Missing data, nonzero residuals, zero... PreviousNext
Mplus Discussion > Missing Data Modeling >
Message/Author
 Roger E. Millsap posted on Wednesday, September 02, 2009 - 6:36 pm
I have a path model that has three endogenous variables, with overlapping exogenous variables except for a subset
of such variables. So it is not saturated, but resembles a multivariate regression. All continuous variables. There are missing data, mostly on the endogenous variables. The output gives (1) zero chi-square, (2) positive df (=13), and (3) nonzero residuals. Some
of the residuals have standardized values above 2.5. Why would there be large residuals with a zero chi-square? The HO and H1 models give the same likelihood values. I can send the output if you want to see it. Thanks.
 Linda K. Muthen posted on Wednesday, September 02, 2009 - 7:34 pm
It's hard to say without seeing the full output. Please send that and your license number to support@statmodel.com.
 Bengt O. Muthen posted on Wednesday, September 02, 2009 - 7:56 pm
Sounds unusual indeed. For missing data to be the cause there would have to be odd instances of zero coverage (where H1 parameters are deleted).
Back to top
Add Your Message Here
Post:
Username: Posting Information:
This is a private posting area. Only registered users and moderators may post messages here.
Password:
Options: Enable HTML code in message
Automatically activate URLs in message
Action: