Professor Stefanie Biedermann

Professional biography

My research focuses on enabling researchers in Medicine and Science to draw accurate conclusions from their experiments. The amount and quality of information gathered from experimental data depends sensitively on the experimental conditions at which the data have been collected, the design of the experiment.

I have also developed a research interest in informative censoring in survival analysis, where I have held a grant by the MRC to develop new modelling approaches for sensitivity analyses.

Here is a quick overview of my research interests.

• Optimal Design of Experiments (methodology)
• Medical applications, for example planning clinical trials for incomplete (missing or censored) data
• Algorithms to find optimal designs numerically
• Survival Analysis (methodology)
• Medical applications, for example sensitivity analyses for potentially informatively censored data
• Mixture experiments (modelling and design), as they occur, for example, in Formulation Chemistry

Here are several examples that showcase my current research plans in more detail:

• Missing data can cause considerable problems in experiments. A particularly problematic situation arises when the missingness is related to the (unobserved) value of the missing response, which may lead to biased conclusions. I am interested in investigating scenarios where some of the missing responses can be “recovered” e.g. through home visits to patients who missed their clinical appointments. This opens up a whole host of opportunities for statistical research, including modelling, inference and experimental design. I have recently been awarded EPSRC grant EP/V00641X/1 in collaboration with Dr Robin Mitra (Cardiff University) to investigate this topic.
• Experiments involving mixtures are conducted in a variety of areas, for example in food processing or in chemical research. The experimental region is constrained naturally, as the proportions of all ingredients sum to one. Khashab, Gilmour and Biedermann (2021) propose a new - parsimonious but flexible - class of non-linear models, based on fractional polynomials (Royston and Altman, 1994), to fit the data from constrained mixture experiments. I am interested in collaborations with researchers in Chemistry to develop these models further, and to test them in realistic scenarios.
• For many years, the field of optimal design of experiments was mostly concerned with theoretical advances. However, in practice, most optimal design problems are too complicated to be solved analytically, and efficient algorithms for numerical design search are required. While the benefits of optimal designs have been well established in various application areas, practitioners cannot use optimal designs unless they are readily available to them. Therefore, it is essential to develop efficient algorithms and to incorporate them into an easy-to-use software package, which can find optimal designs quickly and to a good accuracy. I am interested in collaborations with researchers in Operational Research and Computer Science to do this, and in finding potential users of this research in Science and Industry.

I am currently on the production team for M348 Applied Statistics; I am module team chair for M249 Practical Modern Statistics and deputy chair for M248 Analysing Data.

I have enjoyed/am enjoying fruitful collaborations with many colleagues in the UK and abroad, for example in the US (Prof Min Yang, UIC), Singapore (Dr Selin Damla Ahipasaoglu, SUTD/Southampton), Germany (Prof Holger Dette, Dr Nico Bissantz, RUB) and the UK (Dr Robin Mitra, Lancaster/Cardiff, Prof Steve Gilmour, KCL, Prof Dave Woods, Dr Alan Kimber, Southampton).