Re: st: Constructing a variable from standard deviations

[Maarten buis <maartenbuis@yahoo.co.uk>]

--- On Mon, 22/11/10, M.P.J. van Zaal wrote:
> I am inclining to agree with mr Kolenikov, because the
> study i am (partly) replicating in my thesis does use OLS
> estimates.  

As I just mentioned I am not against Stas' argument, but I
do not find the fact that there exists a publication that 
also uses it a convincing argument: A lot of mistakes get
published and many get perpetuated using exactly this 
argument.
 
> They do use robust standard errors in the Mincerian wage
> regression. In the second step (regression of standard
> deviations on individual characteristics) they use robust
> errors en clustering. Does robust errors and clustering
> indicate that I should use mr Buis his method or can I
> still stick with OLS?

If you use -regress- in this context you must use robust
standard errors. So, the fact that the authors used 
robust standard errors does not help you with the choice
between models. As I mentioned before the main difference 
is in the style of the argument that you want to make.

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/