Measurement Invariance in Lavaan: Correlating error terms of specific groups

[Benedict Antazo]

Good evening everyone.

I'm currently testing the measurement invariance of a resilience scale across 3 countries. However, when I'm using lavaan to fit the model, it seems that there is no option to correlate error terms on specific groups (e.g. correlate Item1 ~~ Item2 only on group 2) rather than to the whole model. Is there a syntax that could help me do that?

Thank you very much,

Ben

[Terrence Jorgensen]

there is no option to correlate error terms on specific groups (e.g. correlate Item1 ~~ Item2 only on group 2)

So you do not even pass configural invariance?  This might be informative:

https://doi.org/10.3389/fpsyg.2017.01455

 

Is there a syntax that could help me do that?

Freeing it in one group but fixing it in another can be done in the syntax:

Item1 ~~ c(0, NA)*Item2 # free in group 2 only

Terrence D. Jorgensen

Assistant Professor, Methods and Statistics

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

http://www.uva.nl/profile/t.d.jorgensen

[Benedict Antazo]

Thank you very much, Prof. Terrence!