2015-05-29

R: Tabulación cruzada y prueba chi-cuadrado con gmodels

Title

Problema

En R deseamos crear una tabla de referencias cruzadas similar a la producida en SAS (SAS Proc Freq) o crosstabs SPSS.

Solución

Utilizamos la función CrossTable del paquete gmodels.

  • SAS
  • CrossTable(mtcars$am, mtcars$gear, dnn = c("am", "gear"), expected = TRUE, format = "SAS")
    
       Cell Contents
    |-------------------------|
    |                       N |
    |              Expected N |
    | Chi-square contribution |
    |           N / Row Total |
    |           N / Col Total |
    |         N / Table Total |
    |-------------------------|
    
     
    Total Observations in Table:  32 
    
     
                 | gear 
              am |         3 |         4 |         5 | Row Total | 
    -------------|-----------|-----------|-----------|-----------|
               0 |        15 |         4 |         0 |        19 | 
                 |     8.906 |     7.125 |     2.969 |           | 
                 |     4.169 |     1.371 |     2.969 |           | 
                 |     0.789 |     0.211 |     0.000 |     0.594 | 
                 |     1.000 |     0.333 |     0.000 |           | 
                 |     0.469 |     0.125 |     0.000 |           | 
    -------------|-----------|-----------|-----------|-----------|
               1 |         0 |         8 |         5 |        13 | 
                 |     6.094 |     4.875 |     2.031 |           | 
                 |     6.094 |     2.003 |     4.339 |           | 
                 |     0.000 |     0.615 |     0.385 |     0.406 | 
                 |     0.000 |     0.667 |     1.000 |           | 
                 |     0.000 |     0.250 |     0.156 |           | 
    -------------|-----------|-----------|-----------|-----------|
    Column Total |        15 |        12 |         5 |        32 | 
                 |     0.469 |     0.375 |     0.156 |           | 
    -------------|-----------|-----------|-----------|-----------|
    
     
    Statistics for All Table Factors
    
    
    Pearson's Chi-squared test 
    ------------------------------------------------------------
    Chi^2 =  20.94467     d.f. =  2     p =  2.830889e-05 
    
  • SPSS
  • CrossTable(mtcars$am, mtcars$gear, dnn = c("am", "gear"), expected = TRUE, format = "SPSS")
    
    
       Cell Contents
    |-------------------------|
    |                   Count |
    |         Expected Values |
    | Chi-square contribution |
    |             Row Percent |
    |          Column Percent |
    |           Total Percent |
    |-------------------------|
    
    Total Observations in Table:  32 
    
                 | gear 
              am |        3  |        4  |        5  | Row Total | 
    -------------|-----------|-----------|-----------|-----------|
               0 |       15  |        4  |        0  |       19  | 
                 |    8.906  |    7.125  |    2.969  |           | 
                 |    4.169  |    1.371  |    2.969  |           | 
                 |   78.947% |   21.053% |    0.000% |   59.375% | 
                 |  100.000% |   33.333% |    0.000% |           | 
                 |   46.875% |   12.500% |    0.000% |           | 
    -------------|-----------|-----------|-----------|-----------|
               1 |        0  |        8  |        5  |       13  | 
                 |    6.094  |    4.875  |    2.031  |           | 
                 |    6.094  |    2.003  |    4.339  |           | 
                 |    0.000% |   61.538% |   38.462% |   40.625% | 
                 |    0.000% |   66.667% |  100.000% |           | 
                 |    0.000% |   25.000% |   15.625% |           | 
    -------------|-----------|-----------|-----------|-----------|
    Column Total |       15  |       12  |        5  |       32  | 
                 |   46.875% |   37.500% |   15.625% |           | 
    -------------|-----------|-----------|-----------|-----------|
    
     
    Statistics for All Table Factors
    
    
    Pearson's Chi-squared test 
    ------------------------------------------------------------
    Chi^2 =  20.94467     d.f. =  2     p =  2.830889e-05 
    
    
     
           Minimum expected frequency: 2.03125 
    Cells with Expected Frequency < 5: 3 of 6 (50%)
    

    Notas

    Argumentos:

    expected - si es TRUE para incluir la prueba chi-cuadrado de Pearson.
    format - indicamos el formato de la salida de datos, SAS —por defecto— o SPSS.
    dnn - nombres de las dimensiones en el resultado.

    Referencias

    No hay comentarios:

    Publicar un comentario

    Nube de datos