davidmanheim

Maybe all studies should be preregistered, including their methods... like this one was?

https://osf.io/vzdh6/

This paper was pretty clearly pre-specified here; https://files.givewell.org/files/DWDA%202009/IPA/Masks_RCT_P...

You might want to click some of the links or read other content on the site to get more of that supposedly missing context.

The source might be: https://medium.com/the-year-of-the-looking-glass/how-do-you-...

There's no cooking allowed on the Sabbath, and women can and frequently do go to synagogue with the men. (But if they stayed home, they were probably setting the table while the men were in the synagogue.)

Interestingly, all ISPs and phone companies in Israel are required to offer server-side filtered internet for all customers who request it, at no extra cost.

This "article" doesn't actually have any substantive evidence for the claim in the title.

Non-interactive zero knowledge proofs.

ZK proofs have a number of good explainers, mostly using graph colorings. Non-interactive versions, however, require quite a bit more than that explanation allows - and despite asking experts, I still haven't found a good, basic explanation.

Right - but there is also a relationship between IQ and educational attainment. So controlling for it means that you're measuring something fundamentally different. That's fine, but you can't just control for things without having a valid theoretical construct to justify it. (Well, OK, you can, and people do, and it's great for getting p<0.05 if you don't care about getting meaningful answers about the underlying questions.)

So does frequentist stats - they are just different axiomatic foundations and assumptions.

Having taught frequentist stats as a TA to grad students, I understand why frequentist stats seems not to make sense. On the other had, my prior on teaching quality, and my data on the relative difficulty of understanding the approaches says with near-certainty that your experience has nothing to do with the approach taken.

Having used Bayesian stats heavily, I'd note that the hard parts are not gone, they are just located elsewhere - in how to actually do the computations, rather than in how to set up problems. Each can be taught poorly or well, but given that MCMC is certainly harder than least-squares, it seems difficult to argue that using Bayesian statistics is easier. (Unless you're not just applying the methods by rote, and letting the computer spit out answers - and if you are, I don't know why you are better off with Bayesian methods. In fact, if that's what you're doing, please stop doing statistics and pay an expert instead.)

What does this have to do with Bayes?

A frequentist assessing probabilities to make decisions about how to respond is in, if anything, a far worse position. The Bayesian would ideally use priors on groups and cross-group correlations to note that the weak evidence that purple -> mean barely shifts from their priors, and the inter-group mean shows that they are likely to be nice, unless your prior is that different colors have nearly independent probabilities of being nice or mean.

That's often helpful, but not sufficient in general. See the dynamics discussed in my article here; https://arxiv.org/abs/1803.04585

This is probably where I should mention my two longer articles that discuss this in depth; https://www.ribbonfarm.com/2016/06/09/goodharts-law-and-why-... https://www.ribbonfarm.com/2016/09/29/soft-bias-of-underspec...

As well as my more recent paper categorizing the different ways it occurs; https://arxiv.org/abs/1803.04585