SRMR PreviousNext
Mplus Discussion > Growth Modeling of Longitudinal Data >
 Daniel Lee posted on Monday, October 24, 2016 - 12:51 pm
Hi Dr. Muthen,
After running a conditional growth model (the time-invariant predictor was whether the participant was in the experimental or control condition), I elected to run a multiple growth model (group = condition/control) to examine parameters (slope, intercept, random effect parameters) in the experimental and control condition. Although model fit generally remained the same (RMSEA, TLI/CFI), the SRMR shot up from .05 to .10.

I was wondering if you can help me understand why SRMR might increase so drastically in this scenario (while other fit indicators remain stable from conditional growth curve model). Does the SRMR naturally increase when multiple groups are introduced? Or is this value sensitive to the inclusion of multiple groups?

Thank you so much!
 Tihomir Asparouhov posted on Monday, October 24, 2016 - 3:39 pm
The definition can be found here

There is no reason for "SRMR to naturally increase". You can use the output:residual; command and see where the increase comes from (in the two group case you will find that the residuals are bigger).

In the single group case fewer sample quantities are compared to the estimated model and some misfits can be "averaged out".
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