The critical thing would be, as @avclub-760e8ab5cbfa2895b4b000a3b1948bad:disqus points out, finding some way of incentivizing it. If you got real credit for replication work, then it would get done. But of course that requires huge cultural shifts, which are not exactly easy to engineer.

I think this self-correction is true in the long term, but over the relatively short term (years, or even decades) it often doesn't apply. Which is an issue when you're thinking about the timeframe of careers and so forth. A true push towards replication would, I think, cut a lot of bullshit, especially from the…

Don't worry, we have.

…No?

Northern Toronto.

Birthday BBQ. Cooked a retarded quantity of food, ate a lot, drank a lot, slept a lot, played lackluster tennis with my kids for the first time, watched the first two episodes of the new Game of Thrones series, uh… I dunno, that's about it, really.

I recently moved into a very Russian Jewish area, and I can assure you that if she lives here, whoever she is, she does not got "the legs", unless "the legs" refers to balls of dough. And the sharp-dressed man is this guy who hangs outside the local discount supermarket chain-smoking in an Adidas tracksuit.

But the things they can do with snakes…

Aren't movies and music, by definition, designed for you to consume them at the exact speed it takes to play them?

"Remember, words were originally created by a bunch of apes to tell other apes where the fresh fruit is."

Shake that university, gurl!

In physics, at least, there is something like an institutional push to encourage replication, of analyses if not experiments.

Well, your original phrasing has a point. There are a number of relatively recent findings in cognitive psychology that I'm 100% confident in, because people saw the original paper, then went on to do an adjustment of it that was still similar enough to replicate the main finding. There's no formal replication process…

True, you're right about that. An increasing number of journals are making room for various types of replication publication, so that's progress of a sort.

Fuck yeah Bad Science. Great blog, great book, great guy.

I nearly cried the first time I read this.

There's a lot of men in my community with them black coats and long beards who would disagree with you.

The undergrad replications are done on things that have been replicated thousands of times, so as you say known results. But what's important is stuff that has never been replicated, or only replicated once or twice (or by the same lab multiple times). There's virtually no systematic way, in many fields, to make…

Agreed with all of those thoughts, which would go a long way towards making p-values interpretable, and probably greatly cut down on false positives. But none of them specifically address the problem of replication, which I think @epicglottisuvularsuspects:disqus has an interesting way of addressing.

I always like being reminded that people like you exist, because it gives me something to feel superior about.