Analisis Diskriminan Menggunakan NCSS

Masukkan datanya
Analysis > Multivariate Analysis . Dicriminan Analysis
Klik Menu Variable
 

 Pada Y; Group Variable: pilih Y

 Pada Independen Variable: pilih X1-X2
 Pada estimation method: pilih linier Discriminant Function
 Klik Menu Report

 Cek listL variable influence, linier discrim coefficient, classification rows, canonical varaite analysis, canonical coeffisien, std canonical coeffieent dan variable variate correlation

 
Klik Run
 Berikut hasilnya:
 
Discriminant Analysis Report
 Dataset                   Untitled
 Dependent              y
 Prior Prob's             Equal
 Variable Influence Section
                               Removed    Removed    Removed    Alone         Alone       Alone            R-Squared
 Variable                  Lambda      F-Value       F-Prob        Lambda      F-Value    F-Prob          Other X's
 x1                           0.072602217.15            0.000000     0.074050225.08          0.000000       0.115754
 x2                           0.785506     4.64            0.045831     0.801176     4.47          0.048779       0.115754
 Linear Discriminant Functions
                                              y
 
Variable                  0                      1
 Constant                 -52.05259          -158.8457
  x1                           19.88132           34.65775
 x2                           1.413607           2.660076
 Regression Coefficients
                               y
 Variable                  0                      1
Constant                 2.225498           -1.225498
x1                           -0.2387484        0.2387484
 x2                           -0.02013968      0.02013968
 Classification Count Table for y
                               Predicted
 Actual                     0                      1                       Total
 0                             10                     0                       10
 1                             0                      10                     10
 Total                       10                     10                     20
 Reduction in classification error due to X's = 100.0%
 Reduction in classification error due to X's = x1, x2
 
Canonical Variate Analysis Section
             Inv(W)B            Ind'l     Total     Canon     Canon                    Numer   Denom      Prob       Wilks'
 Fn        Eigenvalue       Pcnt     Pcnt     Corr        Corr2      F-Value    DF         DF             Level      Lambda
116.192002                100.0    100.0       0.9705     0.9418      137.6     2.0          17.0            0.0000     0.058167
 The F-value tests whether this function and those below it are significant.
 Canonical Coefficients
                               Canonical Variate
 Variable                  Variate1
 Constant                 -13.987558
 x1                           1.935387
 x2                           0.163260
                                                                                                                                                                     
 
Discriminant Analysis Report
 
  Dataset                   Untitled
 Dependent              y
 Prior Prob's             Equal
 Std. Canonical Coefficients
                               Canonical Variate
 Variable                  Variate1
 x1                          1.065733
 x2                           0.512534
 Variable-Variate Correlations
                               Canonical Variate
Variable                  Variate1
x1                           0.878783
x2                           0.123800
Hasil dan pembahasan sama dengan penggunaan software lainnya

Search