Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down at the end of May, and its replacement, statalist.org is already up and running.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: Applying specific aweight to each explanatory variable


From   Steven Samuels <sjsamuels@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: Applying specific aweight to each explanatory variable
Date   Sun, 13 Nov 2011 14:59:31 -0500

It turns out that -eivreg- requires a constant measure of reliability per predictor, so shrinkage looks like the only practical, approach if reliability varies between observations.






Austin,

If the poster is worried about measurement error in the predictors, then -eivreg-  is a direct approach.  But Alice Whittemore showed long ago that the shrinkage estimator you suggested was a good general solution.

Whittemore, A. S. (1989). Errors-in-variables regression using Stein estimates. American Statistician, 226-228.

Steve


On Nov 12, 2011, at 11:09 PM, Austin Nichols wrote:

Steve--
But see answers the first time this question was asked e.g.
http://www.stata.com/statalist/archive/2011-11/msg00551.html

On Sat, Nov 12, 2011 at 9:04 PM, Steven Samuels <sjsamuels@gmail.com> wrote:

Your request does not make sense.  Analytic weights are proportional to the the residual variance of the dependent variable for each observation. Their use minimizes the standard errors of the regression coefficients.  Please look at any regression text.

Steve
On Nov 12, 2011, at 10:01 AM, <N.M.Postel-Vinay@lse.ac.uk> <N.M.Postel-Vinay@lse.ac.uk> wrote:

Hi All,

I'm running simple cross-sectional regressions. I would like to know whether it is possible to make Stata apply a different aweight to each separate explanatory variable.

Usually at the end of the regression I can just write [aweight = exp], but this would apply the same aweight to each explanatory variable. I would like to know whether it's possible to specify a different exp as an aweight for each different explanatory variable, and whether that would make any sense or not. My explanatory variables are financial ratios, so I would need a different weight for each different denominator.

Many thanks,

Natacha




*
*   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/


© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index