From 2013 to 2020, the Ministry of Education required schools to use a preponderance of standards of proof when assessing allegations of sexual assault. [17] A major difficulty with the two mathematical concepts of probative value discussed above is that, for most evidence, it is difficult to obtain the numbers needed to calculate the probability ratio (Allen 1991: 380). Exceptionally, there is quantitative data on baseline levels, as in our blood type example. When objective data are not available, the interviewer must rely on experience and basic knowledge to determine subjective values. In our blood type example, a critical factor in calculating the probability ratio was the percentage of the „suspect population“ that had the same blood type as the accused. „Reference class“ is the general statistical term for the role played by the suspect population in this analysis. How to define the reference class of the „suspect population“? Should we consider the population of the country as a whole or the city or street where the alleged murder took place? What if this happened at an international airport where most people are foreign visitors? Or what if it is proven that the accused and the victim were detained in the same prison at the time of the alleged murder? Should we then take the prison population as a reference class? The distribution of blood groups may vary depending on the reference class chosen. Skeptics of mathematical modelling of probative value point out that data from different reference classes have different explanatory power and that the choice of reference class is open to contextual arguments and should be subject to contextual arguments and requires judgment; There is no a priori way to determine the correct reference class. (On the problem of reference classes in the establishment of legal facts, see, in addition to the references cited in the remainder of this section, Colyvan, Regan and Ferson, 2001; Trollers, 2005; Allen and Roberts, 2007.) Evidence may only be presented in legal proceedings to prove a fact if the fact is relevant. Relevance is a relational concept. No facts are relevant in themselves; It is only relevant in relation to another fact. The term „likely“ is often used to describe this relationship.

We see two examples of this in the following known definitions. Stephen (1886:2, emphasis added): In England and Wales, however, the Courts of First Instance Act 1980, at p. 101, provides that if a defendant invokes an „exception, exemption, reservation, excuse or reserve“ in his defence in summary proceedings, the legal burden of proof of that exception rests with the defendant, if only after weighing the odds. For example, a person charged with being in charge of an intoxicated motor vehicle may raise the defence that there was no likelihood that he or she would drive drunk. [35] The prosecution has the legal burden of proving beyond reasonable doubt that the defendant exceeded the legal alcohol limit and was in control of a motor vehicle. Possession of the keys is usually sufficient to prove control, even if the defendant is not in the vehicle and may be in a nearby bar. According to this evidence, the defendant bears the legal burden of proof after weighing the likelihood that he would probably not drive. [36] As the relevant evidence before us increases, the likelihood of the argument may decrease or increase, depending on how the new evidence amplifies the adverse or favourable evidence; But in both cases, something seems to have improved – we have a more substantial basis on which to base our conclusion. I express this by saying that adding new evidence increases the weight of an argument. New evidence sometimes reduces the likelihood of an argument, but it will always increase its „weight.“ Third, although the Bayesian theorem provides a method for updating probabilities in light of new evidence, it is silent on what the initial probability should be.

In a trial, the initial probability cannot be zeroed, as this means the certainty of the innocence of the accused. No new evidence can make the slightest difference; Regardless of the probability ratio of the evidence, multiplication by zero (the previous probability) always gives an after-the-fact probability of zero. On the other hand, it is also problematic to start with an initial probability. This applies in particular to a criminal case. Starting a trial with a certain probability of guilt means that the investigator initially believes that the accused is guilty, which is not easy to reconcile with the presumption of innocence. (Tribe 1971: 1368-1372; cf. Posner 1999:1514, suggests starting the process with a previous odds of 50:50, criticized by Friedman in 2000. The problem of determining prior probability is supposed to disappear if we base fact-finding simply on probability ratios: Sullivan, 2019: 45-59.) Clear and convincing evidence means that the evidence presented by a party during the trial must be highly and substantially more likely than not likely, and that the trial judge must have a firm conviction or a belief of a factual nature. [25] This standard must provide a higher degree of credibility than the usual standard of proof in civil matters (i.e.