st: Re: regression simulation

[Kit Baum <baum@bc.edu>]

Certainly, the simulate command could be used here, and for more 
complicated simulations it is essential. I wanted to illustrate how one 
could do what was requested from first principles, which in this case 
does not involve much overhead vis-a-vis simulate: only an 
understanding of how the 'post' command works, which is a good thing to 
acquire IMHO.

Kit Baum, Boston College Economics
http://ideas.repec.org/e/pba1.html

On Dec 21, 2004, at 2:33 AM, maartenbuis wrote:


> One way to make Kit's code more elegant is to use the -simulation-
> command:
>
> program define sim, rclass
> drop _all
> set obs 100
> drawnorm y x
> reg y x
> return scalar se = _se[x]
> return scalar t  = _b[x]/_se[x]
> end
>
> simulate "sim" se=r(se) t=r(t), dots reps(1000)
> desc
> sum
*
*   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/