Plan the paper when you plan your project

Advice on paper writing over at NatureJobs is mostly boilerplate but I was struck by the first observation.

 

Consider the final paper when you first plan your project

 

This doesn't seem as obvious to trainees as one might think, particularly postdocs.

 

Sometimes I think it is the most essential role of the mentor to keep harping on "how is this going to contribute to the story, how is this going to fit into a paper?"

 

15 responses so far

  • Beaker says:

    Not just for projects. Heck, I do this for every experiment. If you don't know where you want to go, how can you expect to get there?

  • drugmonkey says:

    Beaker-
    Perhaps because at times, the most logical or most interesting next experiment is not equal to the best experiment to advance towards a manuscript you can submit?

  • Namnezia says:

    I think its a balance. While you shouldn't lose track of your overall goal to make a complete story, there should be some room for following interesting tangents along the way that might form the basis of new projects.

  • darmok says:

    oh boy. That one sentence has been the theme of the last two years of postdoc work I've been doing and why a good bit of it is going to go down the drain. It's not that I didn't know it was good practice to think about projects that way. But the reality of a lot of excitement about a new research direction, and very quick data collection, meant we cut too many corners in the planning stages. The theme was essentially, "just collect the data, we'll figure it out later"

    I had an inkling it that their were some gaps in our planning but didn't speak up as much as I should have. So that's my responsibility. Unfortunately my mentor only facilitated the lack of planning. He was excited and wanted data ASAP. He kept saying that everything was going great. Once we started trying to write things up it became clear that was not the case. Since we were going in a new direction we really should have taken a bit more time in planning.

    I could go on and on, let's just say lesson(s) learned.

  • If you don't know where you want to go, how can you expect to get there?

    The whole point of science is to end up in places that you didn't even know existed until you arrived.

  • drugmonkey says:

    So perhaps the natural tension of trainees following their noses and the PI keeping down to business with their eye on the prize keeps everything humming along, Namnezia?

  • And sometimes it's the opposite: the PI has to nudge the trainee into not just staring at the wheel in front of them, but to look further down the road or on the side.

  • drugmonkey says:

    was that a sports analogy dude? careful, FSP might be lurking about...

  • Namnezia says:

    I had a postdoc who had her "Freaky Friday" experiments where she would follow crazy tangents/ideas that differed from her "main" project. These experiments led to two additional papers and a new prep in the lab.

    And I think that's key. Everyone needs some fucking around time, where you can just try things, even if they don't seem at first likely to work. If they look promising you can then design 'proper' experiments to follow up. I think people don't experiment enough with their experiments.

  • drugmonkey says:

    There are also those for whom your post on "data sink" is laughably absurd, Namnezia, because their "experiments" take place not on the order of hours but rather on the order of weeks to months. Yes, everyone needs some screwing around, exploratory time but there also has to come a time to pinch it off and generate a useful story.

  • Beaker says:

    Before the experiment I imagine the best possible outcome (and its culmination in publication)--and that is what motivates me to put in the effort (or to motivate a trainee). Seldom do the results of any one experiment meet the expectations. But every once in a while they do, and that is the dopamine squirt we all crave. When it leads to the unexpected, even better.

    I did not mean my comment to be viewed through the lens of grantsmanship, where unexpected results are to be avoided and exist mainly as "pitfalls" in the eyes of the reviewers.

  • Alex says:

    Thinking about whether that next step will fit into the paper is fine, but also make sure you think about what steps would be necessary to falsify your results, and perhaps kill the paper. If you design the study to prove a point, you might just guarantee that you prove the point, which is not always the same as guaranteeing good science.

    On the other side, design the study a bit more broadly than that if at all possible. We had a paper rejected because the reviewer said that we proved a very narrow point, and he'd consider the study more interesting in his eyes (we can discuss some other time whether he made the right subjective call on the significance of the paper and the appropriateness for the journal) if we proved a bigger point.

    So we went back, did some work to prove a bigger point...and discovered that there's a whole other dimension to all this that nobody anticipated and nobody has discussed previously in the literature. If we'd been thinking beyond the paper, we would have found this sooner, and gotten an even bigger and better paper out the door sooner.

    Kudos to that reviewer.

  • there also has to come a time to pinch it off

    Dude, I'll take sports analogies over defecation analogies any fucken day.

  • antipodean says:

    DM" "experiments" take place not on the order of hours but rather on the order of weeks to months"

    Try YEARS to complete a proper study in humans. In these cases you could write most of the paper before you started and then fill in the missing spaces once you have the data.

  • Confounding says:

    I was taught this very early on, and it is one of the things I must thank my mentors for. I've seen plenty of projects from my peers run aground with "Crap, how do we publish this thing we've been working on for two years"?

    I think the focus on framing dissertations in terms of n number of papers, rather than a magnum opus, is probably a good start.

Leave a Reply