Comparing the impact of the independe... PreviousNext
Mplus Discussion > Multilevel Data/Complex Sample >
Message/Author
 David Llanos posted on Thursday, November 10, 2011 - 8:52 am
I'd would appreciate any comments on how to compare the relative impact of the independent variables in a multilevel path analysis model with a continuous dependent variable.

In my understanding of the classical path analysis we can have beta weights for each variable that can be directly compared against each other to determine what variable explained more variance of the dependent variable. Another alternative, with some caveats, is asking for standardized coefficients. However, so far it seems that neither of these options are not available in Mplus.

What would the equivalent approach you suggest to compare the substantive impact of the independent variables?

Thanks,

David
 Bengt O. Muthen posted on Thursday, November 10, 2011 - 7:08 pm
You can test the equality of the betas, for instance using Model Test.

Standardized coefficients are available in Mplus and can be tested using Model Constraint in addition to Model Test.

Sorting out variance contributed by a predictor isn't easy when the predictors are correlated.
 David Llanos posted on Friday, November 11, 2011 - 8:01 am
Thanks for the info, I'll explore those options in detail.
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: