Multigroup CFA Invariance Tests (help making it more difficult)

[Gary Burns]

Greetings, my google searches have failed me, so I'm turning this google group instead. 

I understand the great shortcuts that lavaan has in place to help me test different levels of variance, but now I need to turn those off and do some work manually. 

I'll give two examples of things that I could do other programs but that I'm having a hard time doing in lavaan. 

Example 1

I'm comparing 5 items on a single factor across two groups. I have loading variance, as measured by a significant increase in misfit, and I think it comes from item five. I now want to constrain items 1-4 to be invariant across groups but allow 5 to freely estimate. 

Example 2

I'm comparing 5 items on a single factor across three groups. I have loading variance, as measured by a significant increase in misfit, and now I want to explore my data further. Specifically, I want to compare groups 1 and 2 allowing 3 to be freely estimated. (I can do this by subsetting my data and comparing just 1 and 2 without 3, but lets assume I really want three in there).

Suggestions for resources that would help me make those kinds of tests? I'm leaning towards mplus2lavaan or lisrel2lavaan, but if possible I would like to keep it all in the native language. 

[car...@web.de]

Hi, such things are explained very well in the tutorial: http://lavaan.ugent.be/tutorial/groups.html

Am 06.02.19, 23:55, Gary Burns <garyn...@gmail.com> schrieb:

--
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.

[Gary Burns]

I wanted to disagree with you, but there is a paragraph there that I missed. Exactly what I wanted. Now I wish I hadn't spent so much time exploring lisrel2lavaan.

For others, I'm copying it below:

Fixing parameters in some groups, but not all

Sometimes, we wish to fix the value of a parameter in all groups, except for one particular group. In this group, we wish to freely estimate the value of that parameter. The modifier for this parameter is again a vector containing the fixed values for this parameter for each group, but we can use NA to force a parameter to be free in one (or more) group(s). Suppose for example we have four groups. We define a latent variable (say f) with three indicators. We wish to fix the factor loading of indicator item2 to 1.0 in all but the second group. We can write something like

f =~ item1 + c(1,NA,1,1)*item2 + item3