Great find… All this time I’ve gone with ugly alpha hacking and overly smooth kdeplots.

I agree, although I think the same might be said about the anecdotes you are citing, which sound hard to believe and designed to flatter the reader’s own biases. I’m open to being proven wrong but it sounds a bit sketch.

Twitter is now a very unfriendly place for people with left wing views. But I actually think it’s more friendly to moderates than Bluesky. And I’m not just talking about the composition of the population. If you post centrist stuff you have a higher chance of getting trolled here than there.

It’s of course a very different experience for people with left wing views over there, especially on social issues.

I lurk on here a bit, but mostly use Twitter since vaguely moderate views are still more accepted there. I never get trolled there but know I will here. Not the only metric that matters of course, but relevant to me personally.

Idea #3. Bounties to catch non-transparent deviations from preregistration Research that deviates from preregistration is fine when done transparently, but misleading when done non-transparently. Most readers don’t have time to investigate these discrepancies, so let’s create bounties instead!

Idea #2: Bounties for artifact detection in the social sciences. We recently learned that some troubling trends were caused by changes in how the data was collected. Let’s catch these issues sooner by creating bounties for the people who know where the data comes from.

Idea #1: Bounties for fraud whistleblowers. As I argued last week, whistleblowing is undersupplied in science. To fix this, we should borrow an idea from the SEC and copy their popular bounty program.

I don’t understand this victory lap. The people advocating for a more open process knew it was unlikely but thought it was worth trying. What exactly are you celebrating?

Who chooses? One of the options in the screenshot below. Who do they choose? Harris would be the front runner. I can promise you I am aware of the many challenges that might emerge with doing this. My point is that this is not insurmountable. It is possible if people overcome their hangups.