Twitxy Saved

Erik Thulin

@EThulin

Moved. @ethulin.bsky.social

Behavioral social scientist working at @rare_org Center for Behavior & the Environment to get people to do good things
Seattle, WA erik.thulin.us Joined March 2007
1.4K Following 822 Followers 333 Posts

Latest posts 5

Erik Thulin @EThulin
Ryan is saving people's lives at around $5,000 each, and you could too. That's pretty cool stuff.

givewell.org/international/…
www.givewell.org Mass Distribution of Insecticide-Treated Nets (ITNs) Addendum This intervention report was last updated in April 2024. We have since published a more recent version of our cost-effectiveness analysis for insecticide-treated net distribution, which...
Ryan Briggs Verified account @ryancbriggs Jul 16, 2024
I'm hesitant to post but the @givingwhatwecan people have started to convince me that the upsides probably outweigh the down. Anyways, thanks to some youth indiscretion (stock picking) I was in a weird position & was able to give like 16k mosquito nets this year. Thanks Shopify.
View quoted post on X
Erik Thulin @EThulin
I'm winding down my use of Twitter, moving to Threads. I will probably post new papers and job postings here, but that is about it.

If you want to join me on Threads you can find me at the same username: @ethulin

If you are on Threads, please feel free to reply w/ your handle!
Erik Thulin @EThulin
I am looking for a paper that polled happiness/wellbeing experts on causes of wellbeing. I think they were polled on both the impact and malleability of the cause. The paper had a massive color coded table.

My google scholar-fu is failing me. Anyone have any idea?
Erik Thulin @EThulin
But actually do it.

I usually use frequentist statistics. But something I love from bayesian stats culture is the widespread use of 0.89 as a threshold, which I believe @rlmcelreath started.

Why use 0.89 instead of a round 0.90? To remind ourselves how arbitrary the cutoff is.
Null-hypothesis testing, and its handmaiden the p-value, is a paramount tool for applied social science. But it's been stretched to the breaking point—so we're proposing a new threshold for statistical significance: p < .042.
ideas42.org/wp-content/upl…
View quoted post on X