The British and Irish Region would like to invite BIR members (and friends and colleagues!) to an afternoon celebrating Statistical Consultancy, including the 2024 Presidential Address and preceded by the 2024 Region AGM, on 3rd December 2024, at De Morgan House in London.
Registration is now open on Eventbrite at https://www.eventbrite.co.uk/e/statistical-consultancy-the-art-of-the-science-and-2024-agm-tickets-1057045604859 ; the Region's AGM (12:45 to 13:15) is free to attend, but a registration fee will be payable for lunch (13:15 to 14:00) and the scientific meeting (14:00 to 16:30).
Please note we are also providing an option for you to pay your registration fee for the BIR for 2025 at the same time as booking your place at the meeting - the cost for both together is £100.
Finally, for IBS Members only, we have added an option for virtual attendance; tickets are available on the Eventbrite site, priced £25 for the full meeting (no charge for attending the AGM only).
Statistical Consultancy – The Art of the Science
A meeting organised by the British and Irish Region of the International Biometric Society
3rd December, 14:00-16:30
Hardy Room, De Morgan House, London Mathematical Society, 57-58 Russell Sq, London WC1B 4HS
(with the IBS-BIR AGM preceding the meeting at 12:45, followed by lunch at 13:15)
TIMETABLE
12:45 IBS-BIR AGM (IBS-BIR members only – no registration fee required)
13:15 LUNCH & Arrival for Main Meeting
14:00 Introduction
14:05 Andrew Mead
Statistical Design for Modern Experiments – A Consultant's Casebook
14:35 Jennifer Visser-Rogers
Experiences of Clinical Trial Consultancy and the Future Landscape
15:05 Break – refreshments
15:30 Mark Brewer – Presidential Address
Statistical Consultancy and the Use of Plausible Intelligence
16:30 Close
Abstracts:
Andrew Mead
Statistical Design for Modern Experiments – A Consultant's Casebook
The statistical design of experiments has a strong connection with Rothamsted Research, through the pioneering work of R.A. Fisher in the 1920s and 1930s, with considerable contributions also from Frank Yates, Donald Preece, Roger Payne, Rosemary Bailey and many other short- and long-term residents. My own contributions to this Rothamsted history are strongly influenced by the "(Roger) Mead school", as captured in Mead, Gilmour and Mead (2012), and now influence how experiments at a range of scales are approached within the institute. At a recent conference I overheard the suggestion that statistical design was no longer important "because we have linear mixed model analysis methods" - whilst I agree that the advent of these methods has provided more freedom in how we might approach designing many experiments, I strongly defend the importance of good statistical design, at least in applied agri-environmental research. I will talk about the general challenges of designing experiments as part of our consultancy role, illustrated by three relatively recent and complex examples.
Mark Brewer
Statistical Consultancy and the Use of Plausible Intelligence
Does statistical consultancy get the credit it deserves? Is it seen as the poor relation of statistical methodological development? Is it just "easy", applying standard statistical techniques, simple and repeatedly? Can we now replace statistical consultants with Artificial Intelligence?
With examples gathered over many years, Mark will argue that statistical consultancy is commonly a vital part of the scientific process, and itself requires training, skills and experience above and beyond mathematical capability, programming skills and technical competence.