The Department of Energy does pandemic origin investigations? https://www.wsj.com/articles/covid-origin-china-lab-leak-807b7b0a
@m_scribe not too shocked. The DOE mainly handles nuclear weapons design and stockpile stewardship, and has a history of doing things like running bioweapons programs out of their labs. They're reasonably likely to have people with the background knowledge to investigate claims that a pathogen is an escaped bioweapon.
@dsacer yeah I’m reading about this now. It’s interesting that they have “low confidence” in the report.
@skua @m_scribe @dsacer The article is deceptive because "most likely" has a different common meaning than statistical meaning. Many, if not most people, will interpret that as meaning "it was a lab leak".
For the statistical case, you need both the MLE and the uncertainty in that estimate. If you have poor information quality then the uncertainty is so large that the MLE estimate is not likely to be close to the "true" value (if you're a frequentist). In that case it's better to just say that the analysis is inconclusive.
@twitskeptic @skua @m_scribe I expect no less in terms of deceptive description from the WSJ
@twitskeptic @m_scribe @dsacer
I am not familar with the technical terms you use.
But I can see that many preole would be be misled by "most likely".
The phrase "more likely" would have gotten through to nontechnical people much better as it evokes "comparison".
Though WSJ readers would be likely to be more familair with communicating about risk as that is part of dealing in the financial markets.
@twitskeptic @m_scribe @dsacer
AIUI
If I could predict the likely outcome of a coin toss at an average of 51% correct calls then each specific call is more likely to be right than wrong.
But there'd only be low confidence that I'd get any specific call correct.