Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
| From | "Henrique Neder" <hdneder@ufu.br> |
| To | <statalist@hsphsun2.harvard.edu> |
| Subject | st: RES: relevance of instruments |
| Date | Wed, 2 May 2012 00:22:41 -0300 |
I think that you have to do a more detailed inspection of the problem. See Stock and Yogo critical values approach in the help for ivreg2 (Tests of under- and weak identification). Henrique Neder -----Mensagem original----- De: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] Em nome de donsaane dontsi Enviada em: Tuesday, May 01, 2012 6:28 PM Para: statalist@hsphsun2.harvard.edu Assunto: st: relevance of instruments Hi, I run an ivreg2 regression , and I want to know if my instruments are strong. Here is the ouptut below: My question is: Is my instruments are strong because F=16.46 >10? Or there are not because Partial R2 are small? thanks . ivreg2 emploi c30a d30a d30b e30 g30a j30a j30c j30e l30a l30b a2 a4a a6a a7 b2a b2b b2c b5 (k30= fin1k21 fink21), ffirst robust First-stage regressions ----------------------- Summary results for first-stage regressions: Shea Variable Partial R2 Partial R2 F( 2, 254) P-value k30 0.0149 0.0149 16.46 0.0000 NB: first-stage F-stat heteroskedasticity-robust IV (2SLS) regression with robust standard errors ------------------------------------------------ Number of obs = 275 F( 19, 255) = 0.81 Prob > F = 0.7000 Total (centered) SS = 82.96736373 Centered R2 = -0.2291 Total (uncentered) SS = 88.98325588 Uncentered R2 = -0.1460 Residual SS = 101.976442 Root MSE = .61 ---------------------------------------------------------------------------- -- | Robust emploi | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------- -------------+------ k30 | -.3825684 .2103571 -1.82 0.069 -.7948607 .0297239 c30a | -.0519788 .041202 -1.26 0.207 -.1327332 .0287757 d30a | .108799 .0397846 2.73 0.006 .0308226 .1867754 d30b | .0031931 .0348354 0.09 0.927 -.0650831 .0714693 e30 | -.036257 .0293068 -1.24 0.216 -.0936972 .0211832 g30a | .1152195 .058208 1.98 0.048 .0011339 .2293052 j30a | .0429871 .0402775 1.07 0.286 -.0359552 .1219295 j30c | .0356038 .0411333 0.87 0.387 -.0450161 .1162236 j30e | -.0100979 .027043 -0.37 0.709 -.0631013 .0429054 l30a | -.0212033 .0367756 -0.58 0.564 -.0932822 .0508756 l30b | .0310033 .0326002 0.95 0.342 -.0328918 .0948985 a2 | -.0603592 .0425782 -1.42 0.156 -.143811 .0230925 a4a | -.0009478 .0028646 -0.33 0.741 -.0065624 .0046668 a6a | .0933522 .0763294 1.22 0.221 -.0562506 .242955 a7 | .1954874 .1460382 1.34 0.181 -.0907422 .4817169 b2a | .0040452 .0023674 1.71 0.087 -.0005947 .0086851 b2b | .0011903 .002382 0.50 0.617 -.0034784 .0058591 b2c | -.0013251 .006618 -0.20 0.841 -.0142962 .0116459 b5 | -.0002381 .0035678 -0.07 0.947 -.007231 .0067547 _cons | .5296318 7.086394 0.07 0.940 -13.35945 14.41871 ---------------------------------------------------------------------------- -- Hansen J statistic (overidentification test of all instruments): 0.418 Chi-sq(1) P-val = 0.51788 ---------------------------------------------------------------------------- -- Instrumented: k30 Instruments: fin1k21 fink21 c30a d30a d30b e30 g30a j30a j30c j30e l30a l30b a2 a4a a6a a7 b2a b2b b2c b5 ---------------------------------------------------------------------------- -- donsaane * * 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/ * * 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/