Compare path strength in two groups PreviousNext
Mplus Discussion > Multilevel Data/Complex Sample >
Message/Author
 gibbon lab posted on Monday, December 12, 2011 - 5:14 pm
Hi Professor Muthen,

I am running a two group analysis. For one of the paths, the beta is 1.71 with t=3.03 in one group and 4.30 with t=2.91 in the other group. The chi-square difference test shows that they are significantly different. In this case (lower beta with higher t value), which group should be considered to have a stronger path? Thanks a lot.
 Linda K. Muthen posted on Monday, December 12, 2011 - 6:17 pm
Given that they are scaled the same, I think you could say that 4.30 is stronger because it is a larger value and both are significantly different from zero.
 gibbon lab posted on Tuesday, December 13, 2011 - 10:01 am
Hi Professor Muthen,

That path has an exogenous latent variable and an endogenous latent variable. Can I assume the latent variables are scaled the same in both groups? Thanks a lot.
 Linda K. Muthen posted on Wednesday, December 14, 2011 - 12:04 pm
You can if measurement invariance holds.
 gibbon lab posted on Wednesday, December 14, 2011 - 12:25 pm
Is there a command in Mplus to hold the measurement invariance? Thanks a lot.
 Linda K. Muthen posted on Wednesday, December 14, 2011 - 1:15 pm
No. You would need to test for it. See the Topic 1 course handout under multiple group analysis.
 gibbon lab posted on Friday, April 06, 2012 - 1:46 pm
Hi Professor Muthen,

What if the measurement invariance does not hold? Does that mean the comparison between the two groups is invalid? Thanks a lot.
 Linda K. Muthen posted on Friday, April 06, 2012 - 2:52 pm
If measurement invariance does not hold, structural parameters cannot be compared across groups.
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: