.

regress gasto_mon_mens tot_integ

Source SS df MS Number of obs = 425

F(1, 423) = 39.65
Model 2.2883e+09 1 2.2883e+09 Prob > F = 0.0000
Residual 2.4410e+10 423 57707304.9 R-squared = 0.0857
Adj R-squared = 0.0835
Total 2.6698e+10 424 62968080 Root MSE = 7596.5

gasto_mon_~s Coef. Std. Err. t P>|t| [95% Conf. Interval]

tot_integ 1420.497 225.5806 6.30 0.000 977.099 1863.896

_cons 6475.524 866.7213 7.47 0.000 4771.907 8179.14

.
 Prueba Reset Ramsey
Ho: El modelo no ha omitido variables

. ovtest

Ramsey RESET test using powers of the fitted values of gasto_mon_mens

Ho: model has no omitted variables
F(3, 420) = 0.91
Prob > F = 0.4354

El valor-p es alto (mayor a 0.05): NO se rechaza la hipótesis Nula

Existe evidencia: El modelo no ha omitido variables
Sin embargo, el valor del coeficiente intercepto ( 𝛽o) es alto 6475.52 y la bondad
de ajuste R² = 0.0857 es muy bajo.
 Se estiman los residuos

. predict residuals1, r

. regress residuals1 tot_integ ingtrab_mens

Source SS df MS Number of obs = 425

F(2, 422) = 91.63
Model 7.3908e+09 2 3.6954e+09 Prob > F = 0.0000
Residual 1.7019e+10 422 40330214.1 R-squared = 0.3028
Adj R-squared = 0.2995
Total 2.4410e+10 424 57571202.3 Root MSE = 6350.6

residuals1 Coef. Std. Err. t P>|t| [95% Conf. Interval]

tot_integ -902.6973 200.0248 -4.51 0.000 -1295.866 -509.5283

ingtrab_mens .4529775 .0334615 13.54 0.000 .3872056 .5187494
_cons -1889.099 737.884 -2.56 0.011 -3339.485 -438.7133

La variable ingtrab_mens es estadísticamente significativa, puede incluirse en el modelo.

. regress gasto_mon_mens tot_integ ingtrab_mens

Source SS df MS Number of obs = 425

F(2, 422) = 120.00
Model 9.6791e+09 2 4.8396e+09 Prob > F = 0.0000
Residual 1.7019e+10 422 40330214.2 R-squared = 0.3625
Adj R-squared = 0.3595
Total 2.6698e+10 424 62968080 Root MSE = 6350.6

gasto_mon_~s Coef. Std. Err. t P>|t| [95% Conf. Interval]

tot_integ 517.8002 200.0248 2.59 0.010 124.6312 910.9692

ingtrab_mens .4529775 .0334615 13.54 0.000 .3872056 .5187494
_cons 4586.424 737.884 6.22 0.000 3136.039 6036.81
 Prueba Reset Ramsey

Ho: El modelo no ha omitido variables

. ovtest

Ramsey RESET test using powers of the fitted values of gasto_mon_mens

Ho: model has no omitted variables
F(3, 419) = 7.44
Prob > F = 0.0001

El valor-p es bajo (menor a 0.05). Se rechaza la hipótesis Nula.

Existe evidencia: El modelo ha omitido variables
Se estiman los residuos

. regress residuals3 tot_integ ingtrab_mens negocio_mens

Source SS df MS Number of obs = 425

F(3, 421) = 0.00
Model 367327.175 3 122442.392 Prob > F = 0.9998
Residual 1.7019e+10 421 40425137.9 R-squared = 0.0000
Adj R-squared = -0.0071
Total 1.7019e+10 424 40139977.3 Root MSE = 6358.1

residuals3 Coef. Std. Err. t P>|t| [95% Conf. Interval]

tot_integ -.5132378 200.3324 -0.00 0.998 -394.2896 393.2631

ingtrab_mens -.0005723 .0340346 -0.02 0.987 -.0674711 .0663265
negocio_mens .0095012 .099673 0.10 0.924 -.1864176 .20542
_cons -3.193638 739.5112 -0.00 0.997 -1456.788 1450.4

La variación negocio_mens (ingreso independiente) NO es estadísticamente significativa.

