I am using Stata 16.1. When I estimate spatial fixed-effects models or spatial random-effects models using spxtregress, 'z' and 'P>|z|' are not reported for sigma_e for the fixed effects and for sigma_u and sigma_e for random effects, even though '[95% Conf. Interval]' is reported. Could we consider z as coef. / std. err.?
For example, please see the results below.
Code:
. spxtregress lchika Frnk i.year, fe dvarlag(Wid500m) errorlag(Wid500m) ivarlag(Wid500m: Frnk) (3796 observations) (3796 observations used) (data contain 949 panels (places) ) (weighting matrix defines 949 places) Performing grid search ... finished Optimizing concentrated log likelihood: Iteration 0: log likelihood = 3018.3537 Iteration 1: log likelihood = 3019.1078 Iteration 2: log likelihood = 3019.1092 Iteration 3: log likelihood = 3019.1092 Optimizing unconcentrated log likelihood: Iteration 0: log likelihood = 3019.1092 Iteration 1: log likelihood = 3019.1092 (backed up) Fixed-effects spatial regression Number of obs = 3,796 Group variable: _ID Number of groups = 949 Obs per group = 4 Wald chi2(6) = 1567.25 Prob > chi2 = 0.0000 Log likelihood = 3019.1092 Pseudo R2 = 0.0000 ------------------------------------------------------------------------------ lchika | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- lchika | Frnk | -.008818 .0037451 -2.35 0.019 -.0161583 -.0014776 | year | 2008 | .0621063 .0041852 14.84 0.000 .0539035 .0703091 2013 | -.0907423 .0041995 -21.61 0.000 -.0989733 -.0825114 2018 | -.0150216 .0041734 -3.60 0.000 -.0232014 -.0068419 -------------+---------------------------------------------------------------- Wid500m | Frnk | -.0143443 .0070367 -2.04 0.041 -.0281358 -.0005527 lchika | .238697 .0311096 7.67 0.000 .1777233 .2996708 e.lchika | .3770086 .0288928 13.05 0.000 .3203798 .4336374 -------------+---------------------------------------------------------------- /sigma_e | .0814047 .0011026 .079272 .0835947 ------------------------------------------------------------------------------ Wald test of spatial terms: chi2(3) = 544.92 Prob > chi2 = 0.0000 . spxtregress lchika Frnk i.year, re dvarlag(Wid500m) errorlag(Wid500m) ivarlag(Wid500m: Frnk) (3796 observations) (3796 observations used) (data contain 949 panels (places) ) (weighting matrix defines 949 places) Fitting starting values: Iteration 0: log likelihood = 3018.3537 Iteration 1: log likelihood = 3019.1078 Iteration 2: log likelihood = 3019.1092 Iteration 3: log likelihood = 3019.1092 Optimizing concentrated log likelihood: initial: log likelihood = 624.27272 improve: log likelihood = 624.27272 rescale: log likelihood = 624.27272 rescale eq: log likelihood = 1370.7432 Iteration 0: log likelihood = 1370.7432 Iteration 1: log likelihood = 1434.0376 Iteration 2: log likelihood = 1437.9348 Iteration 3: log likelihood = 1437.9941 Iteration 4: log likelihood = 1437.9942 Optimizing unconcentrated log likelihood: Iteration 0: log likelihood = 1437.9942 Iteration 1: log likelihood = 1437.9942 (backed up) Random-effects spatial regression Number of obs = 3,796 Group variable: _ID Number of groups = 949 Obs per group = 4 Wald chi2(6) = 1399.21 Prob > chi2 = 0.0000 Log likelihood = 1437.9942 Pseudo R2 = 0.0052 ------------------------------------------------------------------------------ lchika | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- lchika | Frnk | .0006745 .0038342 0.18 0.860 -.0068404 .0081894 | year | 2008 | .0637408 .0042745 14.91 0.000 .0553629 .0721188 2013 | -.0943856 .0042683 -22.11 0.000 -.1027513 -.0860199 2018 | -.0167213 .0042881 -3.90 0.000 -.0251258 -.0083168 | _cons | 12.5322 .0252436 496.45 0.000 12.48272 12.58167 -------------+---------------------------------------------------------------- Wid500m | Frnk | -.0039038 .007192 -0.54 0.587 -.0179999 .0101922 lchika | .005392 .0036076 1.49 0.135 -.0016787 .0124627 e.lchika | .4980093 .0195698 25.45 0.000 .4596532 .5363653 -------------+---------------------------------------------------------------- /sigma_u | .6189578 .0144549 .5912652 .6479474 /sigma_e | .0814292 .0011169 .0792693 .0836481 ------------------------------------------------------------------------------ Wald test of spatial terms: chi2(3) = 674.49 Prob > chi2 = 0.0000
Thank you.
0 Response to spxtregress: missing 'z' and 'P>|z|' for sigma_u and sigma_e.
Post a Comment