Image: Asim Saeed via flickr.com CC-BY-2.0
Another week, another rejection, right? If you’ve been in science long at all, you almost certainly have a bulging file of rejections for grants, manuscripts, fellowships, and even jobs. Here, for example, is Steve’s truly impressive job-rejection history; and here’s a previous analysis of Andrew’s manuscript rejections.
We were part of a recent Twitter exchange that began when Steve tweeted in celebration of submitting a manuscript – to its third different journal:
Well, maybe “celebration” isn’t quite the right word. Nobody likes being rejected, much less being rejected over and over again. But serial rejection happens. Not that it’s a contest, but Andrew’s record (submission of one paper to 9 journals before publication) beats Steve’s (6 journals). An admittedly unscientific Twitter poll (n=499) suggests that we’re not outliers:
Now, these worst-case data are maxima, and maxima are notoriously difficult to estimate. We have another dataset, though, because Andrew has kept meticulous records of all his manuscript rejections*. The message is similar: serial rejections are common:
Both Andrew and Steve have seen students and colleagues disheartened by having a paper rejected. It’s easy to think each rejection is an indictment of your scientific worth. It isn’t. We think it’s important for scientists to realize that rejection happens to all of us. Both Andrew and Steve have had pretty productive careers; both of us have written papers that we’re proud of. Both of us still get rejected (and we get at least temporarily disheartened too). We don’t know anyone who doesn’t.
What about the impact of those serially-rejected papers? You might imagine that serial rejection marks a weak paper. It might also be that with each rejection, the paper moves from a higher-profile journal to a lower-profile one (from Nature to Ecology Letters to the Malacological Journal of the East Lithuanian Natural History Society) – and as a result, a run of rejections dooms a paper to obscurity. We can test these hypotheses using Andrew’s rejection data, coupled with citation counts from the Web of Science:
Serially-rejected papers show no sign of lower citation impact (log citations/year vs. submissions, P = 0.08, adjusted r2 = 0.018)**. There’s a thrice-rejected paper there with over 30 citations/year – 300 in total – and there are quite a few never-rejected papers that are only lightly cited. Clearly, rejection needn’t presage low influence.
Rejection also doesn’t mean a paper is bad. Papers are rejected for lots of reasons. They may not fit the journal, or at least the editor’s concept of the journal. They may catch a bad-luck assignment of editor or reviewers. The journal may have a huge in-press backlog. A reviewer may sit down with your manuscript before their coffee, at the end of a long day, or just in a bad mood (perhaps because of a serial rejection of their own). And yes, manuscripts even get rejected because they have shortcomings – but nearly all such shortcomings are fixable for resubmission.
Which brings us (finally) to Tubthumping, the 1997 hit from the English band Chumbawamba. You may love this song or hate it, but you have to admit that Chumbawamba understands the strategy of academic publishing:
I get knocked down, but I get up again
You’re never gonna keep me down
I get knocked down, but I get up again
You’re never gonna keep me down***
(By the way, if you’re not in Tubthumping’s corner, you’re welcome to substitute Shine, from the American post-punk Rollins Band (1994):
If I’d listened everything that they said to me,
I wouldn’t be here!
And if I took the time to bleed from all the tiny little arrows shot my way,
I wouldn’t be here!
Call it persistence, call it stubbornness, call it what you will: when your work gets rejected, almost without exception**** the only thing to do is to submit again. Revise, of course, taking advantage of all the help reviewers have offered you (even reviews that seem useless at first glance almost always contain value). Both Andrew’s 9-journal paper and Steve’s 6-journal paper were overhauled along the way, with thorough rewrites and entirely new data analyses. By the time they were published, they weren’t the manuscripts that were first rejected. They were better. Would we choose serial rejection? Of course not – but we’re glad that we persisted.
The Tubthumping strategy isn’t just important with a particular rejected manuscript; it’s important in your career more generally. Don’t let the rejection of one manuscript dampen your enthusiasm for your next one:
Rejection happens to all of us, sometimes repeatedly. Some level of serial rejection might even be optimal if it’s a correlate of aiming for more prestigious or better-read journals. We believe it’s important for more senior scientists to share their rejections, and that’s what we’re doing today. If we get them – and are annoyed but not ashamed – then we hope earlier-career folk will realize that their rejections, just like ours, are normal. (Andrew even makes this a routine part of his lab meetings, with every lab member reporting every rejection.)
Finally: you may accept the inevitablity of rejection, but still prefer not to accumulate too many 6s or 9s. If so, there are some things you can do to reduce your average number of rejections. Submit good work – and not just good work, but well-polished writing that’s benefitted from friendly review by colleagues. Think carefully about the match between your manuscript and your target journal (if you’re early career, you can ask a mentor for advice). And it may not reduce the number of rejections, but it reduces their sting if you plan for them from the get-go: with a prepopulated list of journals, in order, so the next submission is expected and routine.
No matter how well you understand that rejection is normal, it never entirely stops hurting. You can’t let it get to you. So put on Tubthumping or Shine (depending on your musical sensibilities), turn the volume up loud, and when the song is over get right back on the horse.
*^We excluded papers that were unlikely to be rejected at their first journal, such as invited papers, notes on previously published papers, and the like. Including these papers would inflate the “1” bar but change nothing else.
**^Analyses of log-transformed and untransformed data, and of total citations or citations/year, all tell the same story, with all r2 < 0.04. The only significant relationship is for untransformed total citations, which do decline with submission number, but only with P = 0.036 and adjusted r2 = 0.027. There’s nothing to see here.
***^We are officially agnostic about the relevance to publishing of “He drinks a whiskey drink, he drinks a vodka drink/He drinks a lager drink, he drinks a cider drink”.
****^Although we should admit that we haven’t always persisted forever. Each of us has given up on a paper, too (expect a future post, if we can bring ourselves to write it). Our point today is that giving up should be very rare; but for a variety of reasons, it’s occasionally appropriate.