British Irish Region IBS

 View Only

Report - Current Topics in Designed Experiments

By Kirsty Hassall posted an hour ago

  

The British and Irish Region held a meeting at King’s College London on 23 April 2026. Marking the Centenary of RA Fisher’s landmark paper “The Arrangement of Field Experiments”, the meeting focussed on the design of experiments in modern biometrics. Fisher (1926) introduced, at least in outline form, many of the methods that are still being extended today, including randomisation and its relationship with data analysis, replication, blocking, factorial treatment structures and confounding. Three speakers described their recent work. First, Peter Goos (KU Leuven) described a flexible algorithm for sample size and design optimization in the presence of transition costs, motivated by forage harvester experiments and pilot plant experiments from the food industry, in both of which major parts of the cost of experimentation were connected with changing factor levels. Rebecca Walwyn (University of Leeds) then described the use of factorial designs for clinical trials of practitioner-delivered interventions, with the feature that therapist effects were considered as treatment effects, but also random effects in complex multi-stratum and sometimes multi-tiered designs. Finally, Simon Bate (GlaxoSmithKline) described using Hasse diagrams to understand complex experimental designs used in the pharmaceutical industry often involving both complex treatment structures, such as incomplete factorial designs and complex block structures involving both crossing and nesting. All talks could trace their roots to Fisher’s foundational paper and a lively discussion followed from the appreciative audience. The meeting was followed by the Fisher Memorial Lecture.

0 comments
0 views

Permalink