If you look at my publications list, you’ll see that it doesn’t look up to date. The most recent paper on it came out in 2015. And it’s true that it’s not up to date–but only because I’m a co-author on a couple of papers that got accepted in the past week.
Which means that in terms of publishing papers, I went 0-for-2016. I went almost two years between acceptance letters.
That’s for a few reasons. In the fall of 2015, I spent a few months focusing on other things besides writing papers. I’ve always worked best that way; I like having various irons in the fire so that I can switch between them as the mood strikes me. But in retrospect, I probably should’ve forced myself to sit down and write papers in fall of 2015.
Then, when I did go back to writing papers, they all got rejected, in a couple of cases more than once. In some cases, that was just bad luck with reviewers. It happens. But in retrospect, I probably thought too highly of one of the papers concerned and aimed too high with it. Then I stubbornly continued to aim too high with it after the first rejection. Repeatedly aiming too high is the “failure mode” of my approach to publishing. I never intentionally “reach” with papers hoping to get lucky, or try to “sell” my work as more novel or important than I honestly think it is.* Rather, I try to do good fundamental research that will be of interest to a broad range of ecologists and then publish it in good, selective journals read by a broad range of ecologists (what Brian once called “slow science“). That’s in part due to my own preferences. It’s in part due to me playing to my strengths.** And it’s in part due to the fact that I’ll never have the resources to have a big lab, particularly a lab with postdocs. The bottom line is, if you’re going to do shopkeeper science, your “shop” has to make up in quality what it lacks in volume. Which means that, when I do misjudge how good one of my papers is, it’s invariably by aiming too high.
I’m not too worried about having a blank year on my cv, though of course I’m not thrilled with it either. It helps to know I’m not the only postdoc or prof to whom this has happened (blank years are of course common for grad students). I just did a bit of casual googling of Google Scholar profiles, and in a few minutes found seven friends of mine in EEB who had blank years as postdocs or faculty members. None of the blank years were associated with parental leaves or other extenuating personal circumstances. They’re all very successful tenured profs at top research universities now.
I would’ve been more worried had I not gotten those two papers accepted, because at that point I would’ve been running a serious risk of a second consecutive blank year. Fortunately, it’s still a couple of years before my grant comes up for renewal, at which point I’ll find out how much NSERC Discovery Grant reviewers care about that blank year.*** But I’m trying not to worry about that blank year because it’s water under the bridge now.****
In future, I’m planning to adjust my approach to publishing a bit so as to cut the risk I ever have another blank year. Try to crank out a few more small, incremental papers targeted at less-selective venues, so that my bets are better hedged. But really, all I can do is all anyone can ever do–the best I can.
So if you have a blank year on your cv already, or you have one in future, know that you’re not alone, and that in the grand scheme of things it’s probably not that big a deal.
UPDATE: To forestall a possible misunderstanding that came up in the comments, I am not saying that you can get and keep a faculty position in ecology without being productive, or that you can slack off after you get hired or get tenure, or anything like that. All I’m saying is that, in the context of an overall track record of sufficient productivity (where “productivity” is defined more broadly and holistically than just “publication count”), one year with no publications probably isn’t a big deal.
UPDATE #2: To forestall a possible misunderstanding that came up on Twitter, which I discuss at greater length in the comments: I’d have written exactly the same post if I’d been posting unreviewed preprints of my manuscripts. Preprints and peer-reviewed papers are complements, not substitutes. Yes, posting preprints would allow the public to see preliminary versions of my as-yet-unpublished results. But to the extent that I’d worry about a gap in my publication record, it wouldn’t be because the public hasn’t yet seen a preliminary version of my as-yet-unpublished results. Nothing in this post has anything to do with preprints as far as I can see, but if you think I’m missing something on that front please do comment. Always happy to get comments disagreeing with the post.
*And you shouldn’t either.
**For instance, in contrast to someone like Brian, I’m not great at working groups. I’d publish more papers if I was better at it.
***I think and hope they won’t care much, and don’t think they should care (they should care about my record as a whole). But I don’t really know and would welcome comments on this from anyone who’s served on an NSERC Discovery Grant panel.
****Plus, through a combination of being good and being lucky, I’m a tenured full professor. Going forward, even the worst-case career scenarios for me aren’t that bad in the grand scheme of things. The worst-case scenario for me is probably that my grant doesn’t get renewed, in part because of that blank year, and gets rejected again when I reapply. At which point, it would be unlikely I’d ever be able to reapply successfully. I don’t think that scenario is very likely, and of course I’m going to do everything I can to make sure it doesn’t happen. But if it did happen, I’d probably dial back my research program, and voluntarily take on more teaching and administration so that I continued to pull my weight within my department. I wouldn’t get fired, though I’d probably cease to get raises. And none of that would affect the wonderful personal life I’m blessed with outside of work. Further, even if I were pre-tenure or still looking for my first faculty position, I wouldn’t worry too much about a blank year per se. The people who evaluate you for hiring and tenure should, and mostly do, look at your record holistically (see also).
