I think if people click play and see a total run time over 10:00 they bail.

I bail over 4:00, so this 8:47 is also unplayable to me. Let me know if he addresses the failure of reproducible results and their willingness to commit gross crimes against humanity for fascist governments. Otherwise, it's just a fluff video.

People regularly listen to audiobooks at double or even triple speed now, it's probably for those sorts of people. The ones who value quantity of information over quality of experience.

I spent much longer than 8 minutes listening to the audiobook of Baumeister's "Willpower." The man is a goddamn William James fellow. And it was a great book.

On the other hand, here we have Mark McGrath jr. predicting the demise of a field of inquiry as old as civilization, on Youtube, in a hoodie. So yeah, he needs to get to the fucking point before the microwave beeps.

Ok, Sugar Ray's talking some sense.

I did listen to it at 2x. His main points are that the file drawer problem exists (already knew that), psych experiments lack rigor depressingly often (ditto), and that replication hasn't been kind to the famous "cookies & radishes" experiment.

The last part was new to me, but it seemed reasonable to skim past the parts which were just about introducing concepts I've already learned to people who might not know about them yet.

I'm with you on "quality" over quantity, of course. 1000 tweets don't equal long form journalism. I might also call it nuance, or just real investigation. But this video isn't that. The Slate article comes closer, where besides the usual breathless stuff about A Goddamn Crisis in the Sciences, they mention that Roy himself takes issue with how the "letter e" study was designed.

And why not? The other researchers tried to replicate his ego depletion work *with a completely different study.* So, in other words, they didn't replicate it. Brilliant job, guys. But that's psychology for you.

Replication isn't bulletproof. Even data sharing has its pitfalls. A standard product defense industry trick is to "replicate" studies linking chemicals to illness because it's much easier to be sloppy and not get a result than to be really rigorous enough to expose the easily-buried correlation. You can even do this with the same data but a less careful analysis. (Ironically, this is why we have data access laws. Big corporations lobbied for them so that they could take underdog studies linking e.g. Vioxx to strokes and then publish another 5 that use the same data to deny the link. Lay juries hear 5 vs. 1 and suddenly it's game over).

Does that mean ego depletion is real? No. I'm just saying that it makes sense to have a policy of reserving your full attention for people who make nuanced points, or novel ones.

I do think grad students should have to replicate results before being allowed to contribute to original research, both as a way of learning about methodology, and ensuring that the replication gets done. Publication could happen in free electronic journals that the universities maintain.

Isn't that why everybody goes into the field?

HI NOMINAL