------------------------------------------------------------------------------------------------------------ log: d:\panel_msc\mortality.log log type: text . set more off . . clear . use c:\bofe\mortality.dta . . set more off . . su * Variable | Obs Mean Std. Dev. Min Max -------------+-------------------------------------------------------- bord | 7211 2.379004 1.430264 1 11 chld1 | 7211 .3314381 .4707628 0 1 chld2 | 7211 .2872001 .4524871 0 1 chld3 | 7211 .1951186 .3963194 0 1 chld4 | 7211 .1038691 .3051118 0 1 -------------+-------------------------------------------------------- seqid | 7211 49040.27 718.5352 47838 50311 female | 7211 .4562474 .4981166 0 1 mult | 7211 .0165026 .1274068 0 1 agemay | 7211 23.46124 4.239238 13 44 agemays | 7211 568.3984 212.8897 169 1936 -------------+-------------------------------------------------------- chldm4 | 7211 .0804327 .2719807 0 1 dum59 | 7211 .3895438 .4876806 0 1 dum6069 | 7211 .4145056 .4926707 0 1 educpa1 | 7211 .4063237 .4911804 0 1 educpa2 | 7211 .219387 .41386 0 1 -------------+-------------------------------------------------------- educma1 | 7211 .6523367 .4762615 0 1 educma2 | 7211 .1234225 .3289444 0 1 ethnic | 7211 .4902233 .4999391 0 1 hindu | 7211 .4308695 .4952323 0 1 sikh | 7211 .5150465 .4998082 0 1 -------------+-------------------------------------------------------- mort | 7211 .0594924 .2365605 0 1 mort_1 | 7211 .048537 .2149128 0 1 . tab bord birth order | number | Freq. Percent Cum. ------------+----------------------------------- 1 | 2,390 33.14 33.14 2 | 2,071 28.72 61.86 3 | 1,407 19.51 81.38 4 | 749 10.39 91.76 5 | 335 4.65 96.41 6 | 158 2.19 98.60 7 | 57 0.79 99.39 8 | 22 0.31 99.69 9 | 14 0.19 99.89 10 | 6 0.08 99.97 11 | 2 0.03 100.00 ------------+----------------------------------- Total | 7,211 100.00 . tab mort mort_1 if(bord>1), row col +-------------------+ | Key | |-------------------| | frequency | | row percentage | | column percentage | +-------------------+ | mort_1 mort | 0 1 | Total -----------+----------------------+---------- 0 | 4,224 300 | 4,524 | 93.37 6.63 | 100.00 | 94.48 85.71 | 93.84 -----------+----------------------+---------- 1 | 247 50 | 297 | 83.16 16.84 | 100.00 | 5.52 14.29 | 6.16 -----------+----------------------+---------- Total | 4,471 350 | 4,821 | 92.74 7.26 | 100.00 | 100.00 100.00 | 100.00 . . label var mort "1 if child died in infancy" . label var mort_1 "1 if previous died in infancy" . label var female "1 if index child is a female" . label var bord "birth order of index child" . label var mult "index child part of a multiple birth" . label var agemay "mother's age at birth of index child in years" . label var agemays "agemay squared" . label var ethnic "scheduled caste group" . label var dum59 "mother born before 1959 or earlier" . label var dum6069 "mother born during 1960-69" . label var educpa1 "father's education primary or less" . label var educpa2 "father's educ - incomplete secondary" . label var educma1 "mother's edu - primary or less" . label var educma2 "mother's educ - incomeplete secondary" . . . *base - child is 5th or more . *note some of these models should perhaps be only estimated on the . *sample of children born after the first one. . . *static model . *============ . probit mort female dum59 dum6069 /* > */ ethnic hindu sikh educpa1 educpa2 educma1 educma2 agemay agemays /* > */ chld2 chld3 chld4 Iteration 0: log likelihood = -1626.5757 Iteration 1: log likelihood = -1598.2543 Iteration 2: log likelihood = -1597.9002 Iteration 3: log likelihood = -1597.9 Probit regression Number of obs = 7211 LR chi2(15) = 57.35 Prob > chi2 = 0.0000 Log likelihood = -1597.9 Pseudo R2 = 0.0176 ------------------------------------------------------------------------------ mort | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- female | .1239593 .0476721 2.60 0.009 .0305237 .2173948 dum59 | .0721869 .07052 1.02 0.306 -.0660299 .2104036 dum6069 | .0278693 .0671887 0.41 0.678 -.1038182 .1595567 ethnic | .1002422 .052293 1.92 0.055 -.0022502 .2027346 hindu | -.0355094 .1036118 -0.34 0.732 -.2385848 .1675659 sikh | -.0612804 .1027368 -0.60 0.551 -.2626408 .14008 educpa1 | .0649727 .068435 0.95 0.342 -.0691574 .1991028 educpa2 | .0401741 .0722548 0.56 0.578 -.1014427 .181791 educma1 | .2104642 .083934 2.51 0.012 .0459565 .3749718 educma2 | .1859635 .0948896 1.96 0.050 -.0000168 .3719437 agemay | -.0684679 .0475242 -1.44 0.150 -.1616135 .0246778 agemays | .0010201 .000948 1.08 0.282 -.0008378 .0028781 chld2 | -.0804653 .0604871 -1.33 0.183 -.1990178 .0380871 chld3 | -.0455016 .0697925 -0.65 0.514 -.1822923 .0912892 chld4 | .1314566 .0827198 1.59 0.112 -.0306712 .2935845 _cons | -.8347384 .5915922 -1.41 0.158 -1.994238 .3247609 ------------------------------------------------------------------------------ . outreg using mortality, replace noaster nolabel bdec(4) . . *RE static model . *=============== . xtprobit mort female dum59 dum6069 /* > */ ethnic hindu sikh educpa1 educpa2 educma1 educma2 agemay agemays /* > */ chld2 chld3 chld4 Fitting comparison model: Iteration 0: log likelihood = -1626.5757 Iteration 1: log likelihood = -1598.2543 Iteration 2: log likelihood = -1597.9002 Iteration 3: log likelihood = -1597.9 Fitting full model: rho = 0.0 log likelihood = -1597.9 rho = 0.1 log likelihood = -1587.1926 rho = 0.2 log likelihood = -1601.2223 Iteration 0: log likelihood = -1587.1926 Iteration 1: log likelihood = -1574.1605 Iteration 2: log likelihood = -1573.3485 Iteration 3: log likelihood = -1573.3445 Iteration 4: log likelihood = -1573.3445 Random-effects probit regression Number of obs = 7211 Group variable (i): seqid Number of groups = 2390 Random effects u_i ~ Gaussian Obs per group: min = 1 avg = 3.0 max = 11 Wald chi2(15) = 49.95 Log likelihood = -1573.3445 Prob > chi2 = 0.0000 ------------------------------------------------------------------------------ mort | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- female | .1321956 .053285 2.48 0.013 .027759 .2366322 dum59 | .0879044 .0854164 1.03 0.303 -.0795088 .2553175 dum6069 | .0287599 .0811052 0.35 0.723 -.1302035 .1877232 ethnic | .1209946 .0647411 1.87 0.062 -.0058956 .2478848 hindu | -.0652123 .1304364 -0.50 0.617 -.3208628 .1904383 sikh | -.0943849 .1293005 -0.73 0.465 -.3478093 .1590395 educpa1 | .0833441 .0847454 0.98 0.325 -.0827538 .2494421 educpa2 | .0567814 .0889409 0.64 0.523 -.1175396 .2311024 educma1 | .2261162 .1013062 2.23 0.026 .0275598 .4246726 educma2 | .1844061 .1148475 1.61 0.108 -.0406908 .409503 agemay | -.0900522 .054105 -1.66 0.096 -.196096 .0159916 agemays | .00136 .0010772 1.26 0.207 -.0007512 .0034712 chld2 | -.0597807 .0663599 -0.90 0.368 -.1898437 .0702824 chld3 | -.0292058 .0766781 -0.38 0.703 -.1794921 .1210804 chld4 | .1359971 .09089 1.50 0.135 -.042144 .3141382 _cons | -.7701613 .6762555 -1.14 0.255 -2.095598 .5552751 -------------+---------------------------------------------------------------- /lnsig2u | -1.315169 .1812711 -1.670454 -.9598846 -------------+---------------------------------------------------------------- sigma_u | .5181012 .0469584 .433776 .6188191 rho | .2116231 .030243 .1583636 .2769013 ------------------------------------------------------------------------------ Likelihood-ratio test of rho=0: chibar2(01) = 49.11 Prob >= chibar2 = 0.000 . outreg using mortality, append noaster nolabel bdec(4) . . . *simple probit including the first borns . *======================================= . probit mort mort_1 female dum59 dum6069 /* > */ ethnic hindu sikh educpa1 educpa2 educma1 educma2 agemay agemays /* > */ chld2 chld3 chld4 Iteration 0: log likelihood = -1626.5757 Iteration 1: log likelihood = -1581.8545 Iteration 2: log likelihood = -1580.7647 Iteration 3: log likelihood = -1580.7639 Probit regression Number of obs = 7211 LR chi2(16) = 91.62 Prob > chi2 = 0.0000 Log likelihood = -1580.7639 Pseudo R2 = 0.0282 ------------------------------------------------------------------------------ mort | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- mort_1 | .5358543 .0877755 6.10 0.000 .3638175 .7078912 female | .1258308 .0479711 2.62 0.009 .0318092 .2198525 dum59 | .0750701 .0708122 1.06 0.289 -.0637192 .2138594 dum6069 | .0258769 .0675313 0.38 0.702 -.1064819 .1582358 ethnic | .0996956 .0526709 1.89 0.058 -.0035375 .2029286 hindu | -.0334211 .1039429 -0.32 0.748 -.2371454 .1703032 sikh | -.0552922 .1030333 -0.54 0.592 -.2572338 .1466493 educpa1 | .0628685 .0688597 0.91 0.361 -.0720941 .1978312 educpa2 | .0381025 .0727172 0.52 0.600 -.1044206 .1806257 educma1 | .1953511 .0844074 2.31 0.021 .0299158 .3607865 educma2 | .1717505 .0955748 1.80 0.072 -.0155728 .3590737 agemay | -.0654349 .0479921 -1.36 0.173 -.1594977 .0286279 agemays | .0009212 .0009584 0.96 0.336 -.0009572 .0027997 chld2 | -.1119094 .0612294 -1.83 0.068 -.2319167 .008098 chld3 | -.075613 .0703904 -1.07 0.283 -.2135757 .0623497 chld4 | .1065805 .0830698 1.28 0.199 -.0562334 .2693943 _cons | -.862572 .5971268 -1.44 0.149 -2.032919 .3077749 ------------------------------------------------------------------------------ . outreg using mortality, append noaster nolabel bdec(4) . . *simple probit excluding the first borns . *======================================= . probit mort mort_1 female dum59 dum6069 /* > */ ethnic hindu sikh educpa1 educpa2 educma1 educma2 agemay agemays /* > */ chld2 chld3 chld4 if(bord>1) Iteration 0: log likelihood = -1115.3981 Iteration 1: log likelihood = -1062.372 Iteration 2: log likelihood = -1060.9729 Iteration 3: log likelihood = -1060.9711 Iteration 4: log likelihood = -1060.9711 Probit regression Number of obs = 4821 LR chi2(16) = 108.85 Prob > chi2 = 0.0000 Log likelihood = -1060.9711 Pseudo R2 = 0.0488 ------------------------------------------------------------------------------ mort | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- mort_1 | .4588258 .0902176 5.09 0.000 .2820025 .6356491 female | .238379 .0588232 4.05 0.000 .1230876 .3536704 dum59 | .1024962 .0927599 1.10 0.269 -.0793099 .2843024 dum6069 | .0440334 .08959 0.49 0.623 -.1315598 .2196265 ethnic | .1599544 .0654114 2.45 0.014 .0317505 .2881584 hindu | .041735 .1242117 0.34 0.737 -.2017154 .2851854 sikh | -.0117687 .1236209 -0.10 0.924 -.2540612 .2305237 educpa1 | .0576099 .085699 0.67 0.501 -.1103571 .2255769 educpa2 | .0601452 .0899655 0.67 0.504 -.116184 .2364744 educma1 | .0220287 .1114297 0.20 0.843 -.1963695 .240427 educma2 | .1902902 .1218651 1.56 0.118 -.0485611 .4291414 agemay | -.1485593 .0629972 -2.36 0.018 -.2720316 -.025087 agemays | .0020255 .0012004 1.69 0.092 -.0003272 .0043783 chld2 | -.5204387 .1106769 -4.70 0.000 -.7373615 -.303516 chld3 | -.4176735 .1044685 -4.00 0.000 -.6224281 -.2129189 chld4 | -.1728693 .1047837 -1.65 0.099 -.3782415 .0325029 _cons | .7939227 .8469491 0.94 0.349 -.866067 2.453912 ------------------------------------------------------------------------------ . outreg using mortality, append noaster nolabel bdec(4) . . . *==================================== . *xtprobit including the first borns . *include first borns but without any allowance for initial conditions. . *remember - start of the process same as sample start . *=================================== . xtprobit mort mort_1 female dum59 dum6069 /* > */ ethnic hindu sikh educpa1 educpa2 educma1 educma2 agemay agemays /* > */ chld2 chld3 chld4 Fitting comparison model: Iteration 0: log likelihood = -1626.5757 Iteration 1: log likelihood = -1581.8545 Iteration 2: log likelihood = -1580.7647 Iteration 3: log likelihood = -1580.7639 Fitting full model: rho = 0.0 log likelihood = -1580.7639 rho = 0.1 log likelihood = -1583.774 Iteration 0: log likelihood = -1580.7638 (not concave) Iteration 1: log likelihood = -1578.92 (not concave) Iteration 2: log likelihood = -1576.4683 Iteration 3: log likelihood = -1572.2645 Iteration 4: log likelihood = -1572.0906 Iteration 5: log likelihood = -1572.0603 Iteration 6: log likelihood = -1572.0603 Random-effects probit regression Number of obs = 7211 Group variable (i): seqid Number of groups = 2390 Random effects u_i ~ Gaussian Obs per group: min = 1 avg = 3.0 max = 11 Wald chi2(16) = 55.11 Log likelihood = -1572.0603 Prob > chi2 = 0.0000 ------------------------------------------------------------------------------ mort | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- mort_1 | .1920864 .12469 1.54 0.123 -.0523016 .4364743 female | .1315647 .0524057 2.51 0.012 .0288514 .234278 dum59 | .0857235 .0827558 1.04 0.300 -.0764749 .2479219 dum6069 | .0280764 .0785686 0.36 0.721 -.1259153 .182068 ethnic | .1168788 .0626523 1.87 0.062 -.0059174 .239675 hindu | -.0585177 .1256573 -0.47 0.641 -.3048015 .187766 sikh | -.0861694 .1246235 -0.69 0.489 -.330427 .1580883 educpa1 | .0785299 .0819109 0.96 0.338 -.0820125 .2390722 educpa2 | .0530147 .0859867 0.62 0.538 -.1155161 .2215455 educma1 | .2167439 .0983601 2.20 0.028 .0239617 .4095262 educma2 | .1790213 .1113154 1.61 0.108 -.0391528 .3971954 agemay | -.0848849 .0533573 -1.59 0.112 -.1894632 .0196934 agemays | .0012554 .0010627 1.18 0.237 -.0008275 .0033382 chld2 | -.0767624 .0664198 -1.16 0.248 -.2069429 .0534181 chld3 | -.0447476 .0762955 -0.59 0.558 -.1942841 .1047889 chld4 | .1254718 .0898144 1.40 0.162 -.0505612 .3015048 _cons | -.7877652 .6639049 -1.19 0.235 -2.088995 .5134645 -------------+---------------------------------------------------------------- /lnsig2u | -1.55075 .3150683 -2.168273 -.9332277 -------------+---------------------------------------------------------------- sigma_u | .460531 .0725494 .3381938 .6271222 rho | .174978 .0454835 .102636 .2822703 ------------------------------------------------------------------------------ Likelihood-ratio test of rho=0: chibar2(01) = 17.41 Prob >= chibar2 = 0.000 . outreg using mortality, append noaster nolabel bdec(4) . . . *xtprobit excluding the first borns . *================================= . xtprobit mort mort_1 female dum59 dum6069 /* > */ ethnic hindu sikh educpa1 educpa2 educma1 educma2 agemay agemays /* > */ chld2 chld3 chld4 if(bord>1) Fitting comparison model: Iteration 0: log likelihood = -1115.3981 Iteration 1: log likelihood = -1062.372 Iteration 2: log likelihood = -1060.9729 Iteration 3: log likelihood = -1060.9711 Iteration 4: log likelihood = -1060.9711 Fitting full model: rho = 0.0 log likelihood = -1060.9711 rho = 0.1 log likelihood = -1064.9103 Iteration 0: log likelihood = -1060.9711 (not concave) Iteration 1: log likelihood = -1060.7517 (not concave) Iteration 2: log likelihood = -1060.3027 Iteration 3: log likelihood = -1059.9065 Iteration 4: log likelihood = -1059.1389 Iteration 5: log likelihood = -1059.1293 Iteration 6: log likelihood = -1059.1293 Random-effects probit regression Number of obs = 4821 Group variable (i): seqid Number of groups = 2071 Random effects u_i ~ Gaussian Obs per group: min = 1 avg = 2.3 max = 10 Wald chi2(16) = 83.43 Log likelihood = -1059.1293 Prob > chi2 = 0.0000 ------------------------------------------------------------------------------ mort | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- mort_1 | .3443479 .1130279 3.05 0.002 .1228173 .5658784 female | .2458574 .0617534 3.98 0.000 .1248229 .3668919 dum59 | .1093165 .0995752 1.10 0.272 -.0858473 .3044803 dum6069 | .0440738 .0958835 0.46 0.646 -.1438543 .232002 ethnic | .1692718 .071045 2.38 0.017 .0300262 .3085174 hindu | .0367093 .1363046 0.27 0.788 -.2304429 .3038615 sikh | -.0245483 .1357939 -0.18 0.857 -.2906995 .241603 educpa1 | .0615159 .0930395 0.66 0.508 -.1208381 .2438698 educpa2 | .0704765 .0975157 0.72 0.470 -.1206508 .2616039 educma1 | .029893 .1195468 0.25 0.803 -.2044145 .2642005 educma2 | .1930362 .1310323 1.47 0.141 -.0637824 .4498547 agemay | -.1655683 .0672387 -2.46 0.014 -.2973538 -.0337828 agemays | .0023018 .0012746 1.81 0.071 -.0001964 .0048 chld2 | -.5036794 .1176575 -4.28 0.000 -.7342839 -.2730748 chld3 | -.4000247 .1103301 -3.63 0.000 -.6162677 -.1837816 chld4 | -.1560018 .1098984 -1.42 0.156 -.3713988 .0593951 _cons | .9377851 .9026194 1.04 0.299 -.8313165 2.706887 -------------+---------------------------------------------------------------- /lnsig2u | -2.311391 .612104 -3.511093 -1.111689 -------------+---------------------------------------------------------------- sigma_u | .3148385 .0963569 .1728128 .5735876 rho | .0901839 .0502236 .0289983 .2475561 ------------------------------------------------------------------------------ Likelihood-ratio test of rho=0: chibar2(01) = 3.68 Prob >= chibar2 = 0.027 . outreg using mortality, append noaster nolabel bdec(4) . . . *====================================================================== . *include first borns but without any allowance for initial conditions. . *remember - start of the process same as sample start . *allow for different effects for the first born . . foreach x of varlist female dum59 dum6069 ethnic hindu sikh educpa1 /* > */ educpa2 educma1 educma2 agemay agemays { 2. gen a`x'= `x'*chld1 3. } . . *avars1 and avars2 list the variables for the first born . . local avars1 "afemale adum59 adum6069 aethnic ahindu asikh aeducpa1" . local avars2 "aeducpa2 aeducma1 aeducma2 aagemay aagemays" . . xtprobit mort mort_1 chld1 `avars1' `avars2' female dum59 dum6069 /* > */ ethnic hindu sikh educpa1 educpa2 educma1 educma2 agemay agemays /* > */ chld2 chld3 chld4 Fitting comparison model: Iteration 0: log likelihood = -1626.5757 Iteration 1: log likelihood = -1561.2577 Iteration 2: log likelihood = -1559.5393 Iteration 3: log likelihood = -1559.5347 Iteration 4: log likelihood = -1559.5347 Fitting full model: rho = 0.0 log likelihood = -1559.5347 rho = 0.1 log likelihood = -1563.5391 Iteration 0: log likelihood = -1559.5347 (not concave) Iteration 1: log likelihood = -1559.4934 (not concave) Iteration 2: log likelihood = -1558.8423 (not concave) Iteration 3: log likelihood = -1557.6145 Iteration 4: log likelihood = -1555.3718 Iteration 5: log likelihood = -1553.1892 Iteration 6: log likelihood = -1553.1577 Iteration 7: log likelihood = -1553.1573 Iteration 8: log likelihood = -1553.1573 Random-effects probit regression Number of obs = 7211 Group variable (i): seqid Number of groups = 2390 Random