Control Covariates in metaSEM

Posted on
No user picture. Xin Lin Joined: 06/10/2019
Forums

Hi all,

Can we control covariates (i.e., age, SES, problem type) in metaSEM? To be more specific, can I add them directly to the model like regression?

Thank you!

Replied on Wed, 03/18/2020 - 20:36
Picture of user. Mike Cheung Joined: 10/08/2009

Dear Xin Lin,

Yes, you may refer to the following paper.

Jak, S., & Cheung, M. W. -L. (2019). Meta-analytic structural equation modeling with moderating effects on SEM parameters. Psychological Methods. Advance online publication. https://doi.org/10.1037/met0000245

The R code for the examples is available at https://github.com/mikewlcheung/code-in-articles/tree/master/Jak%20and%20Cheung%202019

Best,
Mike

Replied on Thu, 03/19/2020 - 15:54
No user picture. Xin Lin Joined: 06/10/2019

In reply to by Mike Cheung

Dear Dr. Cheung,

Thanks for getting back to me! I have read the paper and used the code with my data. I am trying to conduct a TSSEM with moderators. Yet, that paper only included codes for OSMASEM.

I have tried to modify the OSMASEM code for TSSEM using
random2 <- tssem2(random1, model.name="Moderator with WPF", Amatrix=M1a, Smatrix=S1, intervals.type="z")

and got that Error message
Error in (function (cl, name, valueClass) :
‘name’ is not a slot in class “list”

Can we conduct a moderation analysis with TSSEM?
I appreciate your help!

Best,
Xin

Replied on Fri, 03/20/2020 - 04:54
Picture of user. Mike Cheung Joined: 10/08/2009

Dear Xin,

tssem2() does NOT allow continuous moderators. As shown in the paper, it is osmasem() which allows moderators.

Best,
Mike