Standardized coefficients for mixed

[Liza]

Dear all,
Is it possible to obtain standardized coefficients for mixed?

Thanks in advance.

Best regards,
Liza Vieira

[Clyde Schechter]

There is no built-in option in -mixed- analogous to -regress-'s -beta-. You can standardize the variables in the model yourself before running it But it also raises some definitional questions. For example, is the standardized version of a variable on that is standardized to the entire sample, or just to one of the subsample corresponding to its value of one (or more) of the higher-level variables? And if so, which one? Standardized regression coefficients in multi-level models have all the drawbacks of standardized regression coefficients in flat models, plus additional ones. Do you really need to do this?

[Liza]

Dear Clyde Schechter,

thank you for your response.

Best regards,

Liza Vieira

[Filip VD]

Dear Lize,

You can get beta's after mixed through

mixed dep ind || nesting:,
estadd beta
matrix list e(beta)

From what I can discern this uses the standardization method as suggested by Hox and colleagues and described in formula (2.13) on p. 22 of http://joophox.net/mlbook2/Chapter2.pdf
This is an excerpt from the book
Hox, J. J., Moerbeek, M., & van de Schoot, R. (2010). Multilevel Analysis: Techniques and Applications, Second Edition.

Hope this helps,
Kind regards,
Filip