Puede NO incluirse en el modelo
Bondad de ajuste contradictoria
. regress gasto_mon_mens tot_integ ingtrab_mens negocio_mens

Source SS df MS Number of obs = 425

F(3, 421) = 79.81
Model 9.6795e+09 3 3.2265e+09 Prob > F = 0.0000
Residual 1.7019e+10 421 40425138 R-squared = 0.3625
Adj R-squared = 0.3580
Total 2.6698e+10 424 62968080 Root MSE = 6358.1

gasto_mon_~s Coef. Std. Err. t P>|t| [95% Conf. Interval]

tot_integ 517.287 200.3324 2.58 0.010 123.5106 911.0633

ingtrab_mens .4524052 .0340346 13.29 0.000 .3855063 .519304
negocio_mens .0095012 .099673 0.10 0.924 -.1864176 .20542
_cons 4583.231 739.5112 6.20 0.000 3129.637 6036.825
 Prueba Reset Ramsey

Ho: El modelo no ha omitido variables

. ovtest

Ramsey RESET test using powers of the fitted values of gasto_mon_mens

Ho: model has no omitted variables
F(3, 418) = 7.67
Prob > F = 0.0001

El valor-p es bajo (menor a 0.05). Se rechaza la Hipótesis Nula.

Existe evidencia: El modelo ha omitido variables
Se estiman los residuos

. regress residuals4 tot_integ ingtrab_mens negocio_mens transfer_mens

Source SS df MS Number of obs = 425

F(4, 420) = 8.94
Model 1.3355e+09 4 333867138 Prob > F = 0.0000
Residual 1.5684e+10 420 37341701.5 R-squared = 0.0785
Adj R-squared = 0.0697
Total 1.7019e+10 424 40139111.3 Root MSE = 6110.8

residuals4 Coef. Std. Err. t P>|t| [95% Conf. Interval]

tot_integ -10.99791 192.5495 -0.06 0.954 -389.4786 367.4827

ingtrab_mens .0182097 .0328522 0.55 0.580 -.0463656 .082785
negocio_mens -.0220962 .0958676 -0.23 0.818 -.2105363 .1663438
transfer_mens .1710314 .0285994 5.98 0.000 .1148157 .2272471
_cons -638.6076 718.7259 -0.89 0.375 -2051.355 774.1403

La variable transferencias (transfer_mens) es estadísticamente significativa.

El valor-p es bajo: La variable debe ser incluida en el modelo.
2. Transforme las variables en logarítmicas y obtenga los siguientes modelos:

. gen ingasto_mon_mens=log(gasto_mon_mens)
(1 missing value generated)

. gen intot_integ=log(tot_integ)

. gen iningtrab_mens=log(ingtrab_mens)
(40 missing values generated)

. gen innegocio_mens=log(negocio_mens)
(322 missing values generated)

. gen intransfer_mens=log(transfer_mens)
(163 missing values generated)
 . regress gasto_mon_mens tot_integ ingtrab_mens negocio_mens transfer_mens

Source SS df MS Number of obs = 425

F(4, 420) = 73.74
Model 1.1015e+10 4 2.7537e+09 Prob > F = 0.0000
Residual 1.5684e+10 420 37341701.2 R-squared = 0.4126
Adj R-squared = 0.4070
Total 2.6698e+10 424 62968080 Root MSE = 6110.8

gasto_mon_m~s Coef. Std. Err. t P>|t| [95% Conf. Interval]

tot_integ 506.289 192.5495 2.63 0.009 127.8084 884.7697

ingtrab_mens .4706149 .0328522 14.33 0.000 .4060396 .5351901
negocio_mens -.012595 .0958676 -0.13 0.896 -.2010351 .175845
transfer_mens .1710314 .0285994 5.98 0.000 .1148158 .2272471
_cons 3944.623 718.7259 5.49 0.000 2531.875 5357.371

