What's the point of peer reviews if such errors are not detected?
leephillips•6mo ago
As a reviewer I try to judge whether a paper is scientifically relevant, appropriate for the journal, advances the field, credits previous work, and is not obviously wrong. No reviewer can guarantee that there are no mistakes.
palata•6mo ago
Doesn't "wrong unit" or "wrong decimal" count as "obviously wrong"?
> No reviewer can guarantee that there are no mistakes.
Of course not. But I would assume that the goal is to kind of understand the work. If you review my code by saying "it does compile and the name of the variables suggests that it is relevant for the project", that's not enough. You're suppose to get an understanding of it.
I am not saying you should catch every bug, but if I do "int fraction = 3/5" or "val dist_in_meters = dist_in_kilometers * 100", I would expect that you see it. And for an academic paper, there are more than 1 reviewer, right?
palata•6mo ago
leephillips•6mo ago
palata•6mo ago
> No reviewer can guarantee that there are no mistakes.
Of course not. But I would assume that the goal is to kind of understand the work. If you review my code by saying "it does compile and the name of the variables suggests that it is relevant for the project", that's not enough. You're suppose to get an understanding of it.
I am not saying you should catch every bug, but if I do "int fraction = 3/5" or "val dist_in_meters = dist_in_kilometers * 100", I would expect that you see it. And for an academic paper, there are more than 1 reviewer, right?