Skip to main content

Table 5 Model performance and comparison using simulation study (n=100) with about 40% censored observations

From: Generalized log-logistic proportional hazard model with applications in survival analysis

     Generalized    
  Parameter    log-logistic PH   Weibull PH   Cox PH
Scenarios   True   Mean AB MSE   Mean AB MSE   Mean AB MSE
Scenario 1 β 1 0.50   0.524 0.024 0.024   0.531 0.031 0.024   0.524 0.024 0.025
(True model: β 2 −0.50   −0.538 0.038 0.032   −0.545 0.045 0.033   −0.536 0.036 0.032
generalized β 3 0.75   0.798 0.048 0.088   0.808 0.058 0.090   0.795 0.045 0.091
Weibull) β 4 −0.75   −0.780 0.030 0.081   −0.792 0.042 0.083   −0.782 0.032 0.083
  ρ 0.10   0.148     0.103       
  κ 0.50   0.550     0.508       
  γ −0.10   0.073          
Scenario 2 β 1 0.50   0.516 0.016 0.027   0.518 0.018 0.027   0.532 0.032 0.031
(True model: β 2 −0.50   −0.522 0.022 0.035   −0.523 0.023 0.035   −0.533 0.033 0.039
generalized β 3 0.75   0.785 0.035 0.099   0.788 0.038 0.099   0.813 0.063 0.112
Weibull) β 4 −0.75   −0.765 0.015 0.082   −0.768 0.018 0.082   −0.796 0.046 0.092
  ρ 0.10   0.107     0.106       
  κ 2.00   2.269     2.249       
  γ 0.10   0.006           
Scenario 3 β 1 0.50   0.519 0.019 0.025   0.530 0.030 0.026   0.516 0.016 0.026
(True model: β 2 −0.50   −0.548 0.048 0.038   −0.557 0.057 0.039   −0.547 0.047 0.039
generalized β 3 0.75   0.791 0.041 0.099   0.811 0.061 0.103   0.791 0.041 0.102
Weibull) β 4 −0.75   −0.790 0.040 0.089   −0.811 0.061 0.093   −0.792 0.042 0.095
  ρ 0.10   0.103     0.098       
  κ 2.00   2.130     2.016       
  γ −0.10   0.024