News for hackers. Stuff that matters.

A comment on EPA's proposed new rule would allow it to ignore the best available science. Parent comment

Re:EPA's proposed new rule would allow it to ignore the best... by ouid on Friday July 13, 2018 @06:14PM (#17527313)
New research is overwhelmingly more likely to hurt industrial interests than help them, so adding any hoops at all is likely going to benefit industry. In fact, all that this does, from a regulatory perspective, is provide a pathway for the EPA to accept and reject studies at its own discretion, without having to ever actually call the result into question.
[ Parent | on HN ]
Re:EPA's proposed new rule would allow it to ignore the best... by refurb on Friday July 13, 2018 @06:27PM (#17527405)
The EPA already weights the value of studies in a public way.

In terms of new research are you saying that it’s ok to keep the raw data a secret? Why? We don’t do that for any other type of science.

[ Parent | on HN ]
Re:EPA's proposed new rule would allow it to ignore the best... by ouid on Friday July 13, 2018 @06:42PM (#17527502)
Yes we do, and it doesn't matter. Data isn't relevant to the reproduction discussion. The reproducibility crisis is a crisis of methodology, not falsified or misinterpreted data. If I p-hack a result, then no amount of transparency is going to demonstrate that that's what I've done. The only standard of evidence for reproducibility is reproduction, and they are only using the "reproducibility crisis" as a diversionary tactic to increase the cost of doing science to an unacceptable level, at the same time as pulling EPA funding of science.
[ Parent | on HN ]
Re:EPA's proposed new rule would allow it to ignore the best... by refurb on Friday July 13, 2018 @07:53PM (#17527810)
Give me some examples of science where we make decisions and keep the raw data a secret. I can’t think of any except ones involving national security.

Yes, the best way to prove reproducibility is to reproduce it, but that can’t always be done and at any rate is no excuse for keeping raw data a secret.

[ Parent | on HN ]