2019_Factores predictores exito academico

packages

library("tidyverse")
library("MASS")
library("stargazer")
library("broom")
library("nnet")

las variables

#masculino--0, femenino--1
#no padres--0, con padres--1
#particular..0, estatal--1
#Historial..no--0, si--1
#trabajo, no--0, si--1 
#660 punto corte psu
#60 NEM  

abro df para cuarto y quinto

modelo de regresion para cuarto y quinto

modelo1 <- glm(df1$APROBACION ~ df1$curso1 + df1$Sexo1 + df1$Residencia1 + df1$Establecimiento1 + +df1$Historial1 + df1$Trabajo1 + df1$PSU_lenguaje_1 + df1$PSU_matematicas_1 + df1$PSU_ciencias_1 + df1$NEM_1,
               data=df1,
                family = binomial(logit))

modelo5

modelo1

Call:  glm(formula = df1$APROBACION ~ df1$curso1 + df1$Sexo1 + df1$Residencia1 + 
    df1$Establecimiento1 + +df1$Historial1 + df1$Trabajo1 + df1$PSU_lenguaje_1 + 
    df1$PSU_matematicas_1 + df1$PSU_ciencias_1 + df1$NEM_1, family = binomial(logit), 
    data = df1)

Coefficients:
          (Intercept)             df1$curso1              df1$Sexo1        df1$Residencia1  
              0.91634                0.03926                0.95975                0.10193  
 df1$Establecimiento1         df1$Historial1           df1$Trabajo1     df1$PSU_lenguaje_1  
              0.69021               -0.58776               -0.53823               -0.62442  
df1$PSU_matematicas_1     df1$PSU_ciencias_1              df1$NEM_1  
              0.12887               -0.25453               -1.21279  

Degrees of Freedom: 118 Total (i.e. Null);  108 Residual
Null Deviance:      161.2 
Residual Deviance: 149.5    AIC: 171.5

summary modelo5

summary(modelo1)

Call:
glm(formula = df1$APROBACION ~ df1$curso1 + df1$Sexo1 + df1$Residencia1 + 
    df1$Establecimiento1 + +df1$Historial1 + df1$Trabajo1 + df1$PSU_lenguaje_1 + 
    df1$PSU_matematicas_1 + df1$PSU_ciencias_1 + df1$NEM_1, family = binomial(logit), 
    data = df1)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-1.5327  -1.0045  -0.7454   1.1487   1.8867  

Coefficients:
                      Estimate Std. Error z value Pr(>|z|)  
(Intercept)            0.91634    1.07766   0.850   0.3952  
df1$curso1             0.03926    0.41219   0.095   0.9241  
df1$Sexo1              0.95975    0.41749   2.299   0.0215 *
df1$Residencia1        0.10193    0.42386   0.240   0.8100  
df1$Establecimiento1   0.69021    0.48337   1.428   0.1533  
df1$Historial1        -0.58776    0.48407  -1.214   0.2247  
df1$Trabajo1          -0.53823    0.49809  -1.081   0.2799  
df1$PSU_lenguaje_1    -0.62442    0.40769  -1.532   0.1256  
df1$PSU_matematicas_1  0.12887    0.44454   0.290   0.7719  
df1$PSU_ciencias_1    -0.25453    0.42230  -0.603   0.5467  
df1$NEM_1             -1.21279    0.93964  -1.291   0.1968  
---
Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 161.24  on 118  degrees of freedom
Residual deviance: 149.50  on 108  degrees of freedom
AIC: 171.5

Number of Fisher Scoring iterations: 4

tabla intervalos para cuarto y quinto

stargazer::stargazer(modelo1,  type = "text",
                     no.space=FALSE, 
                     apply.coef = exp, 
                     # apply.se = exp, 
                     ci=TRUE, ci.level=0.95, single.row=TRUE, 
                     omit.table.layout = "n", star.cutoffs = NA) # omit p values FTW!

