Hacker News new | ask | show | jobs
by yayday 3640 days ago
There's already a "trustability" factor: it's called a name. When I search for a new paper and find something interesting, I look at the first author, the name of the PI, and the name of the university or laboratory associated with this paper.

If I'm familiar with the work of the first author then I already have an idea of how much I can trust his/her work. If I don't know this person then I look at the coauthors and especially the PI. If the PI is a heavy hitter in my field then there's my trustability factor right there. If I don't know the PI but I know the laboratory or university it's affiliated with then there's my third line of a trustability metric.

1 comments

You don't consider the journal? You're a better man (or woman) than I am, if so. Otherwise agreed, reluctantly. This implicit trust is frequently manipulated to relax standards by experienced PIs, and in biomedical work that is a source of numerous problems.

In an ideal world, work from an unknown author would receive the same scrutiny as that from an established lab, and vice versa. Obviously this isn't the case, but the further we drift from it, the more likely we become to see sloppy results wasting everyone's time.

It can't just be about the narrative. That simply isn't science. That's storytelling, and confusing the two has caused a great deal of harm to science in the public eye.

I'd be lying if I said the journal didn't matter strictly speaking, but in general as long as it's a "real" journal (as opposed to one of those obviously fake journals I keep getting spammed by) then I don't really care.

Don't get me wrong though, just because the paper has a heavy hitter as the PI or even as first author, doesn't mean I treat the content with less scrutiny. It's more that I know my time is extremely limited so I'm more willing to spend more time dissecting a paper from an established first author or PI than an unknown person (unknown also meaning that it's not a paper someone recommended to me or a paper someone cited in another work).

It's more like, I'll _make_ time to read papers with certain authors on it because I know that they do good work. I'll also _make_ time to read something if it's published in Nature et al for example. But a random paper from a random author in a random group? It's not that I don't trust this person's work, it's more that I'm unfamiliar with that person's work and there's other things I could spend my time doing as well. I'll download the paper, throw it on my iPad, etc, and I'll get around to it eventually.