The information content of a multistate distribution

D. M. Boulton, C.S.Wallace

Journal of Theoretical Biology, Vol.23, No.2, pp.269-278, May 1969

here (click)

In the literature we find two different expressions in use for finding the information content of a population whose members each occupy one of a number of states. The information content of a population is equal to the optimum length of a message which fully describes its members, and the two forms, which are not equivalent, differ in the amount of information which a receiver of the message need know beforehand, in order to be able to decode it. We believe that terms measuring this prior information should be added and show that when this is done the resulting expressions are essentially equivalent.