Chi-square, MLR PreviousNext
Mplus Discussion > Multilevel Data/Complex Sample >
 SAMANEH SADEGHI posted on Monday, November 24, 2014 - 11:15 am
Dear Muthen,

I have run a longitudinal data analysis (autoregressive) cross-lagged. The results indicate poor model fit based on chi-squared as shown below:

Chi-Square Test of Model Fit 297.253*
Degrees of Freedom 6
P-Value 0.0000
Scaling Correction Factor 1.1038
for MLR

Chi-Square Contribution From Each Group
MALE 147.433
FEMALE 149.820

* The chi-square value for MLM, MLMV, MLR, ULSMV, WLSM and WLSMV cannot be used for chi-square difference testing in the regular way. MLMV, WLSMV, and ULSMV difference testing is done using the DIFFTEST option.

Estimate 0.294
Probability RMSEA <= .05 0.000

CFI 0.865
TLI -0.483

Chi-Square Test of Model Fit for the Baseline Model
Value 2225.690
Degrees of Freedom 66
P-Value 0.0000

Value 0.074

There is the option of doing the Satorra-Bentler scaling correction. I am cannot do this as I only have 1 corrected MLR. There is also an option given as DIFTEST which I am not familiar with. I would be grateful if you could explain this and how this can be done.
 SAMANEH SADEGHI posted on Monday, November 24, 2014 - 11:17 am
In addition to the message posted before I would be grateful if you could also explain how I can improve the model by using the model modification indices shown below.


Minimum M.I. value for printing the modification index 10.000

M.I. E.P.C. Std E.P.C. StdYX E.P.C.

ON Statements
W3OPP ON W3DEL 79.835 0.524 0.524 0.401
W3OPP ON W4OPP 116.189 8.101 8.101 7.624

WITH Statements
W3DEL WITH W3OPP 79.835 0.006 0.006 0.438
W3DEP WITH W3OPP 52.852 0.035 0.035 0.357

Many thanks
 Linda K. Muthen posted on Tuesday, November 25, 2014 - 10:17 am
To do a difference test, you need two analyses, one for each of the nested models.

Modification indices should be used that make substantive sense.

Please limit posts to one window.
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