Firstly, I am aware of this question, but: 1) there is no accepted answer there and 2) my situation is not as good as the OP's in that question.

Being new to research, I have always thought it is about the new knowledge you create. Many times, you follow a theory to find out that it doesn't work. I had thought it is OK to publish such findings and to provide a constructive criticism thereof, but recently I got a second rejection which made me revisit myself.

From your practical experience, is it worthy submitting in such cases? if not, how is this situation best dealt with provided that much time and effort are usually invested therein?

I assume it makes sense to submit in principle but I am wondering whether the academic reality agrees with such an assumption.

  • 2
    See this: nature.com/articles/471448e The last sentence sums up the situation pretty well.
    – user9482
    Commented Aug 5, 2020 at 14:42
  • However, registered reports might be an option for you.
    – user9482
    Commented Aug 5, 2020 at 14:45
  • 3
    An answer being accepted sounds impressive, but all it means is that the person asking the question thought that that answer was good. So that is not much of a quality indicator. It depends on how knowledgeable the person asking the question is and if (s)he is even aware that (s)he could accept an answer. So the absence of an accepted answer means pretty much nothing. Commented Aug 5, 2020 at 15:22
  • @MaartenBuis could well be the case and also could not. However, I am aware of that issue hence the reason #2. My situation is "worse" so to say
    – Rafs
    Commented Aug 5, 2020 at 15:58
  • 1
    See academia.stackexchange.com/questions/30995/… Commented Aug 5, 2020 at 20:44

2 Answers 2


Can you make a good theory on why it doesn't work? How it fails? When it fails? Can you extend your showcase to show that? I think this would be more interesting.

The truth is, otherwise it is difficult because far too many people are reducing papers to state of the art results. I don't approve of this but this seems to be far too common.

  • 1
    I didn't quite get what you meant by "reducing papers to sota results", can you explain more please?
    – Rafs
    Commented Aug 5, 2020 at 17:22
  • Sorry for not explaining better, in my field sota: state-of-the-art is basically the best results achieved on for example, a public dataset. So people have theories, they implement those and test it on these datasets, and then you can analyze how your method compares to the best results achieved by others until then (sota). So more than often, in my field, reviewers only consider this, even if you have a different approach, a novel method etc. which might be actually insightful. It is just too easy to dismiss a theory if they aren't working just as well.
    – dusa
    Commented Aug 5, 2020 at 17:25
  • Of course this is bad practice because, again, in my field, most of the time getting good results is very much dependent on access to an abundance of human and computing power, tests and trials that finetune your implementation.
    – dusa
    Commented Aug 5, 2020 at 17:26
  • 1
    My point is, why a theory or method fails can be well written to be an interesting paper, but otherwise I feel it has lower chances
    – dusa
    Commented Aug 5, 2020 at 17:29
  • Reflecting on the original question I get that it is not that worth it proceeding and investing more resources into submitting such inferior results, no? (pragmatically speaking)
    – Rafs
    Commented Aug 6, 2020 at 11:19

There is a big problem called publication bias: only studies that "find something" get published. In part this bias is due to journals being less likely to accept papers with negative findings. In part this is due to researchers not even submitting the papers. So, the fact that you submitted those papers is an important and positive step you have taken to (try to) reduce publication bias. However, it is an uphill battle.

  • I agree that it's a nudge in the correct direction, but most of us are very tight on resources, and such a bias needs to be resolved on higher levels inevitably. How would this bias be reduced if such publications do not see the light of day? alas, slightly if at all
    – Rafs
    Commented Aug 6, 2020 at 11:15

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .