[Limdep Nlogit List] Cox's test
ecptq at bath.ac.uk
ecptq at bath.ac.uk
Wed Apr 4 07:32:10 EST 2007
Hello,
I am doing the Cox's test for nonnest models. I obtained the results as
follows:
namelist;z=one,edu,experien,expsq,wpt;x=one,edu,experien,expsq,wpq$
calc;sz=ess(z,lrwage)/n;sx=ess(x,lrwage)/n$
regress;lhs=lrwage;rhs=x;keep=xb$
regress;lhs=xb;rhs=z;res=u$
calc;szx=sx+sumsqdev/n;c12=n/2*log(sz/szx);v12=(sx/szx^2)*ess(x,u);
list;coxtest=1-phi(c12/sqr(v12))$
?COXTEST = .10000000000000000D+01
namelist;z=one,edu,experien,expsq,wpq;x=one,edu,experien,expsq,wpt$
calc;sz=ess(z,lrwage)/n;sx=ess(x,lrwage)/n$
regress;lhs=lrwage;rhs=x;keep=xb$
regress;lhs=xb;rhs=z;res=u$
calc;szx=sx+sumsqdev/n;c12=n/2*log(sz/szx);v12=(sx/szx^2)*ess(x,u);
list;coxtest=1-phi(c12/sqr(v12))$
?COXTEST = .14594936348616890D+00
I have checked the Limdep manual, and get the information of 'The
hypothesis that X is correct and Z is not is tested by this'. However I
am still not sure how to identify the upper results.
Any help would be much appreciated.
Thanks a lot.
--
Miss Tian Qiu
Department of Economics and International Development
University of Bath
Tel.: (01225) 384864
More information about the Limdep
mailing list