=======================================
                   Dependent variable: 
                  ---------------------
                       APROBACION      
---------------------------------------
curso1            1.040 (0.232, 1.848) 
                                       
Sexo1             2.611 (1.793, 3.429) 
                                       
Residencia1       1.107 (0.277, 1.938) 
                                       
Establecimiento1  1.994 (1.047, 2.942) 
                                       
Historial1        0.556 (-0.393, 1.504)
                                       
Trabajo1          0.584 (-0.392, 1.560)
                                       
PSU_lenguaje_1    0.536 (-0.263, 1.335)
                                       
PSU_matematicas_1 1.138 (0.266, 2.009) 
                                       
PSU_ciencias_1    0.775 (-0.052, 1.603)
                                       
NEM_1             0.297 (-1.544, 2.139)
                                       
Constant          2.500 (0.388, 4.612) 
                                       
---------------------------------------
Observations               119         
Log Likelihood           -74.752       
Akaike Inf. Crit.        171.503       
=======================================

lineal para cuarto y quinto

summary

summary(modelo7)

Call:
lm(formula = df1$APROBACION ~ df1$curso1 + df1$Sexo1 + df1$Residencia1 + 
    df1$Establecimiento1 + +df1$Historial1 + df1$Trabajo1 + df1$PSU_lenguaje_1 + 
    df1$PSU_matematicas_1 + df1$PSU_ciencias_1 + df1$NEM_1, data = df1)

Residuals:
    Min      1Q  Median      3Q     Max 
-0.6749 -0.4087 -0.2504  0.4855  0.8521 

Coefficients:
                       Estimate Std. Error t value Pr(>|t|)   
(Intercept)            0.699235   0.246098   2.841  0.00537 **
df1$curso1             0.008301   0.094396   0.088  0.93009   
df1$Sexo1              0.211639   0.093099   2.273  0.02499 * 
df1$Residencia1        0.015494   0.096550   0.160  0.87281   
df1$Establecimiento1   0.151505   0.110898   1.366  0.17472   
df1$Historial1        -0.129899   0.109673  -1.184  0.23885   
df1$Trabajo1          -0.109151   0.109689  -0.995  0.32191   
df1$PSU_lenguaje_1    -0.138807   0.093255  -1.488  0.13954   
df1$PSU_matematicas_1  0.030032   0.101182   0.297  0.76718   
df1$PSU_ciencias_1    -0.054996   0.096901  -0.568  0.57152   
df1$NEM_1             -0.265990   0.211955  -1.255  0.21221   
---
Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

Residual standard error: 0.4917 on 108 degrees of freedom
Multiple R-squared:  0.09405,   Adjusted R-squared:  0.01017 
F-statistic: 1.121 on 10 and 108 DF,  p-value: 0.3532

abro df para cuarto

df2 <- read.csv("https://docs.google.com/spreadsheets/d/e/2PACX-1vTqhm3ewZFirDV6BHKxF1mqpgADOgt629ND3swbKXGrE559HtV79QEtYJkNTUjIGge1ZtCaJ-cHPSN4/pub?gid=1815207712&single=true&output=csv")

modelo de regresion para cuarto

modelo_cuarto <- glm(df2$APROBACION ~ df2$Sexo1 + df2$Residencia1 + df2$Establecimiento1 + +df2$Historial1 + df2$Trabajo1 + df2$PSU_lenguaje_1 + df2$PSU_matematicas_1 + df2$PSU_ciencias_1 + df2$NEM_1,
               data=df2,
                family = binomial(logit))

modelo_cuarto

modelo_cuarto

Call:  glm(formula = df2$APROBACION ~ df2$Sexo1 + df2$Residencia1 + 
    df2$Establecimiento1 + +df2$Historial1 + df2$Trabajo1 + df2$PSU_lenguaje_1 + 
    df2$PSU_matematicas_1 + df2$PSU_ciencias_1 + df2$NEM_1, family = binomial(logit), 
    data = df2)