. estimates store M1, title(modelo1)

 . regress gasto_mon_mens intot_integ iningtrab_mens innegocio_mens intransfer_mens

Source SS df MS Number of obs = 66

F(4, 61) = 17.52
Model 2.2049e+09 4 551223936 Prob > F = 0.0000
Residual 1.9193e+09 61 31464085.8 R-squared = 0.5346
Adj R-squared = 0.5041
Total 4.1242e+09 65 63449307.3 Root MSE = 5609.3

gasto_mon_mens Coef. Std. Err. t P>|t| [95% Conf. Interval]

intot_integ 499.7484 1324.239 0.38 0.707 -2148.231 3147.728

iningtrab_mens 5629.051 819.5989 6.87 0.000 3990.162 7267.94
innegocio_mens -1450.341 546.7562 -2.65 0.010 -2543.648 -357.0345
intransfer_mens 1035.008 546.6147 1.89 0.063 -58.01599 2128.031
_cons -35099.8 7863.309 -4.46 0.000 -50823.46 -19376.14

. estimates store M3, title(modelo3)

k

k
 . regress ingasto_mon_mens tot_integ ingtrab_mens negocio_mens transfer_mens

Source SS df MS Number of obs = 424

F(4, 419) = 72.20
Model 83.5606176 4 20.8901544 Prob > F = 0.0000
Residual 121.234234 419 .289341846 R-squared = 0.4080
Adj R-squared = 0.4024
Total 204.794851 423 .484148584 Root MSE = .53791

ingasto_mon~s Coef. Std. Err. t P>|t| [95% Conf. Interval]

tot_integ .0815438 .0169784 4.80 0.000 .0481702 .1149173

ingtrab_mens .0000386 2.89e-06 13.34 0.000 .0000329 .0000443
negocio_mens -5.28e-06 8.44e-06 -0.63 0.532 -.0000219 .0000113
transfer_mens .0000121 2.52e-06 4.82 0.000 7.18e-06 .0000171
_cons 8.378693 .0635216 131.90 0.000 8.253832 8.503554

. estimates store M2, title(modelo2)

Cuando la educación (tot_integ), la experiencia (ingtrab_mens), la antigüedad
(negocio_mens) y (tranfer_ mens) son nulos, el ingreso aumenta en promedio
837.86%.
Por cada año que aumenta la educación (tot_integ), el salario (gasto_mon_mens)
umenta en promedio 8.15%, manteniendo constante la experiencia (ingtrab_mens),
la antigüedad (negocio_mens) y (tranfer_ mens).

Por cada año que la experiencia (ingtrab_mens), el salario (gasto_mon_mens)

aumenta en promedio 0.41%, manteniendo constante la educación (educ) y la
antigüedad (tenure).

Por cada año que la antigüedad (tenure), el salario (wage) aumenta en promedio
2.20%, manteniendo constante la educación (educ) y la experiencia (exper).

Las variaciones en las variables independientes: educación (educ), experiencia

(exper), antigüedad (tenure), explican el 31.21% de las variaciones en el logaritmo
natural del salario (Inwage).
 . regress ingasto_mon_mens intot_integ iningtrab_mens innegocio_mens intransfer_mens

Source SS df MS Number of obs = 66

F(4, 61) = 28.39
Model 21.5323133 4 5.38307833 Prob > F = 0.0000
Residual 11.5661223 61 .189608563 R-squared = 0.6506
Adj R-squared = 0.6276
Total 33.0984356 65 .509206702 Root MSE = .43544

ingasto_mon_m~s Coef. Std. Err. t P>|t| [95% Conf. Interval]

intot_integ .1041588 .1027987 1.01 0.315 -.1014 .3097176

iningtrab_mens .5376113 .0636243 8.45 0.000 .4103867 .6648359
innegocio_mens -.1040556 .0424439 -2.45 0.017 -.1889274 -.0191838
intransfer_mens .0865578 .0424329 2.04 0.046 .001708 .1714077
_cons 4.422476 .6104172 7.25 0.000 3.201872 5.643081

