metaSEM

Dear Mike,

When I call summary (random2a, R=5000), z-values and standard errors of some estimates are significantly changed. The estimates and goodness-of-fit indices remain the same.

I didn't try (R=10000), as "R=5000" already took 7 hours to finish;)

However, if the number of replications dramatically influences the z-values and std.errors, are there any criteria for choosing one result for reporting?

Thanks!

Ryan

Dear Mike,

After struggling with my data in the stage 1 of TSSEM, I finally moved into the stage 2. However, the output seems not to be very good.

(1) The results of two approaches of specifying models are different.

Specifically, when using "diag.constraints=TRUE, intervals.type="z"", the result suggests "Amatrix[1,22]" and "Amatrix[1,23]" are significant. But removing the "constraints" and "intervals.type", the result suggests "Amatrix[1,21]" is significant.

In addition, both results have same goodness-of-fit values. Which result I should believe?

Hello All,

As the title suggests, I have a question regarding whether to fix the factor loading or the latent variable variance. Both can be fixed in order to derive a 'scale' for your latent variables. The fit statistics should be the same in both cases.

Hi all,

I encounter a problem of "x is not positive definite" when using Cheung's Two-stage MASEM. Could anyone provide some clues?

I attach part of my dataset as samples. My R-syntax is "random1<-tssem1(entrymode,samplesize,method="REM",RE.type="Diag")" . This command ends with an error information:

Error in (function(x, n, cor.analysis = TRUE, dropNA = FALSE, as.matrix = TRUE), : x is not positive definite!

Hi there,

I am using A/Prof Mike Cheung's TSSEM (See: https://dl.dropboxusercontent.com/u/25182759/Fixed%20and%20Random%20Effects%20Meta%20Analytic%20Structural%20Equation%20Modeling.pdf). The 2nd stage of this method is constructing a path model (a SEM model) based on the A, S, and F matrices of OpenMx.

Mike's examples (as well as the examples of OpenMx) show only 'reflective models' (i.e., the arrows of paths are from a latent variable to its elements.

My question is: can we build formative models (i.e., the arrows of paths are from elements to a latent variable)?

Dr. Mike,

I am running multiple structural models and appear to have the same issue with all the models. In each of the outputs the certain paths appear to be fixed to 1.0 and another value.

For instance in the structural model, EU -> BI -> U (files EUBIU.dat & EUBIU.R), the paths from EU -> BI = 1 and the path from BI -> U = 1.

Another model (files PUEUBI.data & PUEUBI.R), the paths EU -> PU = 1.0, while the paths from EU -> BI and PU -> BI are both equal to 1.0.

I think I may be constraining these paths to be either equal to 1.0 or equal to each other.

Dear all,

After model comparison with 'anova()' command, variation in level 3 turned out to be not significant which means that 3 level approach can be inappropriate for the data.