Faculty Scholarship Repository
A Service of the Ross-Blakley Law Library
| Article |
| Apples and Oranges: Confidence Coefficients |
| David Kaye |
| 73 Cornell L. Rev. 54 (1987) |
| Library Access |
|
Abstract: In his recent article entitled Confidence in Probability Professor Neil Cohen seeks to overthrow ‘the currently accepted probabilistic formulation of the burdens of persuasion.’ Cohen perceives a dissonance between the generally accepted Bayesian decision theory interpretation of the civil preponderance of the evidence standard and a few hypothetical cases in which it is imagined that courts will direct verdicts against a plaintiff whose only evidence is overtly statistical. The resolution of the perceived dissonance resides in the often overlooked distinction between justified and unjustified naked statistical evidence, and in the application of a negative, spoliation-like inference or doctrine in the latter situation. This article spells out the logic and implications of Cohen’s analysis, and applies a similar analysis to a realistic problem-assessing the evidentiary value of polygraph tests.This misunderstanding has important practical implications. In cases involving statistical proof, it can engender a false sense of confidence in the implications of statistical evidence. Also, recent reports of the ‘confidence’ that judges or jurors can have in the results of polygraph tests in criminal cases could be misinterpreted. This article hopes to promote a deeper understanding within the legal profession of the confidence interval's meaning and relation to the burden of persuasion. |
| 2,614 |
| Total Views |