Design for disparate disciplines
Maximising success, minimising failure

A poorly designed experiment can lead to useless results and wasted resources. However, by carefully designing an experiment, it should be possible to minimise information losses and increase the amount of information gained from data. Minimax aims to do just that.

When gathering information from experiments, it is crucial to think about the means of doing so. Known as experimental design, the choices made are shaped by the decisions on what variables to examine, the level at which they should be examined and how they should be examined in relation to other variables. Each decision is made with the aim of finding the optimal design for the experiment. This will maximise the information gathered within the limitations of the resources available. Experimental design is central to the success and validity of any experiment, and is therefore crucial. Nonetheless, there are numerous examples of experiments where better planning could have doubled the precision in the parameter estimation (measured in terms of halved error of estimation). There are also examples of where a poorly planned experiment results in a situation where it is not possible to draw any conclusions at all.

Read the full article by downloading this pdf. Design for disparate disciplines (1972 Kb)

About Ellinor Fackle-Fornius

Ellinor Fackle-Fornius is Assistant Professor at the Department of Statistics, Stockholm University, Sweden. She received a PhD in Statistics in 2008 for the thesis entitled “Optimal Design of Experiments for the Quadratic Logistic Model”. Her major research interests include optimal design of experiments, minimax designs, sequential designs and generalised linear models.

About the project

The project aims to develop a new algorithm to handle the computational difficulties involved in the construction of minimax designs. The main objectives are to develop the algorithm and evaluate its convergence properties, examine the efficiency of minimax designs in comparison with other design methods, and apply the algorithm to practical problems.


Linda Wänström, Division of Statistics, Department of Computer and Information Science, Linköping University, Sweden
Hans Nyquist, Department of Statistics, Stockholm University, Sweden

About International Innovation

International Innovation is the leading global dissemination resource for the wider scientific, technology and research communities, dedicated to disseminating the latest science, research and technological innovations on a global level. More information and a complimentary subscription offer to the publication can be found at: