I'm a grad student in planetary science, and I'm writing my first academic paper. This will be the first paper from my numerical model. My results are timely: recent rumors in my community indicate that my model's predictions may be on the right track.

Papers in my field often fall in one of (at least) two categories:

  • "Methods Papers": These papers aim to completely describe the research methods, equations, data sources, and results in great, reproducible detail. Consequentially, they tend to be long, and results can be buried in the details.

    • In the context of numerical modelling, such papers might derive the important equations used. They might also describe the details of their numerical implementation, including the governing equation(s), boundary conditions, and so on.
  • "Results Papers": These papers aim to be concise and to the point, which is to highlight the results of their research. Sometimes, such papers are rushed in order to present timely results before competing groups do. The methodology is kept to a minimum so that the results take center-stage.

    • In the context of numerical modelling, such papers might cite the method/equations/data/etc. from another paper, perhaps with certain modifications, rather than writing them out explicitly. Only the most important equations would be given.

How should I decide which approach I should take? What factors should I consider?

I see this as really asking how to balance the following factors:

  • Thoroughness — ensuring exact repeatability
  • Conciseness — so that the reader doesn't have to read a bohemoth, and/or doesn't accidentally skim over the important parts
  • Highlighting results — emphasizing the model's predictions/implications
  • Timeliness — getting it published soon

I feel like I can emphasize one, pick two others, and let the remaining factor fall by the wayside. (Did I miss any other factors?)

  • 2
    I should add that I don't think that "Methods Papers" and "Results Papers" are necessarily mutually exclusive. – jvriesem Jul 25 '17 at 18:49
  • Part of the answer to your question (possibly all of the answer), depends upon what direction you and your advisor want to take your thesis. Furthermore, this answer would be highly specific to you. – Richard Erickson Jul 25 '17 at 19:45
  • @RichardErickson: My solution is absolutely specific to me and my situation. It's probably not worth anyone else's time. That's why I'm asking how to decide — what factors should be considered —rather than asking "What should I decide for my situation?" This question should have an answer that is useful to the broader community. – jvriesem Jul 25 '17 at 20:00
  • 1
    @jvriesem That's an interesting way to approach it to make it more broad, though that 'how' will still depend a lot on individual factors. However, note that the title of your actual post is "Should X or Y". – Bryan Krause Jul 25 '17 at 20:08
  • If your claim is big, and depends on a new method, and you have enough time, writing a methods paper first will reduce the ability of reviewers to shoot down your claim in the research paper. If you heard rumors and someone else is also working on the same topic, you ars likely late and want to submit to the same journal as them (and openly talk to them - as it is nice and as they will likely be ahead of you). – tsttst Jul 26 '17 at 3:24

Why not both?

I'm not in your field (and methods versus results papers in my field break on different lines than the ones you describe), but in my field the methods papers can be among the most influential and cited works, but iff they are sufficiently unique and sufficiently generalizable.

If you have a method that fits those categories, and also have data that speaks to a particular result, and there is a tendency to split them in your field, there is nothing wrong with making it into two unique papers.

As an additional caveat: as a student, you should have an academic advisor who is well-positioned to give you advice on things like this. Please consult them before taking too much advice from the Academia.SE folk. We try our best, but we are not your advisor.

  • My dual advisors weren't sure which approach to take, so I'm just getting started and we'll figure that out later. Of course I'll ultimately rely on their advice, but I saw an opportunity to ask a question here that may be useful to others. As the saying goes, "Plans fail for lack of counsel, but with many advisers they succeed" (Proverbs 15:22, The NIV Bible). – jvriesem Jul 25 '17 at 20:05
  • 1
    @jvriesem Good on you - that's the right approach to take, for sure. You might be surprised how many questions we get where the student asker seems to not have considered asking their advisor first, for reasons I don't quite understand. Fear maybe? – Bryan Krause Jul 25 '17 at 20:07
  • I don't think my method would be generalizable. It involves solving a 2D PDE using a finite-difference approach with specific boundary conditions. I imagine the high-level science aspects would be useful to others, while the details of the numerics would be far less useful. – jvriesem Jul 25 '17 at 20:08
  • 1
    Ok, that suggests to me that the methodology is a lower priority for publication then (if it fit those other categories, a new methodology could jump start a whole career). If your field uses the arXiv maybe it would make sense to put a detailed methodology there, while preparing and publishing a more concise results paper? The methodology could always go to a low-impact journal. I really can't speak more than that to norms in your field though. In biology, all papers contain a fairly detailed methods, and often publish supplementary material when necessary to go into methods in depth. – Bryan Krause Jul 25 '17 at 20:11

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