Summary data from individual data mis... PreviousNext
Mplus Discussion > Missing Data Modeling >
 Jason Sudarshan posted on Sunday, July 14, 2019 - 10:21 pm
Apologies if this is a repost - I cannot see my first attempt to post.
I save the sample covariance matrix in a CFA ML run with individual data with missing values. I run the same CFA with the saved sample covariance matrix ( and specify the same sample size). The H0, H1 and chi-square values are different between these two runs. I need the covariance matrix (and means and sample size) to run a multigroup invariance test and I do not have individual data for the other group.How do I get summay data that will produce same fit measures? Thank you very much for any help.(I am using Mplus 7, and have access to Mplus 8.3)
 Tihomir Asparouhov posted on Monday, July 15, 2019 - 7:37 pm
This is not possible in Mplus currently and there are no such summary data in principle (in the presence of missing data). I would recommend that you generate data (with Mplus Montecarlo) using the summary data (possibly even adjust the generated sample by adding constants and multiplying by constants to match the summary as close as possible). You can then conduct the multiple group analysis using raw data for both groups. If you follow this you must use the ML estimator rather than MLR (ML would be the only valid inference).
 Jason Sudarshan posted on Monday, July 15, 2019 - 8:26 pm
Thank you very much for the prompt reply and suggestion.
Back to top
Add Your Message Here
Username: Posting Information:
This is a private posting area. Only registered users and moderators may post messages here.
Options: Enable HTML code in message
Automatically activate URLs in message