3
$\begingroup$

I am quite familiar with mixed effects (multilevel models) in meta analyses.
Now, I want to do a mixed effects (multilevel) meta analysis structural equation model (MASEM) with STATA and the gsem package because my data format is long.

My data look like:
id studyId performance education moderator1 moderator2
1 1 -0.4 0.1 1 0
2 1 0.2 0.1 0 1
3 2 0.5 0.3 0 1
4 3 -0.1 0.4 1 0
5 3 -0.6 0.4 1 0

performance: Firm-Performance, correlation coefficients (Fisher z transformed)
moderator1-moderator2: My moderator variables, have some nusisance variables as well but do not display it here.
education: Pre-dominant education level in a certain country

All variables are weighted by w = fisher z variance + tau

I want to model primary study effects nested in primary studies (results grouped in the related study).

gsem (performance <- moderator1 moderator2 education M1[id]@1 M2[id>studyId]@1)  
(moderator1 <- education M3[id]@1 M4[id>studyId]@1)  
(moderator2 <- education M5[id]@1 M6[id>studyId]@1),  
Latent(M1 M2 M3 M4 M5)  
cov(e.performance M1[id]*M2[id>studyId])
cov(M3[id]*M4[id>studyId])
cov(M5[id]*M6[id>studyId])  

I am not sure about the specification of the model in STATA.
1. Did I model the covariances in a correct way?
2. @1 means I normalize the latent variable. What alternatives do I have?
3. Is it enough just to model the error term (e.performance) of the dependent variable performance or do I have to model all variables there?

Thank you very much for your support :-)

$\endgroup$

closed as off-topic by mdewey, Michael Chernick, Dougal, John, Nick Cox Apr 30 '17 at 9:54

This question appears to be off-topic. The users who voted to close gave this specific reason:

  • "This question appears to be off-topic because EITHER it is not about statistics, machine learning, data analysis, data mining, or data visualization, OR it focuses on programming, debugging, or performing routine operations within a statistical computing platform. If the latter, you could try the support links we maintain." – mdewey, Michael Chernick, Dougal, John, Nick Cox
If this question can be reworded to fit the rules in the help center, please edit the question.

  • 1
    $\begingroup$ The metaSEM package in R provides better support to meta-analytic structural equation modeling. If you really want to do it in STATA, you may contact Tom Palmer. Tom and I are working on a paper on implementing MASEM in STATA. $\endgroup$ – Mike Cheung Nov 29 '16 at 2:21
  • $\begingroup$ Dear Mike, thank you very much for your answer. I sent you an Email. Would be great to receive an answer. $\endgroup$ – Dominik Wagner Dec 4 '16 at 17:15