Calibration and the aggregation of probabilities

In order to avoid the task of assessing a complicated likelihood function, Morris uses an axiomatic approach to develop a multiplicative rule for aggregating a decision maker's and an expert's probabilities. An essential shortcoming of the multiplicative rule is that it does not allow the decision maker to model his beliefs about the dependence between his assessment and the expert's. The root of the problem lies in the fact that the decision maker must calibrate the expert's information. When the calibration is done properly, the decision maker is forced to tackle the task which Morris proposes to avoid.