The comic is very true. See also the movement for papers that can be verified (open data for AI models etc). Oh and I can tell you that you can be even declined from a conference when publishing a continuation of your previous work: “there is already a paper XY on that” “that is mine, that was the first publication on this, this takes it further” “bye, duplicate of XY”
I am not sure I get what you are saying; you have cited the work in question and explained why your current paper is an improvement over the state of the art?
(OTOH, I agree, continuations are typically hard, because the bleeding edge research in some of the topics are of interest only to very few people until it filters down to the rest of the community, and it can be difficult to make the point why your current work is a significant advancement.)
You don’t get to explain anything. You just get a deny from the reviewers that somebody else already published something about that stuff. Although they can totally see that the paper they referenced was also from you and they should’ve just read part of the paper to get the idea..
Just to confirm; you do cite the previous paper in your related work and explain what you do new right? I have had the same problem as a submitter, but I have been also on the other side of the fence as a reviewer, and it can be hard to convince other reviewers that a paper is good when the new contributions are not clear.
In particular, what has worked best for me, is to start by mentioning your previous work in third party, in the introduction itself, and talk about how this paper advances from there.
This is painfully accurate!
The bit about scientist and reviewers not checking references, etc. due to it being time consuming is so true, even from my research days a few years ago. When I started working at the National Cancer Institute I (perhaps naively) spent too much time thoroughly checking all papers/references, and would constantly wonder how others were able to be so much more productive. Evidently cutting corners was rampant.
A shameless (but hopefully relevant) plug: the current project I’m working on is at scite (https://scite.ai), and one of the things we offer for scientists and reviewers is the ability to (quickly) see how papers are cited by others.
In fact, one of our latest additions is the ability to upload a manuscript and see how its references were cited (especially useful when checking if you’re citing heavily disputed or even retracted work).
I thought it’d be relevant to bring up here because we (like many others whom this resonates with) want to improve this system by working within.
Curious to hear what others think…
This is a nice approachable exploration of the problem, but the solutions at the end seem a little naïve.
Scientific journals can raise their standards
Under the current system, people pay to publish their crap in journals, so the journals are motivated to publish as much crap as they can get away with without too much damage to their reputation.
Universities can change their hiring policies
Hopefully they can, but I suspect that using metrics like numbers of published papers saves them a lot of time.
Funders can fund the boring-but-important stuff instead
Public money for research is relatively scarce in most countries, so many researchers are forced to seek funding from industry. Industry generally wants technology they can sell in the near future as a return on their investment. I suspect that we can’t just persuade the current “funders” to fund different things, we need different funders (or at least a different balance of funders, e.g. more public funding for research). At that point it becomes a somewhat political issue, with all the complexities and inertia that that incurs.
Those who have built their careers on the current system are motivated to maintain the status quo, while outsiders face an uphill struggle to be heard at all.
I hope that we will see real change, I just feel like the conclusion of the comic was a little too optimistic and implied that we knew how to solve the problem.
Not much more than sharing a dank meme and plugging some dude’s book. Points for sharing; valid points nonetheless.