. estimates store M4, title(modelo4)

3. Con el criterio de Akaike (AIC) et el
Criterio de información Bayesiana (BIC),
exponga ¿cuál de los modelos D, E, F, G es el
“mejor” modelo?
Criterios de Akaike (AIC) y Criterio de Información Bayesiana (BIC)

. estimates table M1 M2 M3 M4

Variable M1 M2 M3 M4

tot_integ 506.28905 .08154379

ingtrab_mens .47061486 .0000386
negocio_mens -.01259505 -5.280e-06
transfer_m~s .17103144 .00001213
intot_integ 499.74841 .10415878
iningtrab_~s 5629.0509 .53761131
innegocio_~s -1450.3411 -.10405556
intransfer~s 1035.0076 .08655782
_cons 3944.6231 8.3786931 -35099.798 4.4224761
. estimates table M1 M2 M3 M4, b(%8.0g)se(%8.0g)t(%8.0g)p(%8.0g)

Variable M1 M2 M3 M4

tot_integ 506.289 .081544

192.549 .016978
2.6294 4.80278
.008868 2.2e-06
ingtrab_mens .470615 .000039
.032852 2.9e-06
14.3252 13.3412
3.5e-38 4.4e-34
negocio_mens -.012595 -5.3e-06
.095868 8.4e-06
-.13138 -.625722
.895538 .531838
transfer_m~s .171031 .000012
.028599 2.5e-06
5.98026 4.81679
4.8e-09 2.0e-06
intot_integ 499.748 .104159
1324.24 .102799
.377385 1.01323
.707197 .314952
iningtrab_~s 5629.05 .537611
819.599 .063624
6.86806 8.44978
3.9e-09 7.4e-12
innegocio_~s -1450.34 -.104056
546.756 .042444
-2.65263 -2.4516
.010167 .017101
intransfer~s 1035.01 .086558
546.615 .042433
1.89349 2.03988
.063039 .0457
_cons 3944.62 8.37869 -35099.8 4.42248
718.726 .063522 7863.31 .610417
5.48836 131.903 -4.46374 7.24501
7.0e-08 0 .000035 8.8e-10

legend: b/se/t/p
. estimates table M1 M2 M3 M4, b(%8.0g)se(%8.0g)t(%8.0g)p(%8.0g) stats(r2, r2_a, aic, bic)

Variable M1 M2 M3 M4

tot_integ 506.289 .081544

192.549 .016978
2.6294 4.80278
.008868 2.2e-06
ingtrab_mens .470615 .000039
.032852 2.9e-06
14.3252 13.3412
3.5e-38 4.4e-34
negocio_mens -.012595 -5.3e-06
.095868 8.4e-06
-.13138 -.625722
.895538 .531838
transfer_m~s .171031 .000012
.028599 2.5e-06
5.98026 4.81679
4.8e-09 2.0e-06
intot_integ 499.748 .104159
1324.24 .102799
.377385 1.01323
.707197 .314952
iningtrab_~s 5629.05 .537611
819.599 .063624
6.86806 8.44978
3.9e-09 7.4e-12
innegocio_~s -1450.34 -.104056
546.756 .042444
-2.65263 -2.4516
.010167 .017101
intransfer~s 1035.01 .086558
546.615 .042433
1.89349 2.03988
.063039 .0457
_cons 3944.62 8.37869 -35099.8 4.42248
718.726 .063522 7863.31 .610417
5.48836 131.903 -4.46374 7.24501
7.0e-08 0 .000035 8.8e-10

r2 .412569 .408021 .534623 .650554

r2_a .406974 .40237 .504107 .627639
aic 8621.21 682.408 1331.55 82.356
bic 8641.47 702.657 1342.5 93.3042

legend: b/se/t/p