Dynamic Ecology 
Always amazed when tenured (!) academic staff members tell the world how you can succeed with little academic outputs… What should I take as an early carerr researcher from this? That “forced” academic productivity is not a requirement to succeed?
Similar patterns at my own faculty, where newly employed research staffs usually publish way less after they have joined faculty (just heard the other day from a newly appointed lecturer that he is doing more teaching and less research in the next years as this is more rewarding (more students -> more tuition fees and money for the university ).
Of course sometimes projects simply don’t work and you have little to publish so these periods can happen. Or your manuscript is stuck in the rejection “peer-review” hell and circulates around for more than 1-2 years. Still…
Idea for a project:
Query academics with a permanent employment at the university (lecturer, reader, tenured prof ,etc…) and collect their publication rates (nr of publications per year). Maybe weight by the impact factor of the published journal.
Then plot them vs distance from the year of a first permanent position at a university. Use piece-wise regression analysis to test if there is a change in trend before and after that date.
I would be fairly suprised to see that there isn’t any change in slope. With all the oversupply of ecology PhD students in the market and the ever reducing amount of funding available, it simply stimulates ECR academics to publish more because of necessity.
And yes, I have read the previous blog posts at the end of your post, but imho the observed patterns simply reflect that people with better marketing and practice (countless numbers of applications) get more interviews as they manage to write and sound more convincing.
“Always amazed when tenured (!) academic staff members tell the world how you can succeed with little academic outputs”
Oh, I’m not saying you can succeed while producing little! (And just to forestall further possible misunderstanding, nor am I saying that it’s easy to succeed, or that you can just slack off for an extended period and not worry about the consequences.) All I’m saying is that my blank year probably isn’t a big deal *given the context of my overall track record of productivity*. It’s just as you said: “Of course sometimes projects simply don’t work and you have little to publish so these periods can happen. Or your manuscript is stuck in the rejection “peer-review” hell and circulates around for more than 1-2 years.” That’s the only point I wanted to make–that that sort of thing happens to everyone, even experienced tenured profs.
“Idea for a project:
Query academics with a permanent employment at the university (lecturer, reader, tenured prof ,etc…) and collect their publication rates (nr of publications per year). Maybe weight by the impact factor of the published journal.
Then plot them vs distance from the year of a first permanent position at a university. Use piece-wise regression analysis to test if there is a change in trend before and after that date.
I would be fairly suprised to see that there isn’t any change in slope. ”
People have done this sort of analysis, trying to identify the career stage at which scientists are most productive (see, e.g., https://dynamicecology.wordpress.com/2016/06/13/is-research-a-young-persons-game/; we also have other links to this literature in old linkfests but I can’t find them just now). One challenge is that people’s typical mix of “outputs” changes over their careers; “productivity” really isn’t a single variable. So for instance, you get different answers if you ask “at what career stage do scientists publish papers at the highest rate?” vs. “at what career stage do scientists have the highest per-paper chance of writing a paper that will go on to be highly cited?” IIRC, if you use the former as your measure of productivity (or something correlated with the former), you find that productivity typically peaks around the time of tenure, though without declining much thereafter. If you use the latter as your measure, you find that there’s no peak: people are equally productive over their entire careers. But I’m not an expert on this literature, and have only seen a few papers on this.
The other challenge in trying to identify the career stage at which scientists are most productive is that describing changes in scientists’ rate and mix of outputs as their careers progress doesn’t tell you about the causes of those changes. For instance, insofar as scientists publish less after tenure, is that because they relax and slack off? Or because their service and admin duties ramp up, leaving less time to read, think, and write?
“the observed patterns simply reflect that people with better marketing and practice (countless numbers of applications) get more interviews as they manage to write and sound more convincing.”
Hmm…so the hypothesis is that people who submit more faculty job applications in any given year get more interviews *per application submitted*? If that were so, # of interviews received would be a concave-up function of # of applications submitted. Sorry, I don’t see it; the relationship looks dead-on linear to me. That’s not to deny what’s surely the case, that people get somewhat better at putting together applications as they get some practice at it. And it could well be that some people are better than others at writing cover letters, research statements, and teaching statements tailored to the ad. But having sat on search committees myself, my experience is that who gets interviewed is not mostly a matter of “marketing” on the part of applicants.
Via Twitter: https://twitter.com/atiretoo/status/848859877837611008
This is an EXCELLENT read, particularly (I think) for ECRs! I’m also struggling with a paper right now because I’m stubborn and keep aiming high. It’s left me a bit deflated as of late, but I know that it’s a good piece of work (just not THAT good). Than ks for this, Jeremy!
