Dear all,
I am conducting CFA and SEM with WLSMV, which is best option for the ordered-categorical data. However, I am wondering how does WLSMV handle the missing data in Mplus? Can I use FIML or multiple imputation with WLSMV in Mplus to handle the missing data? or Does this estimator only uses pairwise deletion method as a default option in Mplus?
P.S. I am only asking this question under the Mplus context, not other softwares.
Best,
Yes, WLSMV uses pairwise deletion, which is not the ideal method. You could use multiple imputation in Mplus instead and then fit your models to the imputed (complete) data sets.
Gert Lang Mplus uses pairwise deletion with WLSMV. See p. 5 of the document below:
https://www.statmodel.com/download/GstrucMissingRevision.pdf
- Badges
- Science topic
- Similar topics
- Linguistics
- Composition Studies
- Publication