st: degrees of freedom adjustment in sureg

[Kit Baum <baum@bc.edu>]

sacrificial lamb	

Brian wrote

The "small" option only affects whether z and chi2 statistics on one  
hand
and t and F statistics on the other are displayed.  To replicate the
results from -regress- using -sureg-, you must specify the "dfk" or  
"dfk2"
options along with "small":

    . sysuse auto
    . regress price mpg headroom trunk
    . estimates store reg
    . sureg (price mpg headroom trunk), small dfk
    . estimates table reg ., se equations(1)

    ----------------------------------------
        Variable |    reg         active
    -------------+--------------------------
             mpg | -224.35974   -224.35974
                 |  65.275114    65.275114
        headroom | -659.46302   -659.46302
                 |  484.51008    484.51008
           trunk |  126.60486    126.60486
                 |  107.23994    107.23994
           _cons |  11175.774    11175.774
                 |  2431.1345    2431.1345
    ----------------------------------------
                                legend: b/se


When there is a single equation, the "dfk" and "dfk2" options result in
the same degree-of-freedom adjustment, but in general that need not  
be the
case.  Since one is not clearly better than the other, you must specify
which one you want.



Thanks for this clarification. IMHO using the small option should  
either default to one of these options or require that one of them be  
given as well. I think it is very confusing relative to other Stata  
commands to have a -small- option that alters the distribution used  
for evaluation without  recalculating the covariance matrix.

Kit Baum, Boston College Economics and DIW Berlin
http://ideas.repec.org/e/pba1.html
An Introduction to Modern Econometrics Using Stata:
http://www.stata-press.com/books/imeus.html


*
*   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/