Evidence under Bayes theorem

Evidence under Bayes theorem

Among evidence scholars, the study of evidence in recent decades has become broadly interdisciplinary, incorporating insights from psychology, feminism, economics, and probability theory.

One area of particular interest and controversy has been Bayes' theorem. Bayes’ theorem is an elementary proposition of probability theory. It provides a way of updating, in light of new information, one’s prior probability that a proposition is true. Evidence scholars been interested in its application to their field, either to study the value of rules of evidence or to help determine facts at trial.

Suppose, for example, that the proposition to be proven is that defendant was the source of a hair found at the crime scene. Before learning that the hair was a genetic match for the defendant’s hair, the factfinder believes that the odds are 2 to 1 that the defendant was the source of the hair. If she used Bayes’ theorem, she could multiply those prior odds by a “likelihood ratio” in order to update her odds after learning that the hair matched the defendant’s hair. The likelihood ratio is a statistic derived by comparing the odds that the evidence (expert testimony of a match) would be found if the defendant was the source with the odds that it would be found if defendant was not the source. If it is ten times more likely that the testimony of a match would occur if the defendant was the source than if not, then the factfinder should multiply her prior odds by ten, giving posterior odds of 20 to one.

Bayesian skeptics have objected to this use of Bayes’ theorem in litigation on a variety of grounds. These run from jury confusion and computational complexity to the assertion that standard probability theory is not a normatively satisfactory basis for adjudication of rights.

Bayesian enthusiasts have replied on two fronts. First, they have said that whatever its value in litigation, Bayes’ theorem is valuable in studying evidence rules. For example, it can be used to model relevance. It teaches that the relevance of evidence that a proposition is true depends on how much the evidence changes the prior odds, and that how much it changes the prior odds depends on how likely the evidence would be found (or not) if the proposition were true. These basic insights are also useful in studying individual evidence rules, such as the rule allowing witnesses to be impeached with prior convictions.

Second, they have said that it is practical to use Bayes' theorem in a limited set of circumstances in litigation (such as integrating genetic match evidence with other evidence), and that assertions that probability theory is inappropriate for judicial determinations are nonsensical or inconsistent.

Some observers believe that in recent years (i) the debate about probabilities has become stagnant, (ii) the protagonists in the probabilities debate have been talking past each other, (iii) not much is happening at the high-theory level, and (iv) the most interesting work is in the empirical study of the efficacy of instructions on Bayes’ theorem in improving jury accuracy. However, it is possible that this skepticism about the probabilities debate in law rests on observations of the arguments made by familiar protagonists in the legal academy. In fields outside of law, work on formal theories relating to uncertainty continues unabated, and appears to be accelerating. One important development is the acceleration of work on "soft computing," work that is carried on, for example, at Berkeley under Lotfi Zadeh's BISC, Berkeley Initiative in Soft Computing. Another example is the increasing amount of work, by people both in and outside law, on "argumentation" theory. Even work on Bayes nets continues at an accelerating pace. Some of this work is beginning to filter into legal circles. See, for example, the many papers on formal approaches to uncertainty (including Bayesian approaches) in the Oxford journal: Law, Probability and Risk [http://lpr.oxfordjournals.org/] . In retrospect it may appear that the discussion about formal argument about factual uncertainty in law was just beginning in the last quarter of the twentieth century.


Wikimedia Foundation. 2010.

Игры ⚽ Поможем сделать НИР

Look at other dictionaries:

  • Evidence (law) — The law of evidence governs the use of testimony (e.g., oral or written statements, such as an affidavit) and exhibits (e.g., physical objects) or other documentary material which is admissible (i.e., allowed to be considered by the trier of fact …   Wikipedia

  • Evidence-based medicine — (EBM) aims to apply evidence gained from the scientific method to certain parts of medical practice. It seeks to assess the quality of evidencecite journal |author=Elstein AS |title=On the origins and development of evidence based medicine and… …   Wikipedia

  • Bayes's theorem — ▪ probability       in probability theory, a means for revising predictions in light of relevant evidence, also known as conditional probability or inverse probability. The theorem was discovered among the papers of the English Presbyterian… …   Universalium

  • Theorem — The Pythagorean theorem has at least 370 known proofs[1] In mathematics, a theorem is a statement that has been proven on the basis of previously established statements, such as other theorems, and previously accepted statements …   Wikipedia

  • Naive Bayes classifier — A naive Bayes classifier is a simple probabilistic classifier based on applying Bayes theorem with strong (naive) independence assumptions. A more descriptive term for the underlying probability model would be independent feature model . In… …   Wikipedia

  • List of statistics topics — Please add any Wikipedia articles related to statistics that are not already on this list.The Related changes link in the margin of this page (below search) leads to a list of the most recent changes to the articles listed below. To see the most… …   Wikipedia

  • List of mathematics articles (E) — NOTOC E E₇ E (mathematical constant) E function E₈ lattice E₈ manifold E∞ operad E7½ E8 investigation tool Earley parser Early stopping Earnshaw s theorem Earth mover s distance East Journal on Approximations Eastern Arabic numerals Easton s… …   Wikipedia

  • Bayesian inference — is statistical inference in which evidence or observations are used to update or to newly infer the probability that a hypothesis may be true. The name Bayesian comes from the frequent use of Bayes theorem in the inference process. Bayes theorem… …   Wikipedia

  • Prior probability — Bayesian statistics Theory Bayesian probability Probability interpretations Bayes theorem Bayes rule · Bayes factor Bayesian inference Bayesian network Prior · Posterior · Likelihood …   Wikipedia

  • Bayesian network — A Bayesian network, Bayes network, belief network or directed acyclic graphical model is a probabilistic graphical model that represents a set of random variables and their conditional dependencies via a directed acyclic graph (DAG). For example …   Wikipedia

Share the article and excerpts

Direct link
Do a right-click on the link above
and select “Copy Link”