Re: st: Handling big samples

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

--- On Tue, 22/2/11, miguelvg81@yahoo.com wrote:
> Hi. I'm running a regression in stata
> with a big sample. All the estimates turn to be significant,
> but how can I be sure this is due to the test and not to the
> sample size?? Does anyone know where can I find papers that
> talk about this special issue. 

In principle this is not a problem. We already knew before you
ran your regression that all variables will have an effect on
your dependent variable (however small that effect may be), so
we know a priori that the null hypothesis that any of these 
effects is 0 is false. The fact that we sometimes cannot reject
this hypothesis just means that our dataset is not large enough
to detect that effect. So if we get a large dataset, we should
be able to detect all the effects. Done.

well, not quite done: We do not only care whether or not a 
hypothesis that we know to be wrong is wrong or not (hugh, do 
we really?), but also whether the effect is big enough to be
substanitvely/clinicly (depending on your discipline) relevent.
That is actually surprisingly easy, just interpret the 
coefficients, and think about whether or not you think that that
is a big effect. Now you are really done (except for writing your
paper, submitting it, getting over the rejection, submitting it
somewhere else, etc. etc.).

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/