effects u_i ~ Gaussian Obs per group: min = 1 avg = 3.0 max = 11 Wald chi2(29) = 91.35 Log likelihood = -1553.1573 Prob > chi2 = 0.0000 ------------------------------------------------------------------------------ mort | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- mort_1 | .1621844 .1265335 1.28 0.200 -.0858168 .4101855 chld1 | -2.77308 1.872817 -1.48 0.139 -6.443733 .8975729 afemale | -.361308 .1129743 -3.20 0.001 -.5827335 -.1398825 adum59 | -.1095054 .1577361 -0.69 0.488 -.4186624 .1996516 adum6069 | -.0438504 .1489084 -0.29 0.768 -.3357054 .2480047 aethnic | -.2490196 .1225388 -2.03 0.042 -.4891913 -.0088479 ahindu | -.1732669 .2487971 -0.70 0.486 -.6609003 .3143664 asikh | -.0590338 .2460704 -0.24 0.810 -.5413229 .4232553 aeducpa1 | -.0666058 .1581438 -0.42 0.674 -.3765619 .2433504 aeducpa2 | -.116906 .1671619 -0.70 0.484 -.4445372 .2107252 aeducma1 | .3180796 .1897645 1.68 0.094 -.0538519 .6900112 aeducma2 | -.1251146 .2183723 -0.57 0.567 -.5531164 .3028872 aagemay | .2112695 .1679667 1.26 0.208 -.1179393 .5404782 aagemays | -.0041592 .003865 -1.08 0.282 -.0117345 .003416 female | .2503305 .063655 3.93 0.000 .1255691 .375092 dum59 | .1252881 .1052089 1.19 0.234 -.0809177 .3314938 dum6069 | .0494355 .1010398 0.49 0.625 -.148599 .2474699 ethnic | .188088 .0756032 2.49 0.013 .0399084 .3362677 hindu | .017448 .1450554 0.12 0.904 -.2668553 .3017514 sikh | -.0462609 .1445258 -0.32 0.749 -.3295262 .2370045 educpa1 | .0805922 .0987946 0.82 0.415 -.1130416 .274226 educpa2 | .0833067 .1032362 0.81 0.420 -.1190325 .2856459 educma1 | .0342213 .125899 0.27 0.786 -.2125362 .2809788 educma2 | .2013035 .1381253 1.46 0.145 -.0694171 .472024 agemay | -.1792465 .0695152 -2.58 0.010 -.3154938 -.0429992 agemays | .002546 .0013158 1.93 0.053 -.000033 .005125 chld2 | -.4864301 .1220399 -3.99 0.000 -.7256239 -.2472363 chld3 | -.3828824 .1138629 -3.36 0.001 -.6060495 -.1597153 chld4 | -.1437939 .1127132 -1.28 0.202 -.3647078 .07712 _cons | 1.03154 .9377016 1.10 0.271 -.8063218 2.869401 -------------+---------------------------------------------------------------- /lnsig2u | -1.691679 .3565961 -2.390594 -.992763 -------------+---------------------------------------------------------------- sigma_u | .429197 .076525 .302614 .6087294 rho | .1555552 .0468417 .0838928 .2703667 ------------------------------------------------------------------------------ Likelihood-ratio test of rho=0: chibar2(01) = 12.75 Prob >= chibar2 = 0.000 . outreg using mortality, append noaster nolabel bdec(4) . . . local avars1 "afemale adum59 adum6069 aethnic ahindu asikh aeducpa1" . local avars2 "aeducpa2 aeducma1 aeducma2 aagemay aagemays" . . *heckman's correction . reprobhec mort mort_1 chld1 `avars1' `avars2' female dum59 dum6069 /* > */ ethnic hindu sikh educpa1 educpa2 educma1 educma2 agemay agemays /* > */ chld2 chld3 chld4, i(seqid) ldv(mort_1) q(12) rho0(0.08) tht0(1.0) Fitting initial probit Iteration 0: log likelihood = -1626.5757 Iteration 1: log likelihood = -1561.2577 Iteration 2: log likelihood = -1559.5393 Iteration 3: log likelihood = -1559.5347 Iteration 4: log likelihood = -1559.5347 Probit regression Number of obs = 7211 LR chi2(29) = 134.08 Prob > chi2 = 0.0000 Log likelihood = -1559.5347 Pseudo R2 = 0.0412 ------------------------------------------------------------------------------ mort | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- mort_1 | .4588258 .0902176 5.09 0.000 .2820025 .6356491 chld1 | -2.634247 1.751881 -1.50 0.133 -6.067871 .7993777 afemale | -.3439449 .1040215 -3.31 0.001 -.5478233 -.1400665 adum59 | -.1024521 .1472529 -0.70 0.487 -.3910625 .1861583 adum6069 | -.0488678 .1388039 -0.35 0.725 -.3209185 .2231829 aethnic | -.2252542 .1141734 -1.97 0.049 -.4490299 -.0014785 ahindu | -.1779273 .2340073 -0.76 0.447 -.6365731 .2807185 asikh | -.0766259 .2313365 -0.33 0.740 -.5300371 .3767853 aeducpa1 | -.0433004 .1473183 -0.29 0.769 -.3320389 .2454381 aeducpa2 | -.0912829 .156 -0.59 0.558 -.3970373 .2144715 aeducma1 | .3103947 .1765565 1.76 0.079 -.0356496 .656439 aeducma2 | -.1176628 .2030513 -0.58 0.562 -.515636 .2803104 aagemay | .1981333 .1567772 1.26 0.206 -.1091443 .5054109 aagemays | -.0039088 .0035984 -1.09 0.277 -.0109616 .003144 female | .238379 .0588232 4.05 0.000 .1230876 .3536704 dum59 | .1024962 .0927599 1.10 0.269 -.0793099 .2843024 dum6069 | .0440334 .08959 0.49 0.623 -.1315598 .2196265 ethnic | .1599544 .0654114 2.45 0.014 .0317505 .2881584 hindu | .041735 .1242117 0.34 0.737 -.2017154 .2851854 sikh | -.0117687 .1236209 -0.10 0.924 -.2540612 .2305237 educpa1 | .0576099 .085699 0.67 0.501 -.1103571 .2255769 educpa2 | .0601452 .0899655 0.67 0.504 -.116184 .2364744 educma1 | .0220287 .1114297 0.20 0.843 -.1963695 .240427 educma2 | .1902902 .1218651 1.56 0.118 -.0485611 .4291414 agemay | -.1485593 .0629972 -2.36 0.018 -.2720316 -.025087 agemays | .0020255 .0012004 1.69 0.092 -.0003272 .0043783 chld2 | -.5204387 .1106769 -4.70 0.000 -.7373615 -.303516 chld3 | -.4176735 .1044685 -4.00 0.000 -.6224281 -.2129189 chld4 | -.1728693 .1047837 -1.65 0.099 -.3782415 .0325029 _cons | .7939227 .8469491 0.94 0.349 -.866067 2.453912 ------------------------------------------------------------------------------ Fitting Full model rho0 = 0.1 tht0 = 1.0 Log Likelihood from model with rho=0: -1.6e+03 Iteration 0: log likelihood = -1560.8757 (not concave) Iteration 1: log likelihood = -1555.7993 (not concave) Iteration 2: log likelihood = -1554.4814 Iteration 3: log likelihood = -1553.8882 Iteration 4: log likelihood = -1553.1544 Iteration 5: log likelihood = -1553.1014 Iteration 6: log likelihood = -1553.101 Iteration 7: log likelihood = -1553.101 Random Effects Probit with normal heterogeneity and Heckman's model Number of obs = 7211 Wald chi2(29) = 90.97 Log likelihood = -1553.101 Prob > chi2 = 0.0000 ------------------------------------------------------------------------------ mort | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- mort_1 | .1673874 .1267229 1.32 0.187 -.0809849 .4157596 chld1 | -2.767675 1.89495 -1.46 0.144 -6.481709 .94636 afemale | -.3624356 .1143436 -3.17 0.002 -.5865449 -.1383264 adum59 | -.1069991 .158801 -0.67 0.500 -.4182432 .2042451 adum6069 | -.0428759 .1497357 -0.29 0.775 -.3363525 .2506006 aethnic | -.248742 .1233662 -2.02 0.044 -.4905353 -.0069488 ahindu | -.176451 .250734 -0.70 0.482 -.6678806 .3149785 asikh | -.0626663 .2480074 -0.25 0.801 -.5487518 .4234191 aeducpa1 | -.0667355 .1591477 -0.42 0.675 -.3786592 .2451882 aeducpa2 | -.1164551 .1682749 -0.69 0.489 -.4462679 .2133577 aeducma1 | .3254504 .1923901 1.69 0.091 -.0516272 .702528 aeducma2 | -.1236294 .219902 -0.56 0.574 -.5546294 .3073706 aagemay | .2080969 .1706167 1.22 0.223 -.1263056 .5424995 aagemays | -.0041254 .0039275 -1.05 0.294 -.0118231 .0035722 female | .249235 .063392 3.93 0.000 .1249891 .373481 dum59 | .1243272 .1043448 1.19 0.233 -.0801849 .3288392 dum6069 | .0493964 .1002017 0.49 0.622 -.1469953 .245788 ethnic | .187307 .0749077 2.50 0.012 .0404906 .3341233 hindu | .0171712 .1434698 0.12 0.905 -.2640246 .2983669 sikh | -.0454072 .1429657 -0.32 0.751 -.3256148 .2348003 educpa1 | .080839 .0978183 0.83 0.409 -.1108813 .2725592 educpa2 | .0823706 .1022855 0.81 0.421 -.1181054 .2828465 educma1 | .0332194 .1248515 0.27 0.790 -.211485 .2779237 educma2 | .2010544 .1369025 1.47 0.142 -.0672696 .4693785 agemay | -.1770815 .0693653 -2.55 0.011 -.3130349 -.0411281 agemays | .0025124 .0013117 1.92 0.055 -.0000584 .0050832 chld2 | -.4868752 .1212177 -4.02 0.000 -.7244576 -.2492928 chld3 | -.383923 .1132353 -3.39 0.001 -.6058601 -.1619859 chld4 | -.1451462 .1122091 -1.29 0.196 -.3650721 .0747796 _cons | 1.011714 .933125 1.08 0.278 -.8171772 2.840606 -------------+---------------------------------------------------------------- rho | .1454345 .0553115 2.63 0.009 .0370259 .2538431 tht | 1.16695 .5425033 2.15 0.031 .1036631 2.230237 vara | .1701853 .0757399 2.25 0.025 .0384495 .3402006 ------------------------------------------------------------------------------ Likelihood ratio test of rho=0: chibar2(01) = 12.87 Prob>=chibar2 = 0.000 ----------------------------------------------------------------- . outreg using mortality, append noaster nolabel bdec(4) . . *================================================================== . *orme's model . *step 1 . local avars1 "afemale adum59 adum6069 aethnic ahindu asikh aeducpa1" . local avars2 "aeducpa2 aeducma1 aeducma2 aagemay aagemays" . . probit mort `avars1' `avars2' if(bord==1) Iteration 0: log likelihood = -510.59035 Iteration 1: log likelihood = -498.81454 Iteration 2: log likelihood = -498.56578 Iteration 3: log likelihood = -498.56366 Iteration 4: log likelihood = -498.56366 Probit regression Number of obs = 2390 LR chi2(12) = 24.05 Prob > chi2 = 0.0200 Log likelihood = -498.56366 Pseudo R2 = 0.0236 ------------------------------------------------------------------------------ mort | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- afemale | -.1055659 .0857922 -1.23 0.219 -.2737155 .0625838 adum59 | .0000441 .1143635 0.00 1.000 -.2241042 .2241925 adum6069 | -.0048345 .1060197 -0.05 0.964 -.2126292 .2029603 aethnic | -.0652998 .0935784 -0.70 0.485 -.24871 .1181105 ahindu | -.1361924 .1983201 -0.69 0.492 -.5248926 .2525079 asikh | -.0883946 .1955363 -0.45 0.651 -.4716388 .2948495 aeducpa1 | .0143095 .1198263 0.12 0.905 -.2205457 .2491648 aeducpa2 | -.0311376 .1274449 -0.24 0.807 -.2809251 .2186498 aeducma1 | .3324234 .1369511 2.43 0.015 .0640042 .6008426 aeducma2 | .0726273 .1624152 0.45 0.655 -.2457007 .3909554 aagemay | .049574 .1435633 0.35 0.730 -.2318049 .330953 aagemays | -.0018833 .0033923 -0.56 0.579 -.0085321 .0047655 _cons | -1.840324 1.533547 -1.20 0.230 -4.84602 1.165372 ------------------------------------------------------------------------------ . predict xb0, xb . . gen correc=normden(xb0)/(1-norm(xb0)) if(mort==1 & bord==1) (7079 missing values generated) . replace correc=-normden(xb0)/norm(xb0) if(mort==0 & bord==1) (2258 real changes made) . by seqid: replace correc=correc[1] if(bord>1) (4821 real changes made) . . xtprobit mort mort_1 female dum59 dum6069 /* > */ ethnic hindu sikh educpa1 educpa2 educma1 educma2 agemay agemays /* > */ chld2 chld3 chld4 correc if(bord>1) Fitting comparison model: Iteration 0: log likelihood = -1115.3981 Iteration 1: log likelihood = -1060.5438 Iteration 2: log likelihood = -1059.0642 Iteration 3: log likelihood = -1059.0622 Iteration 4: log likelihood = -1059.0622 Fitting full model: rho = 0.0 log likelihood = -1059.0621 rho = 0.1 log likelihood = -1061.6237 Iteration 0: log likelihood = -1059.0621 (not concave) Iteration 1: log likelihood = -1059.0258 (not concave) Iteration 2: log likelihood = -1058.3992 Iteration 3: log likelihood = -1055.6367 Iteration 4: log likelihood = -1055.5254 Iteration 5: log likelihood = -1055.5249 Iteration 6: log likelihood = -1055.5249 Random-effects probit regression Number of obs = 4821 Group variable (i): seqid Number of groups = 2071 Random effects u_i ~ Gaussian Obs per group: min = 1 avg = 2.3 max = 10 Wald chi2(17) = 84.80 Log likelihood = -1055.5249 Prob > chi2 = 0.0000 ------------------------------------------------------------------------------ mort | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- mort_1 | .1896175 .1268521 1.49 0.135 -.0590081 .438243 female | .2479287 .0631935 3.92 0.000 .1240718 .3717857 dum59 | .1222069 .103659 1.18 0.238 -.080961 .3253748 dum6069 | .0502699 .0995808 0.50 0.614 -.1449048 .2454446 ethnic | .1955078 .0748676 2.61 0.009 .0487701 .3422456 hindu | .0447564 .1421367 0.31 0.753 -.2338265 .3233392 sikh | -.0258213 .141585 -0.18 0.855 -.3033227 .2516801 educpa1 | .073939 .0969974 0.76 0.446 -.1161725 .2640504 educpa2 | .0846159 .1016288 0.83 0.405 -.1145729 .2838047 educma1 | -.02347 .1256228 -0.19 0.852 -.2696862 .2227461 educma2 | .1886759 .135923 1.39 0.165 -.0777283 .4550802 agemay | -.1700925 .0690168 -2.46 0.014 -.3053629 -.0348221 agemays | .0024313 .0013069 1.86 0.063 -.0001301 .0049928 chld2 | -.468851 .1216663 -3.85 0.000 -.7073126 -.2303893 chld3 | -.3718629 .113429 -3.28 0.001 -.5941797 -.1495461 chld4 | -.1385744 .1121758 -1.24 0.217 -.358435 .0812862 correc | .1509078 .0578345 2.61 0.009 .0375543 .2642613 _cons | 1.186541 .9347775 1.27 0.204 -.6455888 3.018671 -------------+---------------------------------------------------------------- /lnsig2u | -1.878233 .4672989 -2.794122 -.9623436 -------------+---------------------------------------------------------------- sigma_u | .3909732 .0913507 .2473228 .6180587 rho | .132592 .0537447 .0576427 .2764092 ------------------------------------------------------------------------------ Likelihood-ratio test of rho=0: chibar2(01) = 7.07 Prob >= chibar2 = 0.004 . outreg using mortality, append noaster nolabel bdec(4) . . *================================================================= . *wooldridge's method . *Only uses sample with bord>1 . *need to condition on the initial value . *first run without any covariates except the initial value . . by seqid: gen initial=mort[1] . . xtprobit mort mort_1 female dum59 dum6069 /* > */ ethnic hindu sikh educpa1 educpa2 educma1 educma2 agemay agemays /* > */ chld2 chld3 chld4 initial if(bord>1) Fitting comparison model: Iteration 0: log likelihood = -1115.3981 Iteration 1: log likelihood = -1060.1338 Iteration 2: log likelihood = -1058.6328 Iteration 3: log likelihood = -1058.6308 Iteration 4: log likelihood = -1058.6308 Fitting full model: rho = 0.0 log likelihood = -1058.6307 rho = 0.1 log likelihood = -1061.0682 Iteration 0: log likelihood = -1058.6307 (not concave) Iteration 1: log likelihood = -1058.5984 (not concave) Iteration 2: log likelihood = -1058.2499 (not concave) Iteration 3: log likelihood = -1057.4161 Iteration 4: log likelihood = -1056.5266 Iteration 5: log likelihood = -1054.9802 Iteration 6: log likelihood = -1054.9606 Iteration 7: log likelihood = -1054.9602 Iteration 8: log likelihood = -1054.9602 Random-effects probit regression Number of obs = 4821 Group variable (i): seqid Number of groups = 2071 Random effects u_i ~ Gaussian Obs per group: min = 1 avg = 2.3 max = 10 Wald chi2(17) = 85.68 Log likelihood = -1054.9602 Prob > chi2 = 0.0000 ------------------------------------------------------------------------------ mort | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- mort_1 | .1785782 .1267466 1.41 0.159 -.0698405 .4269968 female | .2485983 .0632869 3.93 0.000 .1245583 .3726384 dum59 | .122189 .1038833 1.18 0.240 -.0814184 .3257964 dum6069 | .0484501 .0997949 0.49 0.627 -.1471443 .2440444 ethnic | .1899268 .0746991 2.54 0.011 .0435192 .3363345 hindu | .0267704 .1425088 0.19 0.851 -.2525418 .3060825 sikh | -.0388357 .1420495 -0.27 0.785 -.3172475 .2395762 educpa1 | .0788609 .0973435 0.81 0.418 -.1119289 .2696507 educpa2 | .0833009 .1018367 0.82 0.413 -.1162952 .2828971 educma1 | .0156291 .1243955 0.13 0.900 -.2281816 .2594399 educma2 | .1971629 .1361862 1.45 0.148 -.0697571 .4640828 agemay | -.1737514 .0691602 -2.51 0.012 -.3093029 -.0382 agemays | .0024655 .0013085 1.88 0.060 -.0000991 .0050302 chld2 | -.4816014 .1211795 -3.97 0.000 -.7191088 -.244094 chld3 | -.3804215 .113181 -3.36 0.001 -.6022523 -.1585908 chld4 | -.1430953 .1121323 -1.28 0.202 -.3628707 .07668 initial | .3419674 .1221965 2.80 0.005 .1024666 .5814682 _cons | .9454502 .930295 1.02 0.309 -.8778946 2.768795 -------------+---------------------------------------------------------------- /lnsig2u | -1.858293 .460223 -2.760314 -.9562725 -------------+---------------------------------------------------------------- sigma_u | .3948906 .0908689 .2515391 .6199377 rho | .1349021 .0537097 .0595068 .2776251 ------------------------------------------------------------------------------ Likelihood-ratio test of rho=0: chibar2(01) = 7.34 Prob >= chibar2 = 0.003 . outreg using mortality, append noaster nolabel bdec(4) . . . *only time varying covariate are agemay and agemays . *so let us use this and see. . by seqid: gen avage=sum(agemay)/_n . by seqid: replace avage=avage[_N] (4814 real changes made) . by seqid: gen avages=sum(agemays)/_n . by seqid: replace avages=avages[_N] (4814 real changes made) . . xtprobit mort mort_1 female dum59 dum6069 /* > */ ethnic hindu sikh educpa1 educpa2 educma1 educma2 agemay agemays /* > */ chld2 chld3 chld4 initial avage avages if(bord>1) Fitting comparison model: Iteration 0: log likelihood = -1115.3981 Iteration 1: log likelihood = -1021.1541 Iteration 2: log likelihood = -1017.2815 Iteration 3: log likelihood = -1017.261 Iteration 4: log likelihood = -1017.261 Fitting full model: rho = 0.0 log likelihood = -1017.261 rho = 0.1 log likelihood = -1020.9763 Iteration 0: log likelihood = -1017.261 (not concave) Iteration 1: log likelihood = -1017.2531 (not concave) Iteration 2: log likelihood = -1016.9944 Iteration 3: log likelihood = -1016.1006 Iteration 4: log likelihood = -1016.0847 Iteration 5: log likelihood = -1016.0828 Iteration 6: log likelihood = -1016.0828 Random-effects probit regression Number of obs = 4821 Group variable (i): seqid Number of groups = 2071 Random effects u_i ~ Gaussian Obs per group: min = 1 avg = 2.3 max = 10 Wald chi2(19) = 158.95 Log likelihood = -1016.0828 Prob > chi2 = 0.0000 ------------------------------------------------------------------------------ mort | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- mort_1 | .1412109 .1258987 1.12 0.262 -.1055461 .3879679 female | .1988457 .0632577 3.14 0.002 .0748629 .3228284 dum59 | -.2897315 .1124455 -2.58 0.010 -.5101206 -.0693424 dum6069 | -.2131039 .1024576 -2.08 0.038 -.4139172 -.0122907 ethnic | .1202715 .0729695 1.65 0.099 -.0227461 .2632892 hindu | .1079914 .1376998 0.78 0.433 -.1618954 .3778781 sikh | .0701817 .1371811 0.51 0.609 -.1986884 .3390517 educpa1 | .0705753 .0946006 0.75 0.456 -.1148386 .2559891 educpa2 | .0662111 .0988783 0.67 0.503 -.1275868 .260009 educma1 | -.0286462 .1217954 -0.24 0.814 -.2673609 .2100685 educma2 | .1705473 .1330448 1.28 0.200 -.0902158 .4313104 agemay | -.6528629 .0996241 -6.55 0.000 -.8481225 -.4576032 agemays | .0091288 .0017621 5.18 0.000 .0056752 .0125824 chld2 | -1.29845 .1521127 -8.54 0.000 -1.596586 -1.000315 chld3 | -.8809142 .125507 -7.02 0.000 -1.126903 -.6349251 chld4 | -.3898631 .1137384 -3.43 0.001 -.6127863 -.16694 initial | .2294399 .1177809 1.95 0.051 -.0014065 .4602863 avage | .7899575 .1570259 5.03 0.000 .4821923 1.097723 avages | -.0120324 .0030005 -4.01 0.000 -.0179133 -.0061515 _cons | -2.257468 1.342455 -1.68 0.093 -4.888631 .3736946 -------------+---------------------------------------------------------------- /lnsig2u | -2.500321 .7399872 -3.950669 -1.049972 -------------+---------------------------------------------------------------- sigma_u | .2864589 .105988 .1387149 .5915636 rho | .0758357 .0518618 .0188786 .2592304 ------------------------------------------------------------------------------ Likelihood-ratio test of rho=0: chibar2(01) = 2.36 Prob >= chibar2 = 0.062 . outreg using mortality, append noaster nolabel bdec(4) . . . xtprobit mort mort_1 female dum59 dum6069 /* > */ ethnic hindu sikh educpa1 educpa2 educma1 educma2 agemay agemays /* > */ chld2 chld3 chld4 initial avage if(bord>1) Fitting comparison model: Iteration 0: log likelihood = -1115.3981 Iteration 1: log likelihood = -1029.6169 Iteration 2: log likelihood = -1026.6601 Iteration 3: log likelihood = -1026.651 Iteration 4: log likelihood = -1026.651 Fitting full model: rho = 0.0 log likelihood = -1026.651 rho = 0.1 log likelihood = -1029.855 Iteration 0: log likelihood = -1026.651 (not concave) Iteration 1: log likelihood = -1026.3551 (not concave) Iteration 2: log likelihood = -1026.054 Iteration 3: log likelihood = -1026.0475 Iteration 4: log likelihood = -1024.9215 Iteration 5: log likelihood = -1024.8984 Iteration 6: log likelihood = -1024.8983 Iteration 7: log likelihood = -1024.8983 Random-effects probit regression Number of obs = 4821 Group variable (i): seqid Number of groups = 2071 Random effects u_i ~ Gaussian Obs per group: min = 1 avg = 2.3 max = 10 Wald chi2(18) = 142.93 Log likelihood = -1024.8983 Prob > chi2 = 0.0000 ------------------------------------------------------------------------------ mort | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- mort_1 | .1495956 .1258252 1.19 0.234 -.0970172 .3962083 female | .1942619 .0634172 3.06 0.002 .0699665 .3185573 dum59 | -.2522661 .1132925 -2.23 0.026 -.4743153 -.0302169 dum6069 | -.1475279 .1016454 -1.45 0.147 -.3467492 .0516935 ethnic | .1183103 .0736955 1.61 0.108 -.0261302 .2627507 hindu | .1029248 .1394229 0.74 0.460 -.1703391 .3761887 sikh | .0577368 .1389358 0.42 0.678 -.2145724 .3300459 educpa1 | .0453156 .0956462 0.47 0.636 -.1421476 .2327788 educpa2 | .0634302 .1000528 0.63 0.526 -.1326698 .2595302 educma1 | -.0337541 .1230117 -0.27 0.784 -.2748525 .2073444 educma2 | .1740201 .1338806 1.30 0.194 -.0883811 .4364213 agemay | -.3756454 .073924 -5.08 0.000 -.5205338 -.2307571 agemays | .0042051 .0013122 3.20 0.001 .0016332 .006777 chld2 | -1.120918 .1449505 -7.73 0.000 -1.405016 -.8368201 chld3 | -.790253 .1231696 -6.42 0.000 -1.031661 -.548845 chld4 | -.3551686 .1135603 -3.13 0.002 -.5777427 -.1325945 initial | .2647245 .1187873 2.23 0.026 .0319056 .4975434 avage | .1668541 .0220089 7.58 0.000 .1237175 .2099907 _cons | 1.628809 .9193998 1.77 0.076 -.1731821 3.430799 -------------+---------------------------------------------------------------- /lnsig2u | -2.282998 .6237051 -3.505438 -1.060559 -------------+---------------------------------------------------------------- sigma_u | .3193399 .099587 .1733021 .5884406 rho | .0925409 .0523769 .0291579 .2572027 ------------------------------------------------------------------------------ Likelihood-ratio test of rho=0: chibar2(01) = 3.51 Prob >= chibar2 = 0.031 . outreg using mortality, append noaster nolabel bdec(4) . . *let us calculate the average partial effects for the above model . *we will need to do a probit . . probit mort mort_1 female dum59 dum6069 /* > */ ethnic hindu sikh educpa1 educpa2 educma1 educma2 agemay agemays /* > */ chld2 chld3 chld4 initial avage if(bord>1) Iteration 0: log likelihood = -1115.3981 Iteration 1: log likelihood = -1029.6169 Iteration 2: log likelihood = -1026.6601 Iteration 3: log likelihood = -1026.651 Iteration 4: log likelihood = -1026.651 Probit regression Number of obs = 4821 LR chi2(18) = 177.49 Prob > chi2 = 0.0000 Log likelihood = -1026.651 Pseudo R2 = 0.0796 ------------------------------------------------------------------------------ mort | Coef. Std. Err. z P>|z| [95% Conf. Interval] -------------+---------------------------------------------------------------- mort_1 | .2814934 .1004231 2.80 0.005 .0846676 .4783191 female | .1862409 .0603527 3.09 0.002 .0679518 .3045301 dum59 | -.2413475 .1049713 -2.30 0.021 -.4470875 -.0356075 dum6069 | -.1369476 .0943015 -1.45 0.146 -.3217752 .04788 ethnic | .1078478 .0674852 1.60 0.110 -.0244208 .2401165 hindu | .1054665 .126424 0.83 0.404 -.14232 .353253 sikh | .0674 .1258888 0.54 0.592 -.1793376 .3141375 educpa1 | .0371269 .0875962 0.42 0.672 -.1345584 .2088122 educpa2 | .0485298 .0916889 0.53 0.597 -.131177 .2282366 educma1 | -.036855 .1141535 -0.32 0.747 -.2605917 .1868818 educma2 | .1720794 .1240469 1.39 0.165 -.0710482 .4152069 agemay | -.3449124 .0676052 -5.10 0.000 -.4774161 -.2124087 agemays | .0037058 .001213 3.06 0.002 .0013284 .0060832 chld2 | -1.107741 .1350788 -8.20 0.000 -1.372491 -.8429917 chld3 | -.7839698 .1159269 -6.76 0.000 -1.011182 -.5567574 chld4 | -.354373 .108196 -3.28 0.001 -.5664332 -.1423128 initial | .1825589 .1004689 1.82 0.069 -.0143565 .3794743 avage | .159954 .0201642 7.93 0.000 .1204329 .1994751 _cons | 1.422468 .8581438 1.66 0.097 -.2594631 3.104399 ------------------------------------------------------------------------------ . . matrix coef=e(b) . predict xb1, xb . su xb1 Variable | Obs Mean Std. Dev. Min Max -------------+-------------------------------------------------------- xb1 | 7211 -1.201459 .7479457 -3.008977 1.426707 . gen den=normalden(xb1) . . . foreach x of varlist female dum59 dum6069 ethnic hindu sikh educpa1 /* > */ educpa2 educma1 educma2 agemay agemays chld2 chld3 chld4{ 2. gen