At 08:11 PM 11/30/2008, Clive Nicholas wrote:
The coefficients are not the same because the models are not the same.
After
. webuse auto
. tab rep78, gen(r)
Compare
. reg mpg weight r1-r5, nocon
with
. reg mpg weight r1-r4
and notice that not even the model diagnostics are the same, since the
first model contains 1 degree of freedom more than the second model.
Actually, I believe the models are the same but they are
parameterized differently. The Anova table and R^2 get zapped when
you say -nocons- because you haven't told regress that you have
included the constant yourself. Instead try
reg mpg weight r1-r5, nocon hascons
and you'll see that the Anova table etc. is identical to what you get with
reg mpg weight r1-r4
-------------------------------------------
Richard Williams, Notre Dame Dept of Sociology
OFFICE: (574)631-6668, (574)631-6463
HOME: (574)289-5227
EMAIL: [email protected]
WWW: http://www.nd.edu/~rwilliam
*
* 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/