Thanks!
This post made me think of sciwo’s old post on apples & research conveyor belts:
I’m curious: do you try to focus on having things at different stages on the conveyor belt? I try to (though, of course, it can be hard to tell how long it will take for a project to progress along it, or how often it will go in reverse), though sometimes this just leads to anxiety when things get bunched up (especially if that means that I have nothing in review). But it sounds like you might intentionally bunch things up a bit. At least, you mentioned intentionally focusing on other things in Fall 2015. I wonder if people differ in whether they aim to use a conveyor belt approach (and, if so, if it’s useful/productive/effective)?
I do like having a sufficient number of things on all stages of the conveyor belt at all times. Don’t take my behavior in Fall 2015 as indicating otherwise. In practice, I have been bunching things lately–submitting a bunch of mss within a short period, then going a while without submitting anything. But that’s not my preferred mode of working and I’m trying to get back to having a steadily-moving conveyor belt. Well, as steadily as the belt can ever move of course. As you say, there are lots of things out of one’s control, such as how long a new project will take, how long a journal will take to review your ms, etc.
Years ago a colleague suggested to me a rule of thumb that I thought at the time sounded right: always try to have at least two mss out for review. The idea was that, if that was always the case for you, that means that you’re doing a good job of moving things along at all stages of the conveyor belt. I’m curious what others think of that rule of thumb, and more broadly about the idea of any quantitative rule of thumb for deciding if your conveyor belt is running well. If not two papers in review at all times, how about 1, or 3, or etc.? And if not papers in review, what metric should you keep an eye on instead? Or is the whole notion of such quantitative metrics actually a bad idea–just a way to cause needless self-imposed stress or something?
When I started at Georgia Tech, I set a goal of submitted one paper from my lab per semester. I figured that, if I did that, I’d be on track for getting tenure (especially assuming I was a coauthor on papers led by collaborators). Now, I don’t focus on that as much, since I’m focused more on getting things out on timelines that make sense for lab members (and my lab is big enough that that on its own tends to keep things moving along). But I think it was useful for me to have that as my goal when I first started.
“I’m curious what others think of that rule of thumb, and more broadly about the idea of any quantitative rule of thumb for deciding if your conveyor belt is running well.”
I think any kind of quantitative rule of thumb like this is setting one up for disappointment & potential stress somewhere along the line because academic careers are so unpredictable. My blank years coincide with upheavals in my personal life: would I really have wanted the additional stress of “why haven’t I got two papers in review at the moment?!” on top of that?
In any case having X number of papers in review at any one time is no guarantee that one won’t have a blank year, if X number of papers are rejected or take a long time to review or revise or be published.
This made me think of my own strategy. I always (since my Masters) combined writing smaller papers for local journals (Brazilian Journal of Biology and the like) and a smaller number of papers for more important journals (I consider Plant Ecology and Austral Ecology as important), either as first author or collaborating with other people. The result is that I don’t have any blank years since my first paper in 2011. Another result is that I don’t really have any high-impact research for now*. I’m in my third post-doc year (it’s a five-year postdoc, we have those in Brazil). What would you say of such a strategy?
* I did get a paper accepted for Methods in Ecology and Evolution this year, as a coauthor 🙂
I think it depends a lot on your long-term goals. Are you publishing the sort of work that is done by people who hold the sort of position that you’d like to hold once your postdoc ends? I think that sort of mixed strategy is very common, I definitely don’t think you should worry that you’re not sending everything you write to Nature or whatever.
I also think that one’s publication strategy is in large part determined by the science one does. For the most part, my “target” journals are determined before I even start the research, not after I’ve obtained the results. Ok, the results do have some effect. Invariably, when they have an effect it’s to cause me to aim lower, or not publish at all, because the results were uninterpretable or very noisy. But mostly, I only bother to do projects that, if they work at all, seem likely to result in a paper in a journal at least as widely-read and selective as Ecology.
But that’s just me, your mileage may vary. Not everyone operates or should operate that way. For instance, I operate that way in part because the microcosm experiments I do generally aren’t worth publishing at all unless they’re of interest to a broad range of ecologists. There’s no community of specialists on microcosmology with its own unselective, specialized, local microcosmology journals in which I can publish microcosm papers of merely specialized or local interest. In that respect, what I do is very different from the sort of work you do. And rarely is it the case that I feel like I have a choice between publishing a bunch of little studies as separate little papers, or holding off until I have enough data to tell a really complete story that would be publishable in a high-profile journal. But again, that’s just me; there certainly are people who operate that way. My microcosm experiments tend to be pretty self-contained. Each experiment is (hopefully) one Ecology-level paper.
Makes sense, thanks! In part I was also waiting until I’m skilled enough and have enough knowledge to plan higher-quality research – I think I’m reaching this stage now – maybe a slow-science approach.
