It doesn’t help with senior lawyers who lack their initial insight into probability theory

Lawyers should receive a basic introduction to how to assess and address uncertainty. This is not what it is.

Practically all evidence is inconclusive. The corollary should be that lawyers get a basic introduction to how to assess and address uncertainty. This is not what it is. While uncertainty theory and probability theory have made their way into most social science studies in recent decades, very little has happened in legal education.

In economics, these topics have been central for nearly a century and in the natural sciences even longer. Medical students learn how probability theory can be used to determine how the reliability of a diagnosis is affected by uncertain symptoms.

Likewise, law students must learn how the possibility of guilt is affected by uncertain evidence. (The elective course in Legal Evidence Theory for Law Students at the University of Oslo is a humble start.)

The Baneheia case can illustrate the importance of uncertainty. Witness statements, DNA and a mobile phone were essential.

Eyewitness accounts are, of course, unconfirmed. This is DNA evidence. There are often deficiencies in the DNA strands themselves, and the conditions in their processing, which make it uncertain whether the DNA originates from a particular person.

Technical evidence is also uncertain. Internationally, a number of studies have shown that experts make errors in their assessment of strands of hair, handwriting, speech, fingerprints, and the relationship between guns and bullets. to mention something.

Easy to get lost

The international literature on evidence assessment shows how easy it is to get lost in uncertainty assessments. An example is the case in which world famous football (American football) player OJ Simpson was accused of murdering his ex-wife and her friend. According to USA Today, The case has been described as “the most talked about criminal case in history”.

Professor Alan M. Dershowitz, nicknamed “The World’s Most Famous Criminal Lawyer,” has appeared as an expert witness for Simpson. He became a professor at Harvard University before he was thirty.

On the question of the possibility that Simpson may have committed the murders, an expert witness stated roughly the following: Well, it is true that Simpson has previously been convicted of violence against his ex-wife, but statistics show that in only one out of a thousand cases did he murder the convicted men committing acts of violence against their ex-wives. This possibility is too small to be given much weight.

Simpson was acquitted. However, he was later convicted.

The problem with Dershowitz’s statement is that it is misleading. The relevant possibility is not that given by the expert witness, but the following possibility: if the ex-wife of the husband had previously practiced violence against her, killingWhat is the probability that the husband is the killer? In the past, this probability was calculated at 80 percent.

Extensive literature on the trial indicates that the prosecution did not mention or consider this possibility appropriate. Perhaps the error was – and was clearly a mistake – caused by a lack of knowledge. Introductory courses in probability theory clearly explain the difference between the general probability that a particular type of person (here a spouse convicted of violent acts) commits an act (murder), and the probability that an actual murder will be committed by a person of that type.

It is impossible to determine whether Dershowitz was ignorant, or whether he intentionally and intentionally misled the jury. In any case, the gathering of senior lawyers failed to apply elementary probability theory.

The key point here is that the confusion in the Simpson case is not an isolated case. There are many examples of court cases from different countries with a similar flawed use of probability. It doesn’t help with senior lawyers who lack their initial insight into probability theory.

Probability theory does not solve all problems of proof, but at least it prevents a number of wrong thinking.

How is the situation in Norway?

Since it is difficult to determine how juries and judges will gather uncertain evidence from actual judgments, some simple tests were conducted among advanced law students. Tests show that students are like most people. They have gone astray. They fail to combine the two subtle possibilities.

The test relates to a fictional case that contains only two unconfirmed pieces of evidence. Only one witness (generally reliable) sees a car accident on a large island with no connection to the mainland, who told the judge that the car was red.

The judge wants to determine the probability that the car is actually red. He gets two pieces of information to help: Trusted witnesses generally get the car colors wrong in 5 percent of these cases. This means that witnesses give 5% false answers. Furthermore, the judge knows that only 0.1 percent of the cars on the island are red. The students were told that they should regard this information as correct, and estimate the probability that the accident vehicle was red.

The vast majority of students appreciated this possibility roughly. 95 percent, clearly consistent with the credibility of the certificate.

Correct is approx. 2 percent. (See, for example, my book Evaluating Evidence – Uncertainty and Possibility» for an explanation.)

The main point, made by psychologists, is that most people allow themselves to be influenced by witness statements (and other readily available evidence), while ignoring more technical underlying circumstances. Most of the students ignored the information that there were relatively few red cars on the island.

Would lawyers, jurors, and judges have responded like most people?

Can judges be misled?

It will be interesting to see what the public review of the Baneheia case will reveal about the assessments made about the uncertainty in Jan Helge Andersen’s witness statement, the number of assailants, DNA evidence, and the location of Christiansen’s phone calls. This applies not least to the Commission for the Re-opening of Criminal Cases.

Out of curiosity, it may be noted that in 2013 the undersigned suggested to the court administration that a test be conducted among the participants of the continuing education course for judges, a test similar to the test I did for law students. Since I received no response, after two years I requested a response to my letter. I’m still waiting for that.

Does anyone fear the prospect of getting indications that judges could allow themselves to be misled in the same way that law students and the general public are misled?