Scientific Publishing

So this week another nature publication evidencing a reduction in scientific advances over the last years, by "connecting the dots" between the latest publishing tendencies and the slow progress in science. Basically, this means that an article in the present day has much lest significance and impact than in the past. Nothing new from my point of view.

However, this is an interesting discussion point, it relates with publishing quality, scientific integrity, and especially with the survival of the fittest in science (a.k.a. publish or perish). All subjects of intense discussion between scientists, with a clear lack of consensus among them.

Starting by the end of my list, the publish or perish paradigm has been, from my understanding, the biggest driver and the most representative descriptor of the problems in Academia, which are mostly related with funding. Funding agencies are not able to find alternative ways to evaluate project proposals and individual grants to the classic number of publication combined with other (often ridiculous) criteria, such as H-index, number of supervisions, and just to name my favourite one ...time abroad (we will talk about this one later).

So what the scientists did? Known to be very clever human beings, decided to go for all the opportunities for fast (and lower-quality) publications in order solve the "immediate" funding needs, with the premise that in the future, once money was not a problem anymore, they would dedicate their life to higher impact journals and more consolidated publications. The problem with this reasoning is the following: 1) there will be always lack of money, 2) it requires a lot of time to publish high quality work, leading to an endless postponing of such studies, and 3) many underestimated that this would become a trend in the entire academia, leading to substantial impacts in our scientific credibility.

I am not even talking about the open access model of mdpi journals, which definitely should have an impacts considering the numbers of publications such publisher can do in small amount of time. I'm rather talking that today everything is publishable, and that doesn't seem right.

During my PhD I attended to a course on "how to write and publish in hydrology". The key message that was repeated 10.000 times was to avoid publishing Least Publishable Units, under the argument that more complete and multidisciplinary approached articles were good for science and for our career, thus resulting in higher quality publications. Today, I continue to believe this is the correct approach. The problem is that this strategy is hard to be applied from index-driven-scientists, under the skeptical eyes of knowledge-driven-scientists that observe an increasingly faster publishing dynamics around them.

In this particular point , is where I believe the mentors and the senior researchers, that have a greater overarching vision of a research path, should contribute the most. Especially for early-stage-researchers, by setting up the example and stimulate the discussions of great research topics. Unfortunately, it takes me very few time to dig up in Scopus several profiles of well-established scientists that either take advantage of their PhD supervisions for pumping up their cv (index wise) though fast and dirty publishing approach, but also of other many scientists that have multiple "collaborations" that results in dozens of additional articles per year (sometimes hundreds) without any significant novelty or advance in science. Here is where the ethics comes in, and in my view is lacking the most.

Unsurprisingly, the authors of the nature study even tried to make an additional analysis over a selection of manuscripts from high-impact journals, in order to leave many of the low quality problems mentioned previously, but the results were the same. As a result they even concluded that this trend is not caused by declining quality, but rather the general focus on "a narrower set of existing knowledge". This is just nonsense. A less diverse publication is also less complete, which also means lower quality, as supported by their own statement "regression models suggest that use of less diverse work, more of one’s own work and older work are all negatively associated with disruption" . But then again...perhaps would be a bit nonsense publishing in Nature and concluding that Nature paper's reduced quality over time too, wouldn't it be?