Here’s a good example of a comment that’s too long to rely to via Twitter, so I’m pulling it into the comment thread and will tweet a link to my reply.
I hesitate to reply at all, because tweets are brief and context-free, and only knowing Erol a very little bit I can’t mentally fill in a lot of the unspoken context. And in my admittedly-limited experience, trying to engage someone on Twitter to get clarification 140 characters at a time would be very difficult and inefficient. So apologies in advance if my reply results in me just talking past Erol, because the likelihood of that is unusually high here. But since Erol chose to comment via the means with which he’s most comfortable (which is totally fine by the way, I have no problem with anyone who prefers to tweet about our posts rather than comment on them), I’m going to reply via the means with which I’m most comfortable.
So, I’m very puzzled why Erol thinks I wouldn’t have had to write this post if I’d been routinely posting preprints of my work. Preprints aren’t substitutes for peer-reviewed publications. That’s for at least two reasons. First, whether Erol likes it or not (and if I had to guess, I bet he doesn’t), many of the people who might want to read my work only want to read the peer-reviewed version, because they filter the literature at least in part by focusing their reading on peer reviewed journals (usually whichever ones have a track record of publishing lots of stuff that the reader in question really wants to read). So posting preprints wouldn’t do much to get my work in front of its intended audience any faster. (aside: yes, lots of people still use “what’s published in peer-reviewed journals” as a way to filter the literature: https://dynamicecology.wordpress.com/2013/10/29/survey-results-how-do-you-find-papers-to-read-when-you-cant-do-a-search/).
Ok, I suppose I could not only post a preprint, but also tweet about it and blog about it so as to get the word out. But personally, I find that very self-promotional and I’m not comfortable doing it (others are comfortable tweeting and blogging about their own work; to each his own.) Plus, I like feeling like I’m earning my readers. If I posted a preprint and then blogged and tweeted about it to get the word out, frankly a fair number of the people who read it (or just read the abstract, or just glanced at the figures, or etc.) would only be doing so because I’m a Famous Online Ecologist. That’s not me earning my readers, that’s me coasting on whatever fame I’ve earned for stuff I did in the past. I don’t want people deciding to read my stuff because they’ve heard of me, although that’s inevitably going to happen to some extent. I want them deciding to read it because there’s a *good reason* for them to read it. Such as “a leading ecologist has decided this paper is of broad interest to ecologists, based in part on the advice of several expert reviewers who read the paper *really* carefully and critically”.
Second, the people who evaluate me and my work (my head of department, the reviewers for the funding agencies to which I apply, hiring committees if I was applying for jobs…) only care about peer-reviewed publications, not preprints. Nobody who formally evaluates me and my work will give me even a smidgen of “partial credit” for preprints that haven’t yet become peer-reviewed papers. Nor should they. For this reason alone, I’d worry about going for an extended period of time without publishing any peer reviewed papers, *even if* I was also posting preprints.
And Erol has now replied via Twitter; starts here:
To which, different strokes for different folks. Personally, posting preprints wouldn’t help me “come to terms” with a productivity gap, or cause me to see it as any less of a gap. I already know perfectly well–sometimes too well!–that I did the experiment (or whatever the research was). And sorry, but I don’t care at all that my own work isn’t visible until after it’s been accepted by a peer-reviewed journal.
Thanks for the post. I agree we shouldn’t worry too much for a blank year. They may actually be good if you have been investing your time on something that will pay off later.
But the consequences of having a blank year in your CV probably vary a lot between tenured people and those on fixed term contracts (i.e. 2-3-year contracts as most postdocs have). Compared to scientists moving on longer-term perspectives, when you have to find a new contract every two years, a blank year may be rather damaging. For example, this PNAS paper (http://www.pnas.org/content/109/14/5213) shows that ‘short-term contracts can amplify the effects of competition and uncertainty making careers more vulnerable to early termination, not necessarily due to lack of individual talent and persistence, but because of random negative production shocks’.
Of course, as you said, some hiring committees will consider other variables and form a holistic view of the candidate, but in all other cases where publication metrics are paramount, having no papers in the last year will most certainly not help anyone, don’t you think? And that may be sad if it’s encouraging people to avoid challenging questions or approaches for the sake of ensuring a continuous publication output.
You raise a good point. In general, it’s true that stochasticity matters most early in one’s career. I’m not sure what could be done about this, besides what people do already (e.g., evaluate early career researchers holistically and with appropriate attention to individual circumstances), given that the stochasticity is ineradicable.
I’m reassured that crude quantitative metrics like number of publications actually have no predictive power for who gets faculty interviews in ecology: https://dynamicecology.wordpress.com/2017/01/04/you-cant-tell-much-about-your-odds-of-getting-an-interview-for-a-faculty-position-from-common-quantitative-metrics/