Professor Stephen Lindsay, University of Victoria
The Credibility Revolution in the Life Sciences
This is an exciting time for life sciences. Developments in statistics and programming have gone hand in glove with technological advances and cultural changes to enable new approaches to collecting, analysing, and sharing data. This nascent renaissance was sparked, in part, by a recent surge in awareness of flaws in the way some life-science researchers use inferential statistics to answer questions (e.g., Does mindfulness meditation enhance wellbeing? Do vitamin D supplements improve health?). Until a few years ago, I was one such researcher. I’m still no statistics maven, but I’ve learned a lot that I am convinced has improved the quality of my science. The primary aims of my talk are to (a) persuade listeners that some areas of the life sciences have a long-standing problem with overstating findings, (b) help them to understand the causes of that problem, and (c) briefly describe methods and introduce tools that improve effect-size estimation.