Thanks for the comments,
Here something about the model:
There is no way to decrease the number of dummy variables
by grouping similar ones, since each of them are unique.
The model is a replication of Lichtenberg (1996) model, he
has more than 400 right hand side dummy variables.
Second, for each category, one of the dummy variable has already
been dropped to prevent collinearity problem. Third,
I've used areg if I do not have to impose constraints, but
with constarints areg comment is not running. Fourth, without the
constraint the regression is running okey and multicollinearity
test fails, no multicollinearity.
After receiving comments, there are two things I need to know:
First: Is there any restriction on the right hand side variables
if you need to put some restriction in model? Becasue, I have
tried the same model for a different data set, where none of the
right hand side variables are dummy (75 right hand site variable),
I got the same message.
Second:
gen x1x2=x1 x2 and then replace x1 and x2 with the new x1x2
and the regression. Is it okey or is it meaningful to add
two dummy variable of the same group? I know that it is fine
if they are not dummy variables.
if the asnwer to this question is pozitive, then I have
already solved my problem. if not, please advise me...
Thanks for everybody who spent their valuable time to
help me figuring out the problem...
Zeynal Karaca
Texas A&M University
Department of Economics
-------- Original Message --------
==> From: Richard Williams <[email protected]>
==> Date: Wed, 10 Aug 2005 21:46:45 -0500
At 09:19 PM 8/10/2005 -0500, Scott Merryman wrote:
>This problem is easy enough to replicate: > >. sysuse auto >(1978
Automobile Data) > >. qui tab fore, gen(Df) > >. const define 1 mpg
= gear > >. cnsreg price mpg gear Df*, con(1) >constraints
insufficient for unique estimate >redundant or inconsistent
constraints >r(412);
Good. And, if it is not obvious to you where the problem is,
instead of cnsreg try
.. gen mpggear = mpg gear
.. reg price mpggear Df*
Source | SS df MS Number of
obs = 74 ------------- ------------------------------
F( 2, 71) = 14.98 Model | 188463446 2
94231723 Prob > F = 0.0000 Residual | 446601950
71 6290168.31 R-squared = 0.2968 ------------- ------
------------------------ Adj R-squared = 0.2770 Total |
635065396 73 8699525.97 Root MSE = 2508
---------------------------------------------------------------------
--------- price | Coef. Std. Err. t P>|t| [95%
Conf. Interval] ------------- ---------------------------------------
------------------------- mpggear | -291.2987 53.43181 -5.45
0.000 -397.8387 -184.7587 Df1 | (dropped) Df2 | 1957.088
705.6254 2.77 0.007 550.1107 3364.065 _cons |
12665.52 1258.366 10.07 0.000 10156.41 15174.63 -------
---------------------------------------------------------------------
--
Notice that Df1 got dropped.
------------------------------------------- Richard Williams, Notre
Dame Dept of Sociology OFFICE: (574)631-6668, (574)631-6463 FAX:
(574)288-4373 HOME: (574)289-5227 EMAIL:
[email protected] WWW (personal):
http://www.nd.edu/~rwilliam WWW (department):
http://www.nd.edu/~soc
* * 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/
*
* 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/
