- This event has passed.
Seminar: Peter McCullagh
On-site-only seminar: Friday, January 30th 2026, from 11.00 – 12.00 AM
Title: Statistics of speciation and the evolution of reproductive isolation
Speaker: Peter McCullagh (University of Chicago)
Location: Room 205 – Viale Morgagni 59
ABSTRACT
This talk is concerned with statistical models for the analysis of an experiment in evolutionary biology. The experiment described by Villa et al. (PNAS 2019) is concerned with adaptive evolution of an organism in response to selective pressure. The catchy title “Rapid experimental evolution of reproductive isolation from a single natural population” implies that evolution is observed on short time scales on the order of 50 generations. I will describe the experimental design, the host-parasite system, the randomization scheme and the sampling protocol. One flaw of the statistical model leading to the advertised conclusion is that it contradicts the randomization scheme. An alternative model (due to Cavalli-Sforza & Edwards (1967)) using Brownian motion for the temporal evolution of a quantitative trait is shown to conform to the randomization scheme. It is also a substantially better fit to the data. Unfortunately, the alternative analysis shows that, while adaptive evolution is indeed possible, the authors’ data provides zero evidence in support of that conclusion.
BIO
Peter McCullagh is a Northern Irish–born statistician and John D. MacArthur Distinguished Service Professor Emeritus in the Department of Statistics at the University of Chicago. He completed his Ph.D. at Imperial College London under the supervision of Anthony Atkinson and David Cox and has made foundational contributions to statistical theory and methods, including work on generalized linear models and tensor methods in statistics.
He is co-author, with John Nelder, of the influential text Generalized Linear Models, widely used across disciplines, and his research spans areas such as quasi-likelihood, analysis of ordinal data, and the mathematical foundations of statistical models.
Professor McCullagh has received numerous honours, including the COPSS Presidents’ Award (1990), the Royal Statistical Society’s Guy Medal in Bronze and Silver (1983 and 2005), and election as a Fellow of the Royal Society and the American Academy of Arts and Sciences.