Coefficients:
          (Intercept)              df2$Sexo1        df2$Residencia1   df2$Establecimiento1  
             -0.08952                1.62037                0.07039                0.94019  
       df2$Historial1           df2$Trabajo1     df2$PSU_lenguaje_1  df2$PSU_matematicas_1  
              0.41168               -0.88203               -1.25392                0.26094  
   df2$PSU_ciencias_1              df2$NEM_1  
              0.22690               -1.19419  

Degrees of Freedom: 69 Total (i.e. Null);  60 Residual
Null Deviance:      94.97 
Residual Deviance: 81.44    AIC: 101.4

summary modelo_cuarto

summary(modelo_cuarto)

Call:
glm(formula = df2$APROBACION ~ df2$Sexo1 + df2$Residencia1 + 
    df2$Establecimiento1 + +df2$Historial1 + df2$Trabajo1 + df2$PSU_lenguaje_1 + 
    df2$PSU_matematicas_1 + df2$PSU_ciencias_1 + df2$NEM_1, family = binomial(logit), 
    data = df2)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-1.6412  -0.9468  -0.5136   1.1210   1.7043  

Coefficients:
                      Estimate Std. Error z value Pr(>|z|)   
(Intercept)           -0.08952    1.52324  -0.059  0.95313   
df2$Sexo1              1.62037    0.61558   2.632  0.00848 **
df2$Residencia1        0.07039    0.56072   0.126  0.90010   
df2$Establecimiento1   0.94019    0.70171   1.340  0.18029   
df2$Historial1         0.41168    0.69413   0.593  0.55312   
df2$Trabajo1          -0.88203    0.69862  -1.263  0.20676   
df2$PSU_lenguaje_1    -1.25392    0.62155  -2.017  0.04365 * 
df2$PSU_matematicas_1  0.26094    0.65398   0.399  0.68989   
df2$PSU_ciencias_1     0.22690    0.58792   0.386  0.69955   
df2$NEM_1             -1.19419    1.35562  -0.881  0.37836   
---
Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 94.973  on 69  degrees of freedom
Residual deviance: 81.436  on 60  degrees of freedom
AIC: 101.44

Number of Fisher Scoring iterations: 4

tabla intervalos para cuarto ao

stargazer::stargazer(modelo_cuarto,  type = "text",
                     no.space=FALSE, 
                     apply.coef = exp, 
                     # apply.se = exp, 
                     ci=TRUE, ci.level=0.95, single.row=TRUE, 
                     omit.table.layout = "n", star.cutoffs = NA) # omit p values FTW!

=======================================
                   Dependent variable: 
                  ---------------------
                       APROBACION      
---------------------------------------
Sexo1             5.055 (3.848, 6.261) 
                                       
Residencia1       1.073 (-0.026, 2.172)
                                       
Establecimiento1  2.560 (1.185, 3.936) 
                                       
Historial1        1.509 (0.149, 2.870) 
                                       
Trabajo1          0.414 (-0.955, 1.783)
                                       
PSU_lenguaje_1    0.285 (-0.933, 1.504)
                                       
PSU_matematicas_1 1.298 (0.016, 2.580) 
                                       
PSU_ciencias_1    1.255 (0.102, 2.407) 
                                       
NEM_1             0.303 (-2.354, 2.960)
                                       
Constant          0.914 (-2.071, 3.900)
                                       
---------------------------------------
Observations               70          
Log Likelihood           -40.718       
Akaike Inf. Crit.        101.436       
=======================================

lineal para cuarto

summary

summary(modelo_6)

Call:
lm(formula = df2$APROBACION ~ df2$Sexo1 + df2$Residencia1 + df2$Establecimiento1 + 
    +df2$Historial1 + df2$Trabajo1 + df2$PSU_lenguaje_1 + df2$PSU_matematicas_1 + 
    df2$PSU_ciencias_1 + df2$NEM_1, data = df2)

