Multigroup Non-Convergence

[Aaron Lim]

Hello all,

I was recently working on a project that would require the fitting of a multigroup SEM model.

I found that the model was able to converge and achieve great fit in all the group when the model was fit individually by filtering the data then fitting the model to the filtered data. However, when I conducted the multigroup analysis, there was non-convergence.

What are the possible reasons for this phenomenon?

[Maher Said]

Hello Aaron, what error have you received?

[Aaron Lim]

Hello!

I did some testing and found that the issue was actually caused by the fact that specifying a multigroup analysis automatically gets lavaan to estimate intercepts (meanstructure = TRUE).

I managed to use one of the provided datasets in lavaan to create an example to illustrate my issue, which is that the a particular model exhibits different levels of fit when intercepts are estimated. 

In this particular case, when the model is fit without specifying "meanstructure = TRUE", I receive a warning.

## Warning in lavaan::lavaan(model = model, data = PoliticalDemocracy, model.type = "sem", : lavaan WARNING: ## the optimizer warns that a solution has NOT been found!

However, the warning seemingly goes away when I specify "meanstructure = TRUE".

Below is the full code to replicate the issue (also attached as a knitted html file). I would think the problem is actually caused by the introduction of the latent factor "common" that is trying to account for common method bias using an unmeasured latent method construct, the problem disappears when I remove that factor.

## The industrialization and Political Democracy Example ## Bollen (1989), page 332 model <- ' # latent variable definitions dem60 =~ y1 + y2 + y3 + y4 dem65 =~ y5 + y6 + y7 + y8 common =~ y1 + y2 + y3 + y4 + y5 + y6 + y7 + y8 # latent correlations common ~~ 0*dem60 + 0*dem65 dem65 ~ dem60 ' fit <- sem(model, data = PoliticalDemocracy)

# fit produces a warning but fit2 does not

fit2 <- sem(model, data = PoliticalDemocracy, meanstructure = TRUE)

Here is also my session info.

## R version 4.0.5 (2021-03-31) ## Platform: x86_64-apple-darwin17.0 (64-bit) ## Running under: macOS Catalina 10.15.7 ## ## Matrix products: default ## BLAS: /Library/Frameworks/R.framework/Versions/4.0/Resources/lib/libRblas.dylib ## LAPACK: /Library/Frameworks/R.framework/Versions/4.0/Resources/lib/libRlapack.dylib ## ## locale: ## [1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8 ## ## attached base packages: ## [1] stats graphics grDevices utils datasets methods base ## ## other attached packages: ## [1] lavaan_0.6-8 ## ## loaded via a namespace (and not attached): ## [1] digest_0.6.27 R6_2.5.0 jsonlite_1.7.2 stats4_4.0.5 ## [5] magrittr_2.0.1 evaluate_0.14 rlang_0.4.10 stringi_1.5.3 ## [9] jquerylib_0.1.3 bslib_0.2.4 pbivnorm_0.6.0 rmarkdown_2.11 ## [13] tools_4.0.5 stringr_1.4.0 xfun_0.22 yaml_2.2.1 ## [17] compiler_4.0.5 mnormt_2.0.2 tmvnsim_1.0-2 htmltools_0.5.1.1 ## [21] knitr_1.36 sass_0.3.1