Re: [Gretl-users] Instrumental variables and Hausman test

Hello everybody, I am trying to perform the Hausman test within a instrumental variables context. Using the smoke.gdt data from the Wooldridge's book, I regress the variable called lincome on a constant, cigs, educ, age and agesq. In order to do this I apply "instrumental variables->2sls", providing one instrument (in particular the variable called restaurn) for cigs, a variable that we suspect that could be correlated with the perturbation. As it can be seen below, the asymptotic test statistic is 5.00976. However, if I do it manually (or using another program like Stata), that value is totally different, getting 2.76 aproximately (the square of the estimates differences, (0.00173057-(-0.0413845))^2, over the variances differences, 0.0260279^2-0.00171372^2). What am I doing wrong? It could be a Gretl bug? Thanks a lot in advance. Javi Model 1: OLS, using observations 1-807Dependent variable: lincome              coefficient    std. error    t-ratio    p-value   -----------------------------------------------------------  const       7.79544       0.170427      45.74     1.05e-225 ***  cigs        0.00173057    0.00171372     1.010    0.3129     educ        0.0603605     0.00789834     7.642    6.10e-014 ***  age         0.0576907     0.00764359     7.548    1.21e-013 ***  agesq      −0.000630589   8.33822e-05   −7.563    1.08e-013 *** Mean dependent var   9.687315   S.D. dependent var   0.712695Sum squared resid    341.8539   S.E. of regression   0.652880R-squared            0.164978   Adjusted R-squared   0.160813F(4, 802)            39.61341   P-value(F)           2.68e-30Log-likelihood      −798.5010   Akaike criterion     1607.002Schwarz criterion    1630.469   Hannan-Quinn         1616.013   Model 2: TSLS, using observations 1-807Dependent variable: lincomeInstrumented: cigs Instruments: const restaurn educ age agesq              coefficient   std. error      z       p-value   ---------------------------------------------------------  const       7.78114      0.228129      34.11    5.52e-255 ***  cigs       −0.0413845    0.0260279     −1.590   0.1118     educ        0.0400241    0.0161607      2.477   0.0133    **  age         0.0932076    0.0236787      3.936   8.27e-05  ***  agesq      −0.00104373   0.000272324   −3.833   0.0001    *** Mean dependent var   9.687315   S.D. dependent var   0.712695Sum squared resid    611.6565   S.E. of regression   0.873306R-squared            0.024886   Adjusted R-squared   0.020022F(4, 802)            22.62942   P-value(F)           9.92e-18Log-likelihood      −7885.899   Akaike criterion     15781.80Schwarz criterion    15805.26   Hannan-Quinn         15790.81 Log-likelihood for income = −8614.4 Hausman test -  Null hypothesis: OLS estimates are consistent  Asymptotic test statistic: Chi-square(1) = 5.00976  with p-value = 0.0252048     JAVIER GARCÍA   Departamento de Economía Aplicada III (Econometría y Estadística) Facultad de Economía y Empresa (Sección Sarriko) Avda. Lehendakari Aguirre 83 48015 BILBAO T.: +34 601 7126 F.: +34 601 3754 www.ehu.es ERNE! Baliteke mezu honen zatiren bat edo mezu osoa legez babestuta egotea. Mezuak badu bere hartzailea. Okerreko helbidera heldu bada (helbidea gaizki idatzi, transmisioak huts egin) eman abisu igorleari, korreo honi erantzunda. Kontuz! Mezua ez bada zuretzat, ez erabili, ez zabaldu beste inori, ez kopiatu eta ez baliatu. ¡ATENCIÓN! Este mensaje contiene información privilegiada o confidencial a la que sólo tiene derecho a acceder el destinatario. Si usted lo recibe por error le agradeceríamos que no hiciera uso de la información y que se pusiese en contacto con el remitente. E-mail hau inprimatu baino lehen egiaztatu inprimatzeko beharra. Antes de imprimir este e-mail piense bien si es necesario hacerlo.     _______________________________________________ Gretl-users mailing list Gretl-users(a)lists.wfu.edu http://lists.wfu.edu/mailman/listinfo/gretl-users