Model Fit Chi-Square in MPLUS versus ... PreviousNext
Mplus Discussion > Confirmatory Factor Analysis >
Message/Author
 Chris Schatschneider posted on Friday, December 16, 2005 - 8:04 am
Hello all,

Please forgive a potentially naive question that may have been answered already. I have fit a very simple 1 factor model with 4 indicators and I wanted to compare the MPLUS output with results from LISREL (and SAS). I am finding that my results differ substantially and I assume that I am not parameterizing the model in MPLUS correctly. Just to aid discussion, I have pasted my MPLUS syntax below.

TITLE: Testing SAS and MPLUS output;
DATA:
FILE IS C:\projects\fuchs\data\mplus_math.dat;
FORMAT IS FREE;
TYPE IS INDIVIDUAL;
NOBSERVATIONS ARE 634;
NGROUPS = 1;
VARIABLE:
NAMES ARE PRCBMDIG ADDTESTC SUBTESTC APRTOT NPRTOT FPRTOT
SPRTOT ;
usevariables are APRTOT NPRTOT FPRTOT SPRTOT;
ANALYSIS:
TYPE IS GENERAL;
ESTIMATOR IS ML;
ITERATIONS = 1000;
CONVERGENCE = 0.00005;
MODEL:
F1 by APRTOT*1 NPRTOT*1 FPRTOT*1 SPRTOT*1 ;
F1 @1;
OUTPUT: SAMPSTAT RESIDUAL STANDARDIZED;


This model gives me a chi-square fit value of 8.304 (df=2) but the “identical” model in LISREL (and SAS) gives me a chi-square of 1.48 (with 2dfs).

Any thoughts would be deeply appreciated.
 Linda K. Muthen posted on Friday, December 16, 2005 - 8:10 am
It is possible that you have hit a local solution for some reason. If you send your Mplus input, data, output, and license number to support@statmodel.com, I will be happy to look at this. Please also inlcude either the SAS or LISREL full output.
 Chris Schatschneider posted on Friday, December 16, 2005 - 9:58 am
Thanks for looking at this!
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: