confidence interval on the boundary

questions concerning analysis/theory using programs M-SURGE, E-SURGE and U-CARE

confidence interval on the boundary

Postby Gourlay » Fri Feb 17, 2012 9:17 am

Hi,

In my model, I have some transition parmeters estimated at the boundary and therefore E-Sugre failed to compute wald-type confidence intervals.

How can I deal with this with E-Surge ?
I found a method using a combination of the venzon and moolgavkar's algorithm and a cubic interpolation method (Gimenez et al 2005), but I'm not statistician and I don't know how apply this with E-Surge ???

Thanks in advance for your answer.

Marie-Lucile
Gourlay
 
Posts: 16
Joined: Wed Feb 09, 2011 5:32 am

Re: confidence interval on the boundary

Postby CHOQUET » Mon Feb 20, 2012 4:38 am

Hello Marie-Lucile,
The only mean to get CI for a boundary parameter is to use non-parametric bootstrap.
However in practice, the existence of local minima in mutli-state model leads to
an unstable algorithm.
So, I have no practical solution for that kind of problem.
Sincerely,
Rémi Choquet
CHOQUET
 
Posts: 211
Joined: Thu Nov 24, 2005 4:58 am
Location: CEFE, Montpellier, FRANCE.


Return to analysis help

Who is online

Users browsing this forum: No registered users and 10 guests

cron