Subtweet to media, especially science writers. When WHO says "there is no evidence", they don't mean something is false. There is no evidence either way. It could just as well be true. Also, "correlation doesn't imply causation" is wrong and stupid, and doesn't debunk anything.
-
-
WHO has flunked the communication of the science of this in many ways, but media folks, too. Basics: absence of evidence is not evidence of absence. "No evidence" doesn't mean false. Correlation hints (sometimes strongly) at causation, but doesn't prove it by itself. Not hard!
Show this thread -
Yeah, another major failure. WHO went to China, traveled with all the minders, and didn't see evidence of asymptomatic transmission and told us there was none instead of telling us we didn't know, and communicate the limits of what they could ascertain.https://twitter.com/BolognaFishMD/status/1257197972539838467 …
Show this thread -
Coverage has already gotten into the horse-race framing! Instead of political polls or sports betting, we now have model tracking etc. It's not productive or healthy, nor will it help guide us out of here. It's misleading. We aren't getting what we need.https://twitter.com/eadhed/status/1257337806009044993 …
Show this thread -
Yes! There are many types of evidence besides randomized-trials. Sometimes it's not even possible to have randomized trials but that doesn't mean we should act like all the other types of evidence don't exist or are meaningless.https://twitter.com/jeremyphoward/status/1257346654409285635 …
Show this thread
End of conversation
New conversation -
-
-
This Tweet is unavailable.
-
I have seen so much "correlation does not imply causation" or some version as if it debunked something lately I am ready to breath fire. Masks! Correlation does not imply causation! Weather! Correlation! BCG! Correlation! No real discussion of uncertainty, what we need, etc.
- Show replies
-
-
-
At epidemiology school we have to learn the Bradford Hill criteria for inferring causality by heart! Strength and consistency of association, timeline, dose-response effect, plausibility, reversibility, coherent with other knowledge, specific to outcome, similar analogies. 1/2
-
Criteria to help distinguish correlation from causation. Various versions of BH criteria for causality, but here's good explanation: https://www.healthknowledge.org.uk/e-learning/epidemiology/practitioners/causation-epidemiology-association-causation … 2/2
End of conversation
New conversation -
-
-
Adjusting for the wrong variable can mask correlation between a cause and effect. Best to read http://bayes.cs.ucla.edu/WHY/ and https://www.hsph.harvard.edu/miguel-hernan/causal-inference-book/ …
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.