Mendelian randomisation - it’s a technique that uses the chance distribution of genes in a population, combined with big data sets, to investigate causative relationships.

But there are a lot of questions we have in The BMJ about how the technique works - the association between genes and apparently non-biologically mediated behaviours, how much the strict rule of not claiming causation based on observational data has actually been overturned, and general confusion about how the non-methodologists amongst us can read these studies.

Neil Davies and George Davey Smith from University of Bristol, and Michael Holmes from the University of Oxford, join us to explain how the technique works, where it can be applied, and what readers should look out for when they're trying to assess the quality of a mendelian randomisation study.

Read their full research methods and reporting paper: