Hacker News Reader

Comment by diminoten | Hacker News Reader

Self experiments like these are a great re-enforcement of how insanely hard it is to do accurate, reproducible science. This submission tells us absolutely nothing due to the lack of rigor applied.

SO many potential confounding variables, there is zero value in what is being done here.

View on HN

Here's a Nature opinion piece arguing in favor of N=1 studies: https://www.nature.com/news/personalized-medicine-time-for-o...

And another more recent one https://www.nature.com/articles/d41586-018-07535-2

The n=1 isn't the problem, it's the environment. "Some guy's room" is hardly a lab, and then, when there is so much that can go wrong, such as in this study, n=1 is terrible.

This is terrible pseudoscience and not worth anyone's time.

whatevs. I never saw the guy make grandiose claims about what he was doing, and if the results generalize to no one besides his future selves, so what?certainly useful to him and better than some half assed anecdotal justification.

and most science is absolutely buried in possible confounds. sheez

So what? So it's actively misleading and draws people to unsubstantiated claims such as what are being flooded in this comment section is what.

how is it actively misleading?

Providing useless data incorrectly asserts the data is useful in any way.

I'm impressed by how low the score of your comment is. I know computer scientists aren't "real" scientists but I thought the lack of rigor in this one was apparent and it was only upvoted only as a conversation starter. I guess I was wrong.

There wasn't a "lack of rigor"

He just never finished. He collected data and then didn't follow up with graphs and conclusions and such.

There's nothing wrong with a blog post that tells you about 2/3 of a N=1 study on yourself, it's interesting. It doesn't owe you completion, it just is not complete.

it's also funny how people assume blog posters owe them anything. get off their lawn!

No one owes me anything, but it's posted on HN and there's a comments section in HN so I am entitled to point out the (substantial) flaws in the submission.

There is a distinct lack of rigor, and there's a lot wrong with a blog post that doesn't accurately describe the environment the experiment is taking place in or any other pertinent details that would allow for a reproduction.

Completion is actually the least of the problems here.

Reply on news.ycombinator.com