Helping medical scientists analyse causal relationships

“Most medical research questions concern causal relationships. Say that we want to study if people who used facemasks during the Covid pandemic developed severe symptoms less often than those who didn’t. If we then observe a difference between the groups, we would typically want to know if it’s attributable to a protective effect of the mask or other differences between the groups, such as age or gender.”

How are you going about this?

“Most of my research is theoretical. I work with mathematical tools and statistical models rather than with real data. But I also support other researchers and assist them in their work.”

What are your main findings to date?

“In the past few years, I’ve been focusing on developing what are called sensitivity analyses, the aim of which is to give probable intervals for the strength of causal relationships. Since there can always be underlying factors, you can never say anything categorical about causal effects. But you might be able to make reasonable speculations. My aim is to help scientists make more reliable evaluations of the strength of such relationships.”