Modification indices from imputed data

88 views
Skip to first unread message

garc...@unlv.nevada.edu

unread,
Jul 30, 2018, 2:02:08 PM7/30/18
to lavaan
Hello, 

I have recently started working on a project where I am fitting several CFA models to a multiple imputation dataset. What I am trying to do is get the modification indices for my various fitted models.

So far in my attempts, I get this error message: 

Error in modindices(fit.mod.c2.wlsmvs.mi) : 
no slot of name "optim" for this object of class "lavaan.mi"

I read in an older post that there may not be a way to call the modification indices from this kind of data set. I am wondering if this is still the case or if anyone has successfully done this and may be able to help me? 

Thank you in advanced for any help!
Breanna Garcia 

Yves Rosseel

unread,
Jul 31, 2018, 5:18:11 AM7/31/18
to lav...@googlegroups.com
I am not sure if this possible. Because each imputed dataset would
produce a different set of modification indices.

I would suggest to fit a model for each imputed dataset (separately),
extract the modification indices, and only retain the ones that are
'high' in most (if not all) imputed datasets.

Yves.
> --
> 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
> <mailto:lavaan+un...@googlegroups.com>.
> To post to this group, send email to lav...@googlegroups.com
> <mailto:lav...@googlegroups.com>.
> Visit this group at https://groups.google.com/group/lavaan.
> For more options, visit https://groups.google.com/d/optout.

Terrence Jorgensen

unread,
Jul 31, 2018, 8:25:42 AM7/31/18
to lavaan
I get this error message: 

Error in modindices(fit.mod.c2.wlsmvs.mi) : 
no slot of name "optim" for this object of class "lavaan.mi"

I read in an older post that there may not be a way to call the modification indices from this kind of data set. I am wondering if this is still the case or if anyone has successfully done this and may be able to help me? 

You can use the modindices.mi() function, which properly pools the score-test statistics (or calculates them using other pooled information, if type = "Rubin").

Terrence D. Jorgensen
Postdoctoral Researcher, Methods and Statistics
Research Institute for Child Development and Education, the University of Amsterdam

Reply all
Reply to author
Forward
0 new messages