www.cmm.bristol.ac.uk/forum

Hi,

I am currently running gradually more complex models using MCMC and comparing the DIC to see whether there is an improvement. I have a very simple question about this. If the preceding model, with which I am comparing, was run for 150,000 iterations, but the following models are only run for 50,000 iterations, are they comparable, or do I need to set all the models up to run for exactly the same number of iterations?

Thanks.

As long as the chains have converged to a stable distribution I believe that the DIC should be comparable. You can find more information on how the DIC is calculated in section 2.2 of the MCMC manual.

Great. Thanks, Chris.