Background
Stephen Senn is a statistician with a long institutional career spanning positions at the University of Glasgow, the University of Edinburgh, and several European research institutions. His academic work centres on statistical methodology in clinical trials, cross-over design, and the philosophy and practice of inference in medical research. He is the author of widely cited texts on statistical thinking in pharmaceutical and clinical contexts, and is known for direct public commentary on methodological failures in the application of statistics to medicine and law.
Senn is associated with a school of statistical criticism that is sceptical of over-reliance on p-values, frequentist pattern-detection in small populations, and the uncritical transfer of clinical-population statistics into forensic settings. These are precisely the methodological concerns that have been raised by multiple independent statisticians in relation to the Letby prosecution.
Statistical-methodology critique
Senn has written and spoken publicly about the general class of problems that arise when statistical coincidence reasoning is used to establish culpability in criminal proceedings. His published commentary addresses, among other issues, the prosecutor’s fallacy — the conflation of the probability of a coincidence given innocence with the probability of innocence given a coincidence — as well as the base-rate problems that arise when shift-coincidence data is analysed without adequate reference to a defined comparison population.
These methodological objections are directly relevant to the Letby case, where the Crown advanced evidence of a statistical association between Letby’s shifts and adverse events, and where independent statisticians have argued that the analysis did not account adequately for survivorship bias, confounding clinical factors, or the multiple-comparison problem. Senn’s published framework for evaluating such arguments provides one of the academic reference points against which the Letby statistical evidence is being assessed.
Commentary on the Letby case
Senn is reported to have commented, in accordance with his broader methodological views, on the category of problems that coincidence-of-presence statistical arguments present in criminal proceedings. While he has not, as far as is publicly documented, produced a formal expert report in the Letby proceedings, his published work constitutes part of the academic literature on which commentators and independent experts have drawn when critiquing the statistical component of the Crown’s case.
Independent statisticians who have engaged directly with the Letby evidence — including Prof. Richard Gill and Prof. Norman Fenton — work within a tradition of statistical criticism of which Senn’s published methodology forms a part. His contributions to the debate about statistical inference in forensic contexts are referenced in academic literature on this class of prosecution.
Read alongside
- Prof. Richard Gill — statistical methodology in the Letby case
- Prof. Norman Fenton — Bayesian evidence critique
- Prof. Jane Hutton — statistical evidence
- Statistical pattern evidence and the off-shift denominator
- Lucia de Berk — statistical-evidence acquittal
Source
Public statements, named-publication articles, court rulings, and our own coverage where applicable.