The Process of Expert Elimination
Last week I deposed yet another expert witness who based his opinion on the so-called methodology of "process of elimination"; sometimes referred to as "differential diagnosis". Apparently, the new fad in Daubert-avoidance is to claim that one's opinion was reached via this unassailable method since he's the fourth expert in a row to so claim. Too many judges, all having run the gauntlet of SATs and LSATs, tend to find the method sound and unexceptional. So for them, here's a test.
What is the most likely cause of plaintiff's cancer? (A) "the evil eye"; (B) "bad humours of the blood"; (C) a Voodoo curse; (D) demonic possession; or, (E) Tetra-methyl-whatchamacallit?
Next, say why.
Expert witnesses who play the "process of elimination" game want you to assume, as good high school students do, that the set of all possibly correct answers is encompassed by choices (A) - (E). Of course, not all possible answers are known in real life and seldom do experts present viable alternatives to (E). Nor do they ever, out of fear of all things Bayesian, attempt to lay odds on the theory nor even say by how much subsequent evidence has affected their confidence in the proffered theory.
Our job often then is to explain to the courts that the other side's expert opinion (E) is not the result of considering evidence for and against (E) but is rather a form of the argument to ignorance. Real claims to knowledge still require evidence and the statement that a hypothesis is probably true because the straw men put up to oppose it are unconvincing is an absurdity. Hypotheses will always stand or fall on the evidentiary foundations upon which they themselves rest rather than upon the foundations of competing ideas.
