How to obtain std. coeff. not using M... PreviousNext
Mplus Discussion > Multilevel Data/Complex Sample >
Message/Author
 Paul Ringler posted on Thursday, July 19, 2007 - 8:21 am
Dear All,

i am conducting a multilevel analysis using MLA (a free multilevel analysis software by Busing et al.).
My model is a random-intercept model, with fixed effects at the individual level and second-level explanatory variables to explain the random variance in the intercept.

I would like to obtain the standardized coefficients of the individual level variables and of the explanatory variables for the random intercept.

I have two urgent questions and I would be happy if someone could give me a hint as to their answer.

Since this program does not automatically provide standardized coefficients of regression coefficients, it is necessary to do this manually.
Hox (2002) and others recommend simply re-running the model using standardized variables.
But should I enter only the standardized explanatory variables or the outcome variable as well? Unfortunately the literature I found was not precise on this subject.

Also, I could not find any clear references about the effects that standardizing the input variables has on the the estimation of a random intercept with explanatory second-level variables.
Literature only mentions that random effects are sensitive to standardization, does this apply to the random intercept as well?

Thank you for your time in advance!
Regards,
Paul Ringler
 Linda K. Muthen posted on Thursday, July 19, 2007 - 9:44 am
I would not work with standardized variables. It is not difficult to standardize the raw coefficients. See Technical Appendix 3.
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: