Abstract:

When he comes to assess what he should reasonably invest in a piece of research, the research buyer may try to apply a relatively simple, "Bayesian" method. It consists first in writing down four figures which are the basic data of his problem (the "Gain", "Loss" and "Bet" at stake on one side, and the contemplated research "Validity" on the other). Then, from these four figures, a relatively simple computation gives him an estimate of the contemplated research "Value" or worth. The following contribution, after reviewing these five notions, gives examples of the way in which research value varies according to each of the other four elements. Finally, after discussing the limitations of the applicability of this method, especially in social research, this contribution concludes that the method, not presented as an exclusive panacea, may be quite useful to the research buyer.