Residuals:
    Min      1Q  Median      3Q     Max 
-0.7045 -0.3888 -0.1079  0.4720  0.7437 

Coefficients:
                      Estimate Std. Error t value Pr(>|t|)   
(Intercept)            0.45031    0.31681   1.421  0.16039   
df2$Sexo1              0.32456    0.12162   2.669  0.00978 **
df2$Residencia1        0.01424    0.12123   0.117  0.90692   
df2$Establecimiento1   0.17911    0.14868   1.205  0.23306   
df2$Historial1         0.08630    0.14483   0.596  0.55348   
df2$Trabajo1          -0.16836    0.14169  -1.188  0.23945   
df2$PSU_lenguaje_1    -0.24943    0.12578  -1.983  0.05194 . 
df2$PSU_matematicas_1  0.06072    0.13716   0.443  0.65959   
df2$PSU_ciencias_1     0.03902    0.12742   0.306  0.76046   
df2$NEM_1             -0.20999    0.27051  -0.776  0.44065   
---
Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

Residual standard error: 0.4832 on 60 degrees of freedom
Multiple R-squared:  0.1754,    Adjusted R-squared:  0.05166 
F-statistic: 1.418 on 9 and 60 DF,  p-value: 0.201

abro df para quinto ao

df3 <- read.csv("https://docs.google.com/spreadsheets/d/e/2PACX-1vTqhm3ewZFirDV6BHKxF1mqpgADOgt629ND3swbKXGrE559HtV79QEtYJkNTUjIGge1ZtCaJ-cHPSN4/pub?gid=1429924662&single=true&output=csv")

modelo de regresion para quinto

modelo_quinto <- glm(df3$APROBACION ~ df3$Sexo1 + df3$Residencia1 + df3$Establecimiento1 + +df3$Historial1 + df3$Trabajo1 + df3$PSU_lenguaje_1 + df3$PSU_matematicas_1 + df3$PSU_ciencias_1 + df3$NEM_1,
               data=df3,
                family = binomial(logit))

modelo_quinto

modelo_quinto

Call:  glm(formula = df3$APROBACION ~ df3$Sexo1 + df3$Residencia1 + 
    df3$Establecimiento1 + +df3$Historial1 + df3$Trabajo1 + df3$PSU_lenguaje_1 + 
    df3$PSU_matematicas_1 + df3$PSU_ciencias_1 + df3$NEM_1, family = binomial(logit), 
    data = df3)

Coefficients:
          (Intercept)              df3$Sexo1        df3$Residencia1   df3$Establecimiento1  
               2.4293                 0.4743                -0.2802                 0.5924  
       df3$Historial1           df3$Trabajo1     df3$PSU_lenguaje_1  df3$PSU_matematicas_1  
              -1.9875                 0.1227                -0.5973                 0.1703  
   df3$PSU_ciencias_1              df3$NEM_1  
              -0.4832                -1.3305  

Degrees of Freedom: 48 Total (i.e. Null);  39 Residual
Null Deviance:      66.27 
Residual Deviance: 56.85    AIC: 76.85

summary modelo quinto

summary(modelo_quinto)

Call:
glm(formula = df3$APROBACION ~ df3$Sexo1 + df3$Residencia1 + 
    df3$Establecimiento1 + +df3$Historial1 + df3$Trabajo1 + df3$PSU_lenguaje_1 + 
    df3$PSU_matematicas_1 + df3$PSU_ciencias_1 + df3$NEM_1, family = binomial(logit), 
    data = df3)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-1.4687  -0.8879  -0.6399   0.9731   1.9759  

