gatopingado

That one egg was forty eggs?

Cool experiment!

Once the author started taking pills independently of their stress level, the variance of differences diminished a lot. I'd wager this supports the mean reversion hypothesis.

Also, while I agree with their general conclusion that theanine probably doesn't reduce stress, I'd give assign more probability to the hypothesis that theanine does work, but in other design settings. For example: drinking tea instead of taking pills, or measuring stress levels after a day instead of an hour, or evaluating the difference across time instead of in time chunks.

Thanks for noting! This is the right link for [1]: https://arxiv.org/abs/2210.01948

This is correct. There's been a lot of interest in e-values and non-parametric confidence sequences in recent literature. It's usually refered to as anytime-valid inference [1]. Evan Miller explored a similar idea in [2]. For some practical examples, see my Python library [3] implementing multinomial and time inhomogeneous Bernoulli / Poisson process tests based in [4]. See [5] for linear models / t-tests.

[1] https://arxiv.org/abs/2210.0194

[2] https://www.evanmiller.org/sequential-ab-testing.html

[3] https://github.com/assuncaolfi/savvi/

[4] https://openreview.net/forum?id=a4zg0jiuVi

[5] https://arxiv.org/abs/2210.08589

A logistic regression is the same as a Bernoulli regression, which is the single trial case of a Binomial regression [1].

[1] https://www.pymc.io/projects/examples/en/latest/generalized_...

For anyone interested in anytime-valid testing, I wrote a Python library [1] implementing multinomial and time inhomogeneous Bernoulli / Poisson process tests based in [2].

[1] https://github.com/assuncaolfi/savvi/

[2] https://openreview.net/forum?id=a4zg0jiuVi

FYI, "Why we sleep" is not a reliable source [1].

[1] https://guzey.com/books/why-we-sleep/