drizopoulos
commented
2 years ago Regarding the cumulative association, the formulation used in JMbayes and JMbayes2 is not exactly the same. In the latter the integral is divided by t; see here for more info.
Regarding the value, have you checked whether the model fitted with jm() has converged?
Hello! First, I want to thank you for the great and helpful work!! I used JMbayes package before in a project, and now I started using JMbayes2 since I need to used more than one longitudinal markers. Since I am using almost the same data in both projects, I have noticed that jm function gives similar results to jointModelBayes, but with a difference by factor of 10 in opposite direction for current value and cumulative association structure. For the current value, using jointModelBayes, I got α =0.13, Whereas when using jm α= 0.013. The value obtained by jointModelBayes is the one closer to what is known in the literature in my case. Whereas for the cumulative association structure, using jointModelBayes, α =0.01, while using jm α= 0.18. I would appreciate your feedback on this problem.