Professoriblogissa otetaan kantaa yliopistojen ja tutkimuslaitosten ajankohtaisiin asioihin.

22.11.2021

The Rule of Three

At least half of all the frustrations we experience in academic science are connected in some way to publication: editorial rejections with a bland, formulaic justification and no actionable feedback; writers' block over the nth version of a cover letter about the same manuscript, for yet another high-impact journal; manuscript submission systems that require signed CoI declarations from every single author even before editorial assessment can be initiated; PDF conversions that never complete, leading to a timeout of the entire submission process; bot-generated letters asking us to review manuscripts far outside our area of expertise (and not only for predatory journals); long delays to decision followed by a curt rejection, even though the 'comments to authors' seem neutral or even positive; pages-long peer-reviews that present huge and often mutually contradictory lists of changes required, but with no prioritization or indication of what is really needed to get the paper accepted; editors ignoring our recommendations without explanation, after we spent many hours on a peer-review; errors introduced by typesetters and graphics 'experts' that invert a crucial data figure or its meaning (or simply cannot follow the syntax of Finnish university addresses); last-minute requests for 20-word blurbs or graphical abstracts that challenge our poetic or artistic imaginations; billing that far exceeds what we thought we owed, due to not having read the small print about fees for colour illustrations in the print version that nobody ever reads, or open-access charges that we believed were covered by our institution but, in the event, only represent a small discount or none at all if we were too late in the year. 

I'm sure readers of this blog could insert many other pet gripes. But since I am an editor as well as a peer-reviewer and an author, I'll focus on just one of these that affects us all: the long, unfocused and over-inclusive lists of things to fix in a manuscript, in order to get the editor and reviewers to click 'Accept', next time around. 

The reviewers whom I engage for my current journal - and I always make sure that they are as diverse as possible - generally do a fantastic job, typically dissecting a manuscript in great detail, and often listing dozens of changes that they deem necessary or desirable. But once the reviews are in, it's my job to look over what they have to say, and decide what is really crucial, if a revision is to be invited: which is the usual outcome. In so doing I first look for common points raised by at least two of the three reviewers, then look down each list while carefully checking the manuscript myself. I then decide for myself what is really essential, what is important and what would just be nice to do. My rule of thumb is as follows: I try to distil the essentials into maximally just three clear instructions to the authors. Though I always avoid telling them what results they should get, which is a frequent cause of erroneous data making it into print. Just the controls that must be done, or the crucial follow-up experiments to make sense of what was found. In rare cases where I cannot give such plain direction, I reject the manuscript.

I think this clarity is appreciated. Almost all authors comply, even if it requires months of extra experimental work or even a complete re-write. In fact, most authors are then happy to implement as much as possible of the many other, more minor points raised by the reviewers. The result: everyone is happy and the paper is much improved.

Although I am writing here from the perspective of a journal editor, the underlying concept I am putting forward can easily be embraced by the wider scientific community. Reviewers themselves can assist, by indicating clearly which of the issues they raise are essential. But above all, authors too can pick up the baton, and make polite demands of editors to clarify decision letters which are too diffuse or list too many major changes than could possibly be implemented in a reasonable time frame. 'What must I do to get my paper published?' is a legitimate question to ask. 'Just satisfy the reviewers' is an inadequate answer, a cop-out that negates the role of academic editors entirely. Authors should ask for a maximum of three fixes that must be implemented, even if arduous. If the editor can't specify 'so few', they should simply have rejected the manuscript, not placed the author in indefinite 'remanded on bail' limbo.