ap`x'= den*_b[`x'] 3. } . . su ap* Variable | Obs Mean Std. Dev. Min Max -------------+-------------------------------------------------------- apfemale | 7211 .0363877 .0245518 .0008034 .0742994 apdum59 | 7211 -.0471544 .0318164 -.0962837 -.0010412 apdum6069 | 7211 -.0267568 .0180536 -.0546342 -.0005908 apethnic | 7211 .0210713 .0142174 .0004652 .0430251 aphindu | 7211 .020606 .0139035 .000455 .042075 -------------+-------------------------------------------------------- apsikh | 7211 .0131686 .0088852 .0002908 .0268887 apeducpa1 | 7211 .0072538 .0048944 .0001602 .0148115 apeducpa2 | 7211 .0094817 .0063976 .0002094 .0193606 apeducma1 | 7211 -.0072007 .0048585 -.014703 -.000159 apeducma2 | 7211 .0336208 .0226849 .0007423 .0686497 -------------+-------------------------------------------------------- apagemay | 7211 -.0673889 .0454692 -.1376002 -.0014879 apagemays | 7211 .000724 .0004885 .000016 .0014784 apchld2 | 7211 -.2164302 .1460316 -.4419248 -.0047787 apchld3 | 7211 -.1531718 .1033494 -.3127587 -.003382 apchld4 | 7211 -.0692373 .0467164 -.1413744 -.0015287 . . end of do-file . do "C:\DOCUME~1\User\LOCALS~1\Temp\STD0b000000.tmp" . reg mort mort_1 female dum59 dum6069 /* > */ ethnic hindu sikh educpa1 educpa2 educma1 educma2 agemay agemays /* > */ chld2 chld3 chld4 initial avage if(bord>1) Source | SS df MS Number of obs = 4821 -------------+------------------------------ F( 18, 4802) = 10.56 Model | 10.6122707 18 .589570595 Prob > F = 0.0000 Residual | 268.090903 4802 .055829009 R-squared = 0.0381 -------------+------------------------------ Adj R-squared = 0.0345 Total | 278.703174 4820 .057822235 Root MSE = .23628 ------------------------------------------------------------------------------ mort | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- mort_1 | .0514898 .0143156 3.60 0.000 .0234247 .0795549 female | .0195179 .0069253 2.82 0.005 .0059411 .0330947 dum59 | -.0245716 .0118371 -2.08 0.038 -.0477777 -.0013655 dum6069 | -.0125785 .010759 -1.17 0.242 -.033671 .008514 ethnic | .0096524 .0075462 1.28 0.201 -.0051416 .0244465 hindu | .0133938 .0150658 0.89 0.374 -.0161421 .0429296 sikh | .009498 .014956 0.64 0.525 -.0198225 .0388186 educpa1 | .0040211 .0098722 0.41 0.684 -.0153329 .0233752 educpa2 | .0044022 .0103917 0.42 0.672 -.0159702 .0247747 educma1 | -.0095639 .0118291 -0.81 0.419 -.0327542 .0136265 educma2 | .0124934 .0134209 0.93 0.352 -.0138177 .0388044 agemay | -.0430934 .0081442 -5.29 0.000 -.0590597 -.027127 agemays | .0004786 .0001447 3.31 0.001 .0001949 .0007623 chld2 | -.1325026 .0161104 -8.22 0.000 -.1640864 -.1009188 chld3 | -.0941729 .0139326 -6.76 0.000 -.1214872 -.0668586 chld4 | -.0454548 .0137941 -3.30 0.001 -.0724976 -.018412 initial | .025252 .0136076 1.86 0.064 -.0014251 .0519292 avage | .0193358 .0023647 8.18 0.000 .0146998 .0239717 _cons | .4509113 .1033181 4.36 0.000 .2483604 .6534622 ------------------------------------------------------------------------------ . outreg using mortality, append noaster nolabel bdec(4) . end of do-file . do "C:\DOCUME~1\User\LOCALS~1\Temp\STD0b000000.tmp" . drop ap* . end of do-file . do "C:\DOCUME~1\User\LOCALS~1\Temp\STD0b000000.tmp" . foreach x of varlist mort_1 female dum59 dum6069 ethnic hindu sikh educpa1 /* > */ educpa2 educma1 educma2 agemay agemays chld2 chld3 chld4{ 2. gen ap`x'= den*_b[`x'] 3. } . . end of do-file . do "C:\DOCUME~1\User\LOCALS~1\Temp\STD0b000000.tmp" . su ap* Variable | Obs Mean Std. Dev. Min Max -------------+-------------------------------------------------------- apmort_1 | 7211 .0100601 .0067878 .0002221 .0205415 apfemale | 7211 .0038134 .002573 .0000842 .0077865 apdum59 | 7211 -.0048008 .0032392 -.0098027 -.000106 apdum6069 | 7211 -.0024576 .0016582 -.0050181 -.0000543 apethnic | 7211 .0018859 .0012725 .0000416 .0038508 -------------+-------------------------------------------------------- aphindu | 7211 .0026169 .0017657 .0000578 .0053433 apsikh | 7211 .0018557 .0012521 .000041 .0037892 apeducpa1 | 7211 .0007856 .0005301 .0000173 .0016042 apeducpa2 | 7211 .0008601 .0005803 .000019 .0017562 apeducma1 | 7211 -.0018686 .0012608 -.0038154 -.0000413 -------------+-------------------------------------------------------- apeducma2 | 7211 .0024409 .001647 .0000539 .0049841 apagemay | 7211 -.0084196 .0056809 -.0171918 -.0001859 apagemays | 7211 .0000935 .0000631 2.06e-06 .0001909 apchld2 | 7211 -.0258883 .0174676 -.0528609 -.0005716 apchld3 | 7211 -.0183995 .0124147 -.0375696 -.0004063 -------------+-------------------------------------------------------- apchld4 | 7211 -.008881 .0059922 -.0181339 -.0001961 . end of do-file . drop ap* . do "C:\DOCUME~1\User\LOCALS~1\Temp\STD0b000000.tmp" . foreach x of varlist mort_1 female dum59 dum6069 ethnic hindu sikh educpa1 /* > */ educpa2 educma1 educma2 agemay agemays chld2 chld3 chld4{ 2. gen ap`x'= den*_b[`x'] 3. } . . . su ap* Variable | Obs Mean Std. Dev. Min Max -------------+-------------------------------------------------------- apmort_1 | 7211 .0100601 .0067878 .0002221 .0205415 apfemale | 7211 .0038134 .002573 .0000842 .0077865 apdum59 | 7211 -.0048008 .0032392 -.0098027 -.000106 apdum6069 | 7211 -.0024576 .0016582 -.0050181 -.0000543 apethnic | 7211 .0018859 .0012725 .0000416 .0038508 -------------+-------------------------------------------------------- aphindu | 7211 .0026169 .0017657 .0000578 .0053433 apsikh | 7211 .0018557 .0012521 .000041 .0037892 apeducpa1 | 7211 .0007856 .0005301 .0000173 .0016042 apeducpa2 | 7211 .0008601 .0005803 .000019 .0017562 apeducma1 | 7211 -.0018686 .0012608 -.0038154 -.0000413 -------------+-------------------------------------------------------- apeducma2 | 7211 .0024409 .001647 .0000539 .0049841 apagemay | 7211 -.0084196 .0056809 -.0171918 -.0001859 apagemays | 7211 .0000935 .0000631 2.06e-06 .0001909 apchld2 | 7211 -.0258883 .0174676 -.0528609 -.0005716 apchld3 | 7211 -.0183995 .0124147 -.0375696 -.0004063 -------------+-------------------------------------------------------- apchld4 | 7211 -.008881 .0059922 -.0181339 -.0001961 . end of do-file . exit, clear