Skip to main content

Table 1 Model and parameters in simulation study

From: A comparison of time dependent Cox regression, pooled logistic regression and cross sectional pooling with simulations and an application to the Framingham Heart Study

Longitudinal model

Y ij  = Ui1 + Ui2 * t ij  + τ * Age + ε ij

Survival model

h(t) = λ(t)exp{α1Age + α2Sex + γY ij }

Covariance matrix for random effects (Ui1 , Ui2)

\( G = \left[\begin{array}{cc}\hfill 0.29\hfill & \hfill -0.00465\hfill \\ {}\hfill -0.00465\hfill & \hfill 0.000320\hfill \end{array}\right] \)

# of exams

Random effects (Ui1 , Ui2)

Residual error (σ2)

Age (α1)

Sex (α2)

Link (γ)

6

(4.250, 0.250)

0.116

0.050

−0.500

(0.000, 0.500, 1.000)

  1. Y ij : Observed Longitudinal Measures; λ(t): Baseline Hazard Function; h(t): Hazard Function