Chi-SQ Significant; RMSEA and CFI Acc... PreviousNext
Mplus Discussion > Structural Equation Modeling >
Message/Author
 Chao Yawo posted on Tuesday, February 13, 2018 - 11:35 am
Hello,

I am fitting a structural model, from a complex survey data consisting of 2933 usable cases. My DV is categorical, so are almost all of my IVs (except two)

Here are my analysis commands.

Analysis:
Type = Complex;
Estimator = WLS;
parameterization = theta;

My model fit results:
-------------------------------
Number of Free Parameters 107
Chi-Square (df=20; P-value=0.0175) = 35.511

RMSEA (Root Mean Square Error Of Approximation)

Estimate 0.016
90 Percent C.I. 0.007 0.025
Probability RMSEA <= .05 1.000

CFI/TLI

CFI 0.995
TLI 0.970

-------------------------------------------------------------

This is the results of my my full model. I am inclined to report this model, but only interpret the significant path coefficients. The trouble is my chi-sq is significant, but the RMSEA and CFI/TLI seems acceptable.

1. Can i go ahead and use the model, despite the significant chi-square.

2. No SRMR values were reported - is there anyway to obtain this?

Thanks - cY
 Bengt O. Muthen posted on Tuesday, February 13, 2018 - 6:01 pm
This looks like a well-fitting model to me. The p-value is pretty good given the large sample.

If you use Version 8 and SRMR values are not reported, they are not available and may not be easy to compute yourself.
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: