Should hype-train-riding pre-print articles be considered notable?

Question

Inspired by a comment on this question.

Some research basically re-examines some previous claim with slightly changed methods or in slightly changed conditions, acting on assumption that previous research missed some nuance and the results might differ with this update. Some of such articles even manage to successfully overturn previous assumptions; but some of them are just busywork that would be rejected by any competent peer review. Peer review is important in culling the weak. On the other hand it increases the publishing cycle length, and increases chances that you start working on a problem someone is already getting ready to publish researc on; that spawned a practice of distributing preprints - a way to tell other scientist "hey, I took this subject, if you start on it, you are competing with me". This practice was limited by its very nature - without centralised publishing, there were no way for these publications to reach significant notoriety outside of its respective field (or even within it) other than by word-of-mouth - in other words, it also passed through a kind of peer review; and efforts to increase this non-reviewed exchange of information tended to meet resistance from the scientific community.

But with advent of Internet, everything changed. Now preprints can easilybe published for everyone to see, and research related to high-profile events can just as easily make headlines of mainstream media bypassing any review by people who can actually understand what the authors are talking about.

What does it mean for this site? It means that "busywork research" I mentioned can't be shut down on the journal review stage, now it can be grabbed by sensation-hungry mainstream media off preprint sites (thus becoming notable) and spawn a stream of questions with only a slight change in claims, and with standard answers along the lines of "the research isn't sound and will be rejected during review for publishing".

Are questions with claims based on research articles that are yet to pass peer review notable enough for this site?

Answer 1

I think this question is shining a bright light in the wrong place.

It discusses the importance of peer-review, the changing way that publications are peer-reviewed, and the way that claims can be plucked from low-quality articles - and there is plenty more to discuss and highlight.

However, our question of "notability" (as we define it here) is not "is this claim well-founded?", "is the claim carefully reviewed?", "is this claim just being made for sensational headlines?", "is this research well-conducted?".

Our question of "notability" is "Is this widely believed?" That is very hard to show, so we accept a looser definition: "Has this claim been widely read or seen?"

Applying this notability check, we see that most pre-print articles are not notable because they aren't widely read. However, when the "sensation-hungry mainstream media" grab some claims (and possibly twist them) and published them in their pages that is what makes them notable.

Whether the claims are originally sourced come from peer-reviewed meta-studies, unreviewed pre-prints, politician's spin doctors or made up whole cloth by celebrities, if they are widely believed/widely read, they are considered notable here.

Now, that means we do get questions about bad science and Not Even Wrong claims that can be frustrating, but tackling those claims (and more scientifically plausible ones) is our raison d'etre.

When someone posts a nonsense link on Facebook claiming that some 10-year-old pre-print proves that Albert Einstein was an android, I want to be able to post a Skeptics.SE link that explains, with references, why it is nonsense.

Note: Our scope explicitly rejects research-level questions - which is to say, if there isn't a scientific consilience on a topic, you can't expect a rag-tag crew of Skeptics.SE users to resolve it here. You can use the "Unresolved Current Event" close reason, if that is the case, but answering it with evidence that it is an undecided question is even better.