Talk:Evidence under Bayes theorem

From Wikipedia, the free encyclopedia

This article is within the scope of WikiProject Statistics, which collaborates to improve Wikipedia's coverage of statistics. If you would like to participate, please visit the project page.

This material was broken out of the article on evidence, as it is tertiary to the legal aspects of evidence. BD2412 T 00:00, 29 December 2005 (UTC)

The assertion that probability theory is "tertiary" to the "legal aspects" of evidence in litigation is incorrect. Expert evidence based on the hard sciences is frequently admitted in trials. Today almost all hard sciences make use of probability theory; and, in some instances, the hard sciences make use of Bayesian analysis.

  • I wonder if Wikipedia is capable of treating this tricky topic -- Bayesianism in the law school clasroom and Bayesianism in the courtroom -- with the necessary degree of subtlety. (The fact that judges and law teachers, as a rule, don't handle probability theory very well is no excuse.)
    • The point is not that they don't handle it well, but that they don't handle it at all - ask most any lawyer about the Bayes theorem, and you'll draw a blank stare; ask them about the law of evidence, and you'll hear about authentication, witnesses, hearsay, and the like. Cramming Bayes theorem into the evidence article would be like putting in information about how much weight different materials can bear (a frequent topic raised in construction trials), or what the meaning of a particular contract clause should be - almost anything can come up as evidence in a trial, but an article on evidence under the law must have practical limitations the curtain it from going beyond where it will remain of use to the community of interest for that topic. BD2412 T 05:27, 1 January 2006 (UTC)
  • Unpersuaded observer: I take it everyone will agree that DNA evidence is of practical importance (in both criminal and civil litigation). DNA evidence when used to establish identity or non-identity is probabilistic -- and overtly so. It now turns that the same is true even of fingerprint evidence. If many lawyers are unable or unwilling to comprehend such important forms of evidence, this only shows that many lawyers are ill-trained and not that an encyclopdeia -- a storehouse of knowledge -- should mirror such ignorance.
  • In any event many trial lawyers have in fact learned that they must learn quite a bit about probability theory and statistical inference and, for the sake of profit and professional competence, have done so. See Barry Scheck and the many lawyers who have followed his example.
  • By the way, the importance of statistical and probabilistic evidence is not limited to criminal cases. Statistical and probabilistic evidence is used extensively, for example, in antitrust litigation, in trademark cases, in employment discrimination litigation, in environmental litigation (e.g., involving sampling for pollutants), wrongful death cases (e.g., to predict probable duration of life terminated prematurely in automobile accident), medical malpractice cases, toxic tort litigation (see Agent Orange case, tobacco litigation, asbestos cases), and in a wide variety of other cases in which scientific evidence is used.