Ask HN: What Is a Good Hierarchy of Proof?

1 point

5 years ago

I've gradually come to see that the Gell-Mann Amnesia Effect is an abstraction even though it refers (with good reason) just to newspapers. I've become hyper skeptical (always wait for more evidence, get the other side of the story, etc.), and the evidence so far supports its effectiveness. Generally my belief is proportional to my knowledge.

But what if you need assign a strong belief to a topic that you don't have the bandwidth (time, mental, etc.) to put into learning about it? What I want to ask about is how you structure this, from the highest (I can more or less read this conclusion and basically accept it) to lowest (I wouldn't accept this conclusion if my life depending on it).

I tried to outline my own hierarchy:

* Meta-analysis of papers that have been reproduced.

* Paper that has been reproduced.

* Meta-analysis of papers that haven't been reproduced.

* Paper that hasn't been reproduced.

* Books: As far as I've seen most book reviews are generally "Did I like what I just read?" and have little to do with fact checking. This [instance](https://www.thecut.com/2019/05/naomi-wolf-interview-book-error-bbc-interview.html) comes to mind.

* Anecdotes: Dismissed out of hand unless it's addressing the equivalent of an existence proof.