Re: st: svyreg vs. reg

[Richard Williams <Richard.A.Williams.5@nd.edu>]

At 02:36 PM 5/14/2004 -0400, Paul von Hippel wrote:
> In analyzing weighted survey data, I'm surprised that the following commands
>         svyset [pweight=weight]
>         svyreg y x
> don't give the same results as
>         reg y x [pw=weight]
> or
>         reg y x [pw=weight], hc3
> or even
>         reg y x [aw=weight]
>
> I'm left confused as to what svyreg is doing. Any insights most appreciated.
Check out pages 357-358 of the Stata 8 Users Guide, i.e. Section 30.10, 
Differences Between the SVY Commands and Other Commands.  It states that 
"The point estimates from the svy commands are exactly the same as the 
weighted point estimates from the non-svy commands.  Despite their 
similarities, the svy commands and corresponding non-svy commands have a 
number of differences."

It then lists 9 differences.  For your case, I think the most critical 
points may be #2 (degrees of freedom used to test significance of 
coefficients) and #4 (different multipliers are used for the variance 
estimator which results in slightly different standard errors.)


-------------------------------------------
Richard Williams, Notre Dame Dept of Sociology
OFFICE: (574)631-6668, (574)631-6463
FAX:    (574)288-4373
HOME:   (574)289-5227
EMAIL:  Richard.A.Williams.5@ND.Edu
WWW (personal):    http://www.nd.edu/~rwilliam
WWW (department):    http://www.nd.edu/~soc

*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/