CFI index PreviousNext
Mplus Discussion > Confirmatory Factor Analysis >
Message/Author
 J.D. Haltigan posted on Saturday, August 20, 2011 - 7:21 pm
This may have been addressed before in the context of other parameters from various platforms, so apologies if so. My question pertains to fit indexes between LISREL/Mplus/Rstats(Lavaan package). Specifically, while the CFI between Mplus/Lavaan is the same, the CFI in LISREL is on the order +.10. This is also in the context of other fit statistics which are generally close between all three (e.g., SRMR). Any idea as to the 'more favorable' CFI generated by LISREL?
 Linda K. Muthen posted on Monday, August 22, 2011 - 3:05 pm
The baseline model for Mplus is means and variances of both dependent and independent variables and the covariances of independent variables. Factor indicators are dependent variables.

LISREL uses a baseline model of means and variances of all variables and no covariances among any variables.

This is why the LISREL CFI is better.
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: