Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

> Both the claim "Spartans did this" and "Spartans did not do this" are affirmative claims that require evidence.

Hmm. That is an interesting perspective, but I do not agree.

Isn't "Spartans did not do this" actually a negative claim? And aren't negative claims notoriously difficult to prove?

What if this were consorted with extraterrestrials or some other random claim? Then it's easy to dismiss lack of evidence with "does not mean the phenomenon itself was absent".

Do we really need to find evidence that Spartans did not do this for every claim? Or rather, perhaps we should insist on proving the negative only of claims for which the evidence is strong in the first place.

Shouldn't we take as the null hypothesis "Spartans did not do this" and then evaluate evidence that they did?

In this specific instance, because of 2000 years of tradition, the unchallenged null hypothesis has been "Spartans suffered only healthy, well-formed children to live". This is the only reason people are saying anything like "We need more evidence they did not" despite the weakness of the evidence for its being true.



Null hypothesis should be: we don't know!


Well, sure, we don't know for sure that they did not kill deformed newborns, but we also don't know for sure that Spartans were not magical invisible goats, either. After all, there's no proof they weren't, am I right?


Confidence is a scale, not binary. What does "sure" mean for you and how much of it do you need to be convinced? Does it depend on the situation? For example, how sure do you need/want to be when you cross the street?

Before I consider that, you'd have to show that magical invisible goats (can) exist.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: