[Limdep Nlogit List] Chi2 tests in SEMI;... $
Jadwiga Kostrzewska
dwakotki at wp.pl
Wed Sep 30 18:06:19 EST 2009
Hello Everybody,
In the semiparamtric binary choice model by Kleins and Spadys method:
SEMI ; Lhs=SELEKCJA; Rhs=W; Kernel=Normal$
I have an output like this:
+---------------------------------------------+
| Semiparametric Binary Choice Model |
| Maximum Likelihood Estimates |
| Dependent variable SELEKCJA |
| Weighting variable None |
| Number of observations 12314 |
| Iterations completed 29 |
| Log likelihood function -7753.998 |
| Number of parameters 8 |
| Info. Criterion: AIC = 1.26068 |
| Finite Sample: AIC = 1.26068 |
| Info. Criterion: BIC = 1.26550 |
| Info. Criterion:HQIC = 1.26229 |
| Restricted log likelihood -8520.398 |
| McFadden Pseudo R-squared .0899489 |
| Chi squared 1532.800 |
| Degrees of freedom 8 |
| Prob[ChiSqd > value] = .0000000 |
| Model estimated: Sep 30, 2009, 07:01:51AM |
| Hosmer-Lemeshow chi-squared = ********* |
| P-value= .00000 with deg.fr. = 8 |
| Normal kernel fn. Bandwidth = .20810 |
+---------------------------------------------+
I do not know what does Chi2 statistic test:
1.
| Chi squared 1532.800 |
| Degrees of freedom 8 |
| Prob[ChiSqd > value] = .0000000 |
and:
2.
| Hosmer-Lemeshow chi-squared = ********* |
| P-value= .00000 with deg.fr. = 8 |
Does anybody know?
I would like to test the probit model vs. the semi-parametric model.
Any ideas are welcome.
One more question: can I add ; Start=... to give my own start values for
SEMI?
Thanks,
Jadwiga Kostrzewska
----------------------------------------------------
Dołącz się do Manify!
Wyjdźmy na ulice, w WALCE O TAŃSZE PIWO!
http://klik.wp.pl/?adr=www.zatanisci.pl&sid=856
More information about the Limdep
mailing list