14

I've started a research, and after replicating methods of other researchers (running experiments on my data), I've realised that I cannot add anything new to the field. The work contains no improvement of the methodology, I just used the established methods to analyse my data and I ended up in having some results. Consequently, I cannot publish a paper out of it (which I am almost done writing).

Nevertheless, the obtained results are new (applied to a specific problem), so they could be of interest. Furthermore, I don't want that my time and the possibly valuable results end up in the trash bin.

What can I do with it? Is arXiv suited for this?

  • 2
    What is your field? Why do you think you have to improve the methodology in every paper? That seems fundamentally unsustainable. In many disciplines researchers will use a de facto standard methodology, because it makes their results comparable to the results of others. – Rikki Sep 6 '15 at 16:32
  • 1
    There are TONS of papers here people dont improve a methodology but shows great achievements using a methodology or a combination of methodologies. Of course, this is field dependant: wont happen in maths, but very likely will happen in chemical engineering. (random examples, no specific feeling against either of them) – Ander Biguri Sep 7 '15 at 10:18
24

You say that your work on methods is uninteresting, but that you believe the results to be valuable. That's great! You should write a paper about the results!

Sometimes people feel like the amount of manuscript text they produce for each aspect of their work should be proportional to the time and energy put into that aspect, e.g., if you spent 3/4 of your time working on the methods, then your text should be approximately 3/4 about the methods. That is incorrect. The amount of space something takes up in the text should be proportional to how interesting it is to the reader. Sometimes you have to put in a lot of work on something that is very simple to explain, and that's OK. Focus on the bits that will be interesting to the reader, and instead tell your friends the stories of how long and hard you worked on the "uninteresting" bits to get there.

14

What to do with results that ended up primarily or entirely as replication studies depends a bit on the culture of the field.

  • A preprint server such as arXiv or PeerJ should always be a suitable place to report on replication studies. Of course, for CV and visibility reasons, a reviewed publication would be preferable in most cases.
  • Some journals, specifically PeerJ and PLOS ONE, have a reviewing policy of judging papers solely on technical correctness, but not on novelty. That is, a paper that is sound but "only" replicates results that were previously already known is perfectly acceptable in these journals. The reasoning is that replication studies actually often provide significant value to the community, as independent replications can increase the trustworthiness of results.
  • Some more traditional journals are also (more or less) happy to accept replication studies, but the extent and conditions vary by journal, and, more importantly, by field. In medicine, for instance, replication studies are extremely common, and practically all journals accept them as far as I know. In applied computer science, replications are (unfortunately) almost impossible to publish outside of PeerJ Computer Science.
  • Even journals that per se do not allow for replication studies sometimes have special issues dedicated to them. Of course you need to get lucky to find a suitable special issue with a topic and timeline that suits you.
5

If the method you used to obtain the results -- albeit established -- are not traditional, then by all means you may publish them. Make sure you include relevant literature review to prove you've done the necessary ground work in the manuscript.

Any fact that you believe might benefit the community of your research may well be deserved to be submitted for publication. To what extent it deserves to be published depends on the reviewers.

1

Publish the results of the experiments in a conference or journal. If it won't get accepted to your original target because of the lack of methodological innovation, look around for another.

From the outside (and with very little actual detail), it sounds like a somewhat interesting observation that you tried all of the typical methodologies in your area and found now (current) way to improve them. I think you would have to structure your evaluation of those methods (i.e. what were the criteria that helped you judge that they did everything you needed?), but you could contribute to your community by identifying that. If it's wrong, the peer review you will receive will guide where to target future work.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.