In this event, Nigel Marriott (CStat MAE) will give an interactive talk on the Prosecutor's Fallacy. He will describe a case he has worked on as a statistical expert and the audience will have the chance to vote on the likelihood of guilt as each piece of evidence is described. He will also discuss the Sally Clark case and why he thinks statisticians have drawn the wrong lesson from this case.
The prosecutor’s fallacy confuses the probability of finding the evidence on an innocent person with the probability that a person on whom the evidence is found is innocent” -
The Inns of Court College of Advocacy joint publication with the Royal Statistical Society on Statistical Evidence page 22, 2019
When statistical evidence in a case indicates a very low p-value, it doesn’t automatically mean the case could be a prosecutor fallacy. What matters is how the p-value is interpreted alongside the other evidence in the case.
In this talk, you will get the chance to be the jury! Based on a case Nigel testified in as a statistical expert, you will first hear the statistical evidence he presented followed by 6 more pieces of evidence from the prosecution and defence. After each piece of evidence, you will have the chance to vote on the likelihood of the defendant being guilty of the charge. There will then be a chance to discuss why you voted the way you did before Nigel reveals the actual verdict and its implications.
Nigel will then discuss how his experience in this case led him to reconsider the infamous Sally Clark case. He will explain why he thinks statisticians have drawn the wrong lesson from this case and others.
For booking in-person, see the following link: https://rss.org.uk/training-events/events/events-2025/local-groups/you-the-jury-is-this-case-a-prosecutor-fallacy-or/
Nigel Marriott CStat MAE
Nigel Marriott is an independent Chartered Statistician based in Bath. He is a fellow of the Royal Statistical Society and a member of The Academy of Experts. Over a career spanning 35 years, he has worked with non-statisticians in nearly all industries and functions. He regards his role as helping clients to do the basics right when turning their data into insights and decisions. In support of this goal, he
blogs extensively on a number of statistical topics in the public domain.
His first instruction as an expert witness was in 2007 and set the tone for a wide variety of instructions since. The majority of his instructions involve designing or critiquing samples and surveys but detecting and verifying disparities in discrimination claims has accounted for a larger share of his work in recent years.
He has also written guidance for non-statisticians on how to draw conclusions from data. One example is “
Comparing soil contamination data with a critical concentration” (2020) which inspired the RSS webinar “
Dr Groundlove, or how I learned to stop worrying and love the Central Limit Theorem” in 2021. Another example in 2022 was the draft version of the government
guidance for employers who wish to analyse their ethnicity pay gap.
Nigel is currently Chair of the Business & Industry section of the RSS. He has previously been Chair of the Quality Improvement Section and a member of the Professional Affairs Committee.