such a path model, when I run it using the

random effects model. Say, I run the model in Figure 7.4 using a random effects model

similar to the one run in page 269. Then I want to test whether coefficients of this model

are significantly different for my groups (of the moderator). Please let me know if such

a modeling can be done.
> One way could be, I run random effects models for each of my groups. Then how do

I test for significance?
> Thanks.

Views: 79

Replies to This Discussion

Thank you so much Dr. Cheung. 

Also, excuse me I spammed all inboxes, instead of creating a new discussion.


© 2017   Created by Blair T. Johnson.   Powered by

Badges  |  Report an Issue  |  Terms of Service