Re: st: sample size

[Ron�n Conroy <rconroy@rcsi.ie>]

Anuja wrote:

> Hi Statalist,
>
> I have merged two datasets, one with a sample size of
> 10,696 observations and the other with n<1000.
>
> However when I run a regression, the variables are not
> significant coz of the sample size. 
>  
Do you mean that there are variables in the large dataset which predict 
the outcome you are interested in but the same variables do not predict 
outcome in the smaller set?

There are many things you don't tell us. Are the effect sizes 
(regression coefficients) similar in both samples, but the confidence 
intervals wider in the small sample (in which case the lack of 
statistical significance could well be becoz of the sample size) or are 
the effect sizes different also.

And the two samples, are they from the same or different populations? 
Why combine them?


> Should I run two seperate regressions? or is there a
> way I link both of them, as the dep var is the same.
>
>
>  
I think we need more detail before any practical suggestions can emerge.

> Many thanks,
> Anuja 
> =====
> WE ARE ALL IN THE GUTTER,SOME OF US LOOK AT THE STARS.
>
>
>
> 		
> __________________________________
> Do you Yahoo!?
> Yahoo! Mail Address AutoComplete - You start. We finish.
> http://promotions.yahoo.com/new_mail 
> *
> *   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/
>
>
>  

--

Ronan M Conroy (rconroy@rcsi.ie) 
Senior Lecturer in Biostatistics 
Royal College of Surgeons 
Dublin 2, Ireland 
+353 1 402 2431 (fax 2764) 
-------------------- 
Just say no to drug reps 
http://www.nofreelunch.org/

*
*   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/