FIi Indices PreviousNext
Mplus Discussion > Growth Modeling of Longitudinal Data >
Message/Author
 Jaime Maerten-Rivera posted on Saturday, February 20, 2010 - 9:11 am
I am running a model that has a non-singificant chi-square statistic, thus providing evidence for model fit. Is is still standard procedure to report other fit indices (i.e., CFI, TLCI, RMSEA, SRMR)? Also is there any literature out there evaluating how these other fit indices will perform given that the model fits according to the chi-square statistic? The model is a growth model- not sure that it matters.
 Linda K. Muthen posted on Saturday, February 20, 2010 - 9:55 am
If chi-square is good, then the other fit statistics should also be good given that they are based on chi-square. I believe multiple fit statistics are reported.
 Jaime Maerten-Rivera posted on Tuesday, March 02, 2010 - 1:24 pm
In the model, the chi-square is ns and most of the statistics suggest good fit with the exception of the SRMR which is 0.116 (N = 191, but there are missing data):

The residual output does not show values that are very high- the highest is 0.052. I think that I am looking at the correct table and that the values are standardized - I am looking at the table captioned "Residuals for Covariances/Correlations/Residual Correlations." What could be causing the SRMR to be so high? Is it somehow weigthed or is something else taken into account when it is calculated?
 Linda K. Muthen posted on Wednesday, March 03, 2010 - 11:05 am
I would have to see the output to answer your question. Please send it and your license number to support@statmodel.com.
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: