Statistical modeling and Bayesian inference, cognitive psychology, and sundry other things

Medium member since May 2018

Jan 18 · 1 min

Thanks! The 6% was based on the first figure. I overlayed the ANSUR description of the hand measurement with the outline of Trump’s hand, and aligned the parts that should not be affected as much by the fingers spreading: one side and the bottom of the palm. I then aligned the middle fingers. Measuring from the tip of the finger to the ANSUR…

Jan 5 · 1 min

Thanks for an insightful set of comments, as ever!

2

1

Hi Zoltan,

I don’t have any problem, in principle, with lowering the default alpha, though I’m not convinced it is necessary. My main problem is with the argument that I think misses the relationship between Bayesian evidence (formal, related to changes in credences) and frequentist evidence (informal, related to error…

1

Jan 2 · 1 min

Very interesting post, thank you!

1

Hi Paul, The Bayes factor in the first example is a ratio of marginal likelihoods. It’s easy to prove; just write the posterior odds out as the ratio of two integrals, and you’ll have to correct for the prior odds in the integral over the negative and positive parameter spaces. The ratio of marginal likelihoods pops out as the factor by which the…

1 response

Nov 27, 2017 · 1 min

You wrote: “ What I find surprising — and indicative of the problems we face across the sciences —…

2

Hi Björn,

That’s not really the same point. Of course one can aggregate statistical evidence and set a meta-analytic criterion of .5%; however, I am arguing that “discovery” of an effect is a broad scientific issue, not a narrow statistical one. It requires a constellation of results, some positive, some negative — or a…

7

Jan 12, 2017 · 1 min

I think it’s particularly bad when sites like InfoWars are likely to take liberties with the…

1

Given that the primary source is on the internet and citable, I can’t see any justification for citing Infowars — and not the primary source — except that you want people to read InfoWars’ take instead. It shows very poor scholarly practices, at the very least.

1 response

Dec 13, 2016 · 2 min

Should we cite blog posts during peer review?

4

2

The strength of a point is not changed a whit by where it is written.

Suppose you are responding to reviews on a paper and you are justifying using method X for an analysis rather than more common method Y. The reviewer would like you to use the more common method X, but let’s say that in your situation, application of Y…

1