ABSTRACT
This chapter introduces a model for the occurrence of basic events that share a common parameter (failure probability or rate). The information dependence breaking theorem discusses only models for the parameter consisting of a single interval. The chapter provides some simple example that illustrates common parameter or information dependence. If the amount of data about the unknown parameter is large, then the distribution will be very close to deterministic. Sampling from the Bayesian distributions produces deterministic belief functions for each type parameter. To move to the general case of k events with random interval information about the parameters, the Monte Carlo approximation technique used in WASH-1400 can be expanded. Furthermore, although the motivation for breaking common parameter dependence comes from trying to calculate the system failure beliefs for systems represented by fault trees, the exact representation of the system is not important at this time.
