On the applicability of the 'number of possible states' argument in multi-expert reasoning

The aim of this paper is to explore the applicability of the 'number of possible states' argument in inferential problems in multi-expert reasoning. The argument is Bayesian and it is similar in spirit to the one used to derive the maximum entropy inference process. Under certain conditions a particular way of applying it surprisingly suggests that the weighted arithmetic mean should be used in meta-analysis with unexplained heterogeneity.