Hi all
I have around 50 cross countries sample with various level of datasets and
trying to estimates one model using ivreg2 command. My model is something
like this (all in log) and I put some restriction to capture only quality
variable included
ivreg2 lngdp (lnfdi=lnpat lnusdiff lnopenk lnsave) lnip lnrnd if iprs>2.3 &
rnd>0.01,ffirst small orthog(lnip lnrnd)
My main attention is to identified the quality impact of IP and rnd posses
by certain countries (i.e based on restriction made in the equation). My
question is by looking at the output that I have, is it valid to say that my
model is sufficient and acceptable?
I need comment on this matter.
Thanks
My output shown below:
Sargan statistic (overidentification test of all instruments): 5.012
Chi-sq(3) P-val = 0.17095
And the tested instrument variable produced the output as below:
Sargan statistic for unrestricted equation: 4.678
Chi-sq(1) P-val = 0.03055
C statistic (exogeneity/orthogonality of specified instruments):0.333
Chi-sq(2) P-val = 0.84645
By invoking all possible test for ivhettest, the output as below:
. ivhettest,all
IV heteroskedasticity test(s) using levels of IVs only
Ho: Disturbance is homoskedastic
Pagan-Hall general test statistic : 7.546 Chi-sq(6) P-value = 0.2733
Pagan-Hall test w/assumed normality : 8.008 Chi-sq(6) P-value = 0.2375
White/Koenker nR2 test statistic : 10.138 Chi-sq(6) P-value = 0.1190
Breusch-Pagan/Godfrey/Cook-Weisberg : 15.419 Chi-sq(6) P-value = 0.0172
And invoking for overid test;
Tests of overidentifying restrictions:
Sargan N*R-sq test 5.012 Chi-sq(3) P-value = 0.1709
Sargan (N-L)*R-sq test 4.510 Chi-sq(3) P-value = 0.2114
Basmann test 4.727 Chi-sq(3) P-value = 0.1929
Sargan pseudo-F test 1.503 F(3,36) P-value = 0.2301
Basmann pseudo-F test 1.576 F(3,33) P-value = 0.2138
Test for ivendog produced output as below:
ivendog lnfdi
Tests of endogeneity of: lnfdi
H0: Regressor is exogenous
Wu-Hausman F test: 9.22515 F(1,35) P-value = 0.00449
Durbin-Wu-Hausman chi-sq test:8.34380 Chi-sq(1) P-value = 0.00387
*
* 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/