Coefficients:
                      Estimate Std. Error z value Pr(>|z|)  
(Intercept)             2.4293     1.8134   1.340   0.1804  
df3$Sexo1               0.4743     0.7173   0.661   0.5085  
df3$Residencia1        -0.2802     0.7570  -0.370   0.7113  
df3$Establecimiento1    0.5924     0.7575   0.782   0.4342  
df3$Historial1         -1.9875     0.8606  -2.309   0.0209 *
df3$Trabajo1            0.1227     0.8386   0.146   0.8837  
df3$PSU_lenguaje_1     -0.5973     0.6973  -0.857   0.3917  
df3$PSU_matematicas_1   0.1703     0.6857   0.248   0.8039  
df3$PSU_ciencias_1     -0.4832     0.6824  -0.708   0.4789  
df3$NEM_1              -1.3305     1.6365  -0.813   0.4162  
---
Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 66.266  on 48  degrees of freedom
Residual deviance: 56.847  on 39  degrees of freedom
AIC: 76.847

Number of Fisher Scoring iterations: 4

tabla intervalos para quinto

stargazer::stargazer(modelo_quinto,  type = "text",
                     no.space=FALSE, 
                     apply.coef = exp, 
                     # apply.se = exp, 
                     ci=TRUE, ci.level=0.95, single.row=TRUE, 
                     omit.table.layout = "n", star.cutoffs = NA) # omit p values FTW!

========================================
                   Dependent variable:  
                  ----------------------
                        APROBACION      
----------------------------------------
Sexo1              1.607 (0.201, 3.013) 
                                        
Residencia1       0.756 (-0.728, 2.239) 
                                        
Establecimiento1   1.808 (0.324, 3.293) 
                                        
Historial1        0.137 (-1.550, 1.824) 
                                        
Trabajo1          1.131 (-0.513, 2.774) 
                                        
PSU_lenguaje_1    0.550 (-0.816, 1.917) 
                                        
PSU_matematicas_1 1.186 (-0.158, 2.529) 
                                        
PSU_ciencias_1    0.617 (-0.721, 1.954) 
                                        
NEM_1             0.264 (-2.943, 3.472) 
                                        
Constant          11.351 (7.797, 14.905)
                                        
----------------------------------------
Observations                49          
Log Likelihood           -28.423        
Akaike Inf. Crit.         76.847        
========================================

modelo lineal quinto

modelo_lineal <- lm(df3$APROBACION ~ df3$Sexo1 + df3$Residencia1 + df3$Establecimiento1 + +df3$Historial1 + df3$Trabajo1 + df3$PSU_lenguaje_1 + df3$PSU_matematicas_1 + df3$PSU_ciencias_1 + df3$NEM_1,
               data=df3)

summary(modelo_lineal)

Call:
lm(formula = df3$APROBACION ~ df3$Sexo1 + df3$Residencia1 + df3$Establecimiento1 + 
    +df3$Historial1 + df3$Trabajo1 + df3$PSU_lenguaje_1 + df3$PSU_matematicas_1 + 
    df3$PSU_ciencias_1 + df3$NEM_1, data = df3)

Residuals:
    Min      1Q  Median      3Q     Max 
-0.6401 -0.3273 -0.1875  0.3791  0.8722 

Coefficients:
                      Estimate Std. Error t value Pr(>|t|)  
(Intercept)            0.98574    0.38543   2.557   0.0145 *
df3$Sexo1              0.08369    0.14982   0.559   0.5796  
df3$Residencia1       -0.05612    0.16262  -0.345   0.7319  
df3$Establecimiento1   0.12286    0.16810   0.731   0.4692  
df3$Historial1        -0.42169    0.17996  -2.343   0.0243 *
df3$Trabajo1           0.01922    0.18260   0.105   0.9167  
df3$PSU_lenguaje_1    -0.11583    0.15155  -0.764   0.4493  
df3$PSU_matematicas_1  0.03302    0.15280   0.216   0.8300  
df3$PSU_ciencias_1    -0.10451    0.15210  -0.687   0.4961  
df3$NEM_1             -0.24899    0.34990  -0.712   0.4809  
---
Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

Residual standard error: 0.4985 on 39 degrees of freedom
Multiple R-squared:  0.1811,    Adjusted R-squared:  -0.007917 
F-statistic: 0.9581 on 9 and 39 DF,  p-value: 0.4883