st: bootstrap multiple equations?

["Frakt, Austin" <Austin.Frakt@va.gov>]

I'd like to bootstrap the coefficient standard errors for N different models (they're -probits-, but that's not relevant). I could execute N separate -bootstrap- commands, but it would be more efficient to just -bootstrap- once and, somehow, tell -bootstrap- to gather up all the results from all N models. Is that possible? What's the most efficient way?

I tried to implement this by defining an -eclass- program, gathering up all the coefficients from each of the N models into one vector (call it "big_b" and (e)returning that vector. In this case, the -bootstrap- command I tried looked something like:

bootstrap big_b = e(big_b), reps(200) seed(123): bsprog

But I get type mismatch errors. Is this even possible? Am I going about it the right way?

Thanks in advance for your ideas.

-Austin Frakt

*
*   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/