David Spiegelhalter

Sir David John Spiegelhalter (born 1953) occupies a unique position in modern statistics as both a leading Bayesian researcher and one of the most effective public communicators of statistical ideas. A Winton Professor of the Public Understanding of Risk at the University of Cambridge, he co-developed the BUGS (Bayesian inference Using Gibbs Sampling) software that democratized Bayesian computation, pioneered Bayesian methods in clinical trials and medical decision-making, and developed innovative ways to communicate risk and uncertainty to the public. His work has had an extraordinary impact on both the practice and the public perception of statistics.

Education and Early Career

Spiegelhalter studied mathematics at the University of Oxford and received his PhD from University College London. He spent many years at the Medical Research Council (MRC) Biostatistics Unit in Cambridge, where he developed Bayesian methods for medical applications, before being appointed to the Winton Professorship at Cambridge in 2007.

BUGS Software

Spiegelhalter was a principal developer of the BUGS (Bayesian inference Using Gibbs Sampling) software, first released in the early 1990s, along with its successor WinBUGS and the related OpenBUGS. BUGS allowed researchers to specify complex Bayesian models in a simple language and automatically performed MCMC sampling to obtain posterior distributions. For the first time, researchers without deep expertise in computational statistics could fit hierarchical models, random effects models, and other complex Bayesian specifications. BUGS was transformative, making Bayesian analysis accessible across medicine, ecology, social science, and engineering.

Bayesian Clinical Trials

Spiegelhalter made major contributions to the design and analysis of clinical trials using Bayesian methods. He developed frameworks for incorporating prior evidence (from previous studies or expert opinion) into clinical trial design, for adaptive trial designs that allow modification as data accumulate, and for health technology assessment using Bayesian decision theory. His work influenced both the pharmaceutical industry and regulatory agencies.

“Statistics is the science of changing your mind under uncertainty. Bayesian statistics is the formal version.”— David Spiegelhalter

The DIC and Model Comparison

Spiegelhalter, along with Best, Carlin, and van der Linde, introduced the Deviance Information Criterion (DIC) in 2002, a Bayesian analogue of the Akaike Information Criterion for model comparison. The DIC measures model fit penalized by effective complexity and can be computed easily from MCMC output. Despite some limitations, it became one of the most widely used tools for Bayesian model comparison.

Public Understanding of Risk

In his role as Winton Professor, Spiegelhalter became a tireless communicator of statistical concepts to the public. He has written popular books including The Art of Statistics (2019), appeared extensively in media, and developed innovative visualizations for communicating risk. He was particularly active during public health debates, helping people understand probabilities associated with medical treatments, environmental hazards, and everyday risks.