st: outreg help

[Scott Cunningham <scunning@gmail.com>]

I am using bootstrapping to correct for overdispersion.  After  
calculating these SE, I want to export the table into a comma  
delimited file using -outreg- so that I can work with it in Excel.  I  
get this error, though:

. bootstrap "xi:xtpoisson rp lsr age2 hgc hhd1 i.year $xextra, fe i 
(id)" "_b[ls
> r] _b[age2] _b[hhd1]", cluster(id) reps(500) level(95)

command:      xi:xtpoisson rp lsr age2 hgc hhd1 i.year , fe i(id)
statistics:   _bs_1      = _b[lsr]
              _bs_2      = _b[age2]
              _bs_3      = _b[hhd1]

------------------------------------------------------------------------ 
------
Variable     |  Reps  Observed      Bias  Std. Err. [95% Conf. Interval]
------------- 
+----------------------------------------------------------------
       _bs_1 |   500 -1.280915 -.0134737  .6760543  -2.609178   . 
0473491   (N)
             |                                      -2.584038  -. 
0551219   (P)
             |                                      -2.584038  -. 
0551219  (BC)
       _bs_2 |   500 -.0263555 -.0000388  .0110487  -.0480632  -. 
0046477   (N)
             |                                      -.0489532   -. 
006063   (P)
             |                                      -.0489532   -. 
006063  (BC)
       _bs_3 |   500 -.4402171  .0338045    .28297  -.9961766   . 
1157423   (N)
             |                                      -.9619598   . 
1259881   (P)
             |                                      -.9842739    . 
057382  (BC)
------------------------------------------------------------------------ 
------
Note:  N   = normal
       P   = percentile
       BC  = bias-corrected

. outreg using bm.out, nocons coefastr se replace
variable _bs_1 not found
r(111);

end of do-file


The problem appears to be that it is renaming the variables _bs_1,  
and so -outreg- cannot find these.  How do I get around this so that  
outreg will run smoothly?

Also, for the econometricians - can anyone give me the rule of thumb  
for the number of replications to aim for when calculating empirical  
standard errors by bootstrapping?  Is 500 replications excessive?  It  
takes a long time, and I've got several of these to do, so if I can  
get away with fewer, I'd prefer that.

thanks,


scott cunningham
dept. of economics
university of georgia
athens, ga
*
*   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/