t-threads
private beta
search about

Infectious disease epidemiologist and microbiologist, aspirational barista. mlipsitc@hsph.harvard.edu Director @CCDD_HSPH

May 13, 2020 4:13 AM
copied to clipboard

This gets to the heart of our agreement (and the next one about virtues becoming vices). If EBM contributes rigor to thinking, great. I’d push back a little and say I see greater rigor in Bayesian updating than in p value worship practiced by some EBMers

Jonathan Fuller (@JonathanJFuller):
My main point was to describe different virtues, attitudes and strategies in epidemiology. 6/
https://twitter.com/JonathanJFuller/status/1260401879055511554


And more rigor in causally sophisticated analysis of observational (and RCT) data than in a naive view that randomization cures all ills (@LauraBBalzer pointed this out in initial discussion of @JonathanJFuller s piece).


Those who take observ data seriously are often more rigorous on questions of external validity. So I’d say we should all vote for rigor and see EBM as a set of heuristics that sometimes are rigorous and sometimes limit clear thought. All this related but diff from articles