The validity of subjective probability of success forecasts by R & D project managers

Models and techniques to aid management in planning and controlling R&D projects frequently use subjective probability of success forecasts as one of the major inputs. An experiment was conducted at the research laboratories of Monsanto Company to measure the predictive validity and consistency of such forecasts. The results indicate that the eventual success or failure of certain types of R&D projects can be predicted by measuring the time shape of polled probability of success forecasts. Probability of success forecasts appear to yield more valid advance warning indicators than several commonly used project status measures. These results tend to support the hypothesis that R&D planning and control models that are based on subjective probability estimates may realiably be used by management to aid in early identification of eventually failing projects, as well as to aid in project selection and project funding.