TSTW Thesis

Santiago Torres

6 de mayo de 2018

Modelo econométrico

Modelo

\[\lambda(t,x)=lim_{dt0}\frac{Pr(t<T<t+dt|T>t,x)}{dt}=\frac{g(t,x)}{1-G(t,x)}\] -\[lnT_s=min(x_s,+u_1,ln\tau_s)\] -\[S=Y\]

Resultados

## Loading required package: rJava
## Loading required package: xlsxjars

Plots

Correlation

## 
## Call:
## lm(formula = x$lnTs ~ x$edad_st + x$v_mujer * x$v_hijos + x$padre_instituto + 
##     x$padre_universidad + x$madre_instituto + x$madre_universidad + 
##     x$instituto + x$universidad, data = x)
## 
## Residuals:
##      Min       1Q   Median       3Q      Max 
## -2.26464 -1.00884 -0.05874  0.93290  3.01621 
## 
## Coefficients: (1 not defined because of singularities)
##                            Estimate Std. Error t value Pr(>|t|)   
## (Intercept)               -0.918250   1.666030  -0.551  0.58178   
## x$edad_st                 -0.007016   0.003899  -1.799  0.07258 . 
## x$v_mujer1                 0.774613   1.375107   0.563  0.57349   
## x$v_mujerHombre            4.560001   1.666242   2.737  0.00644 **
## x$v_mujerMujer             4.530070   1.665231   2.720  0.00676 **
## x$v_hijos                  0.443958   0.387449   1.146  0.25243   
## x$padre_instituto          0.070685   0.077514   0.912  0.36228   
## x$padre_universidad        0.231208   0.216186   1.069  0.28539   
## x$madre_instituto         -0.001865   0.002927  -0.637  0.52439   
## x$madre_universidad        0.097888   0.348390   0.281  0.77885   
## x$instituto                0.186817   0.169955   1.099  0.27223   
## x$universidad              0.009561   0.160392   0.060  0.95249   
## x$v_mujer1:x$v_hijos             NA         NA      NA       NA   
## x$v_mujerHombre:x$v_hijos -0.371605   0.393485  -0.944  0.34544   
## x$v_mujerMujer:x$v_hijos  -0.346604   0.394491  -0.879  0.38005   
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 1.246 on 478 degrees of freedom
## Multiple R-squared:  0.3701, Adjusted R-squared:  0.3529 
## F-statistic:  21.6 on 13 and 478 DF,  p-value: < 2.2e-16

Output

## Version:  1.36.23
## Date:     2017-03-03
## Author:   Philip Leifeld (University of Glasgow)
## 
## Please cite the JSS article in your publications -- see citation("texreg").
## 
## ====================================
##                            Model 1  
## ------------------------------------
## (Intercept)                 -0.92   
##                             (1.67)  
## x$edad_st                   -0.01   
##                             (0.00)  
## x$v_mujer1                   0.77   
##                             (1.38)  
## x$v_mujerHombre              4.56 **
##                             (1.67)  
## x$v_mujerMujer               4.53 **
##                             (1.67)  
## x$v_hijos                    0.44   
##                             (0.39)  
## x$padre_instituto            0.07   
##                             (0.08)  
## x$padre_universidad          0.23   
##                             (0.22)  
## x$madre_instituto           -0.00   
##                             (0.00)  
## x$madre_universidad          0.10   
##                             (0.35)  
## x$instituto                  0.19   
##                             (0.17)  
## x$universidad                0.01   
##                             (0.16)  
## x$v_mujerHombre:x$v_hijos   -0.37   
##                             (0.39)  
## x$v_mujerMujer:x$v_hijos    -0.35   
##                             (0.39)  
## ------------------------------------
## R^2                          0.37   
## Adj. R^2                     0.35   
## Num. obs.                  492      
## RMSE                         1.25   
## ====================================
## *** p < 0.001, ** p < 0.01, * p < 0.05

Cox proportional hazard model

## Call:
## coxph(formula = Surv(x$w_inicio_st, x$w_fin_st, x$fail) ~ x$edad_st + 
##     x$padre_instituto + x$padre_universidad + x$madre_instituto + 
##     x$madre_universidad + x$instituto + x$universidad, data = x)
## 
##   n= 492, number of events= 479 
## 
##                          coef exp(coef)  se(coef)      z Pr(>|z|)   
## x$edad_st            0.010957  1.011018  0.003811  2.875  0.00403 **
## x$padre_instituto   -0.081898  0.921366  0.229696 -0.357  0.72143   
## x$padre_universidad -0.343643  0.709182  0.173512 -1.981  0.04765 * 
## x$madre_instituto    0.004065  1.004073  0.002811  1.446  0.14819   
## x$madre_universidad -0.424628  0.654013  0.283276 -1.499  0.13388   
## x$instituto         -0.445242  0.640669  0.145553 -3.059  0.00222 **
## x$universidad       -0.137782  0.871289  0.131940 -1.044  0.29636   
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
##                     exp(coef) exp(-coef) lower .95 upper .95
## x$edad_st              1.0110     0.9891    1.0035    1.0186
## x$padre_instituto      0.9214     1.0853    0.5874    1.4453
## x$padre_universidad    0.7092     1.4101    0.5047    0.9964
## x$madre_instituto      1.0041     0.9959    0.9986    1.0096
## x$madre_universidad    0.6540     1.5290    0.3754    1.1395
## x$instituto            0.6407     1.5609    0.4817    0.8522
## x$universidad          0.8713     1.1477    0.6728    1.1284
## 
## Concordance= 0.613  (se = 0.017 )
## Rsquare= 0.055   (max possible= 0.998 )
## Likelihood ratio test= 27.7  on 7 df,   p=0.0002491
## Wald test            = 24.43  on 7 df,   p=0.000956
## Score (logrank) test = 26.37  on 7 df,   p=0.0004328

KM

Densidades

Densidades