Re: st: stata and weighting

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

At 07:59 AM 3/11/2004 -0500, Nick Winter wrote:
> More generally, my sense is that some other statistics packages 
> (especially SPSS; I can't speak for others), don't distinguish between 
> types of weights, and therefore make it very difficult to even know what, 
> exactly, you are getting.  Stata makes you think about what you really 
> want your weights to do, which IMHO is a feature.
Yes, I would say that what SPSS does is the equivalent of 
iweights.  Whoever provides the weights may have computed them in such a 
way that they become the equivalent of aweights.  Or, you have to rescale 
the weights yourself to make them aweights.  I was struck by the User's 
Guide discussion of how aweights have historically been misused, because 
what it described is what I've basically done with SPSS for years.

I think it is a little odd that SPSS isn't more like Stata in this 
respect.  Rescaling the weights so that weighted N = real N is pretty basic 
but SPSS makes you do the computations yourself.  And, SPSS has no built-in 
feature for robust standard errors (basically aweights + robust 
s.e.s).  SPSS seems to think that if you are weighting it must be because 
you are analyzing data that has been collapsed into a table.  Big edge to 
Stata on the handling of weights.

More generally, Stata often surprises me with the output it does and does 
not provide.  It often does seem inexplicable at first.  But usually, there 
is a good reason for it, albeit not necessarily one that I completely 
understand.


-------------------------------------------
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/