This thesis examines two possible orders of combining multiple experts in elicitations with multiple de-composed events: Should experts be combined early or later in the decision process? This thesis is in conjunction with the paper (Baker & Olaleye, 2012) where we show that it is best to combine experts early as later combination leads to a systematic error. We conduct a simulation to more fully flesh out the theoretical model. We also conduct a theoretical analysis aimed at determining how significantly these two methods differ. We find that all results are in accordance with the theory but combining experts later might lead to less error in some cases due to randomness. We then conduct an empirical evaluation of the two methods using data from a previous study. We show that the experts exhibit some form of correlation. The impact of using the two methods of combining experts is then evaluated using an optimal R&D investment portfolio model. We find that the elicitation inputs have a significant effect on the outcome of the optimal portfolio and that there is an advantage from combining experts early.