partialInvarianceCat() error for non-congeneric models

[christophe...@nicholls.edu]

I'm looking for advice in resolving an error message that I've received. 

Briefly, I've been trying to resolve a measurement non-invariance issue. Specifically, one of my measures seems to fail metric invariance and I'm trying to use the partialInvarianceCat() function. The error message that I receive is below: 

Error in partialInvarianceCat(mi, type = "metric") : 

  The model is not congeneric. This function does not support non-congeneric model.

My model is below:

mi.model <- '

#For each factor, the factor loading of one marker variable is fixed (usually as 1). Other factor loadings are freely estimated across groups.

##Method factors

CM =~ CM1 + CM2 + CM3 + CM4 + CM5 + CM6 + CM7 + CM8

PA =~ PA1 + PA2 + PA3 + PA4 + PA5 + PA6 + PA7 + PA8 + PA9 + PA10 

Na =~ NA1 + NA2 + NA3 + NA4 + NA5 + NA6 + NA7 + NA8 + NA9 + NA10 + CM7 + CM8

AP =~ PA1 + PA2 + PA3 + PA4 + PA5 + PA6 + PA7 + PA8 + PA9 + PA10 + NA1 + NA2 + NA3 + NA4 + NA5 + NA6 + NA7 + NA8 + NA9 + NA10

#Synonyms and Antonyms

CM1~~CM2

CM3~~CM4

CM5~~CM6

CM7~~CM8

#Bifactors freely estimated

PA ~~ 0*Na

PA ~~ 0*AP

Na ~~ 0*AP

'

Do you have any advice or guidance to offer? I can provide more information if needed.

Chris

[Edward Rigdon]

"Not congeneric"--you have items that load on multiple factors. That makes your model not congeneric.

--
You received this message because you are subscribed to the Google Groups "lavaan" group.
To unsubscribe from this group and stop receiving emails from it, send an email to lavaan+un...@googlegroups.com.
To post to this group, send email to lav...@googlegroups.com.
Visit this group at https://groups.google.com/group/lavaan.
For more options, visit https://groups.google.com/d/optout.

[Christopher Castille]

Thanks for your response, Ed. I’m aware that multiple factor loadings makes my model non-congeneric. I’m just not sure how to deal with this issue. 

[Terrence Jorgensen]

I’m just not sure how to deal with this issue. 

Again, these automation functions were only designed for the simplest cases. I'm eventually going to generalize the whole process, but it will take a while to finish coding it up.  In the meantime, you will just have to fit the models manually to test each item by freeing its loading to test whether that model fits as well as the configural model.

Terrence D. Jorgensen

Postdoctoral Researcher, Methods and Statistics

Research Institute for Child Development and Education, the University of Amsterdam

UvA web page: http://www.uva.nl/profile/t.d.jorgensen