Re: st: Standardized Coefficients after Regress with Cluster Adjustment

["Roberto Liebscher" <roberto.liebscher@ku-eichstaett.de>]

I think I got the point. The problem is that we use an estimate for the
standard deviation of each variable and assume that this standard
deviation stems from the same population. But this only makes sense if the
standard deviation in each cluster is the same.

Have I got you right?

Thanks for your kind help.

Roberto

> On Mon, Jun 18, 2012 at 9:28 AM, Roberto Liebscher wrote:
>> When you want to get the standardized coefficients you are usually
>> allowed
>> to add the option -beta- to the regress command in stata. However this
>> does not work when you use the option -cluster(group)- to adjust the VCE
>> for cluster correlated disturbances.
>>
>> What is the reason for this and is it allowed to calculate the
>> standardized coefficients by hand in this case? Unfortunately I can not
>> understand why this should not be allowed as standardization is just a
>> transformation of the coefficients.
>
> You are allowed to do whatever you like. However, with such freedom
> comes responsibility. It is now up to you to convince others that what
> you did actually makes sense.
>
> To me the transformation implicit in standardized coefficients makes
> (some) sense when there is a meaningful estimate of the standard
> deviations for each variable. If that is in doubt, e.g. because you
> have a clustered sample, than standardized coefficients cease to be
> meaningful.
>
> Hope this helps,
> Maarten
>
> --------------------------
> Maarten L. Buis
> Institut fuer Soziologie
> Universitaet Tuebingen
> Wilhelmstrasse 36
> 72074 Tuebingen
> Germany
>
>
> http://www.maartenbuis.nl
> --------------------------
> *
> *   For searches and help try:
> *   http://www.stata.com/help.cgi?search
> *   http://www.stata.com/support/statalist/faq
> *   http://www.ats.ucla.edu/stat/stata/
>


*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/