Using IRT factor scores for SEM

Hi everybody,

I did some extended research about this question but didn't find a good answer: Is it ok to use IRT results (factor scores from package mirt) as input variable for (classical) SEM (lavaan) regression analysis?


In my educational (pre-)study I measured several psychological scales via IRT-methods (mostly GRM) and got valid models and factor scores. I also measured an learning achievement test which I evaluated with MIRT, too. Now I want to know which psych factor has the biggest impact on learning achievement - a classical SEM problem, isn't it? But can I use my IRT data directly for the lavaan-model or are there some caveats to consider?

I am also glad if you can refer me to some appropriate literature.

Thanks a lot in advance,


This topic was automatically closed 21 days after the last reply. New replies are no longer allowed.

If you have a query related to it or one of the replies, start a new topic and refer back with a link.