Go
to Peter Tillers' Home Page
Go
to Home Page for Fact Investigation
Course in Fact Investigation
Cardozo Law School
Professor Peter Tillers
Note on Assessing the Value of Unknown Evidence
As far as I can determine, some of the recipes for assessing the value of unknown information are not intrinsically or conceptually complicated under certain certain circumstances, under certain assumptions. For example, one recipe is the expected value formula. This recipe involves
(i) assigning weights to the possible outcomes [which in this case would be the possible outcomes of alternative sequences of investigative activity, I assume]; i.e., the decision maker -- here, the investigator -- must determine how much utility or disutility the possible alternative outcomes have (in someone's eyes) if those outcomes happen or exist;
and
(ii) taking into account -- discounting for, so to speak -- the degree of uncertainty about the realization of the pertinent alternative outcomes.
See, e.g., the brief explanation provided by the Decision Analysis Society
The trouble I have in even thinking about applying such a recipe to decision making about fact investigation is essentially twofold:
1. The number of alternative investigative paths and, thus, the number of alternative investigative choices, it seems to me, is enormous. See P. Tillers, "Is Proof in Litigation Predictable?: Some Obstacles to Systematic Assessment of Decisions about Proof in Litigation
2. The uncertainty in fact investigation is in part uncertainty about the evidence that an investigative path will uncover. Although there is a theory -- value of information theory, see, e.g., R. Shachter, "Efficient Value of Information Computation" -- that purports to provide a coherent ("rational") recipe (or procedure) for making decisions about investing (or not investing) in the acquisition of unknown and uncertain information, in some contexts -- particularly in the context of exploratory investigation -- the idea of using such a recipe makes me uncomfortable because evidence and information are often suggestive of possibilities and hypotheses -- i.e., evidence & information frequently have "abductive force." I have difficulty imagining a method for rigorously assessing the value of evidence on hypotheses that do not and cannot yet exist -- and whose nature, to a substantial extent, cannot be anticipated before the unknown evidence is acquired or better known. See P. Tillers, "The Explosive Dynamic Complexity of Evidentiary Processes." See also P. Tillers, "Spotty Semiotics" (But I warmly welcome any student who knows more about value of information theory than I do and who wishes to show or suggest how that theory or method might be used to help guide decisions in investigation about investigation.)
FINIS
Go
to Peter Tillers' Home Page
Go
to Home Page for Fact Investigation