Friday, March 10, 2017

Do we really need more journals?


Nature Publishing Group continues to spawn "Baby Natures" like crazy.

I was disappointed to see that Physical Review is launching a new journal Physical Review Materials. They claim it is to better serve the materials community. I found this strange. What is wrong with Physical Review B? It does a great job.
Surely, the real reason is APS wants to compete with Nature Materials [a front for mediocrity and hype] which has a big Journal Impact Factor (JIF).
On the other hand, if the new journal could put Nature Materials out of business I would be very happy. At least the journal would be run and controlled by real scientists and not-for-profit.

So I just want to rant two points I have made before.

First, the JIF is essentially meaningless, particularly when it comes to evaluating the quality of individual papers. Even if one believes citations are some sort of useful measure of impact, one should look at the distribution, not just the mean. Below the distribution is shown for Nature Chemistry.

Note how the distribution is highly skewed, being dominated by a few highly cited papers. More than 70 per cent of papers score less than the mean.

Second, the problem is that people are publishing too many papers. We need less journals not more!
Three years ago, I posted about how I think journals are actually redundant and gave a specific proposal of how to move towards a system that produces better science (more efficiently) and more accurately evaluates the quality of individuals contributions.

Getting there will obviously be difficult. However, initiatives such as SciPost and PLOS ONE, are steps in a positive direction.
Meanwhile those of us evaluating the "performance" of individuals can focus on real science and not all this nonsense beloved by many.


  1. I perfectly agree with most of your post (Isn't it motivated by this, by the way ?) but I regret that it lacks some nuance. Things are not that black and white. Not all papers in Nature journals are wrong or oversold, not all papers in Physical Review are masterpieces. It is also a bit of a caricature to put the blame on the journals only. A huge responsability is on the shoulders of those who almost prostitute themselves to publish in journals with the highest IF. The whole system certainly pushes in that direction, which is indeed a big problem, but overselling and biased presentations are also a matter of ethics of individuals. Furthermore, I must say that, over the last ten years, I have had consistently more competent reviewers at Nature journals than at PRL. I found unprofessional editors on the two sides...

  2. I understood (from talking to PRB editors) that the new PR journal was going to be aimed at materials work that currently does not end up in PRB.

    Reading the scope of PRmat that does not seem the case.

    I think they should better define the differences between PRB and PRmat before I can judge whether this is a good idea.
    ALso, it's not clear to me whether PRmat is going to be a PRX type of flashy journal or a PRB type of "proper" papers.

  3. Ross, I suspect the distribution of citations you plotted has a similar shape for all journals. I know it is the same for the ACS journal for which I am an editor (albeit with a lower median than Nature Materials!).

    One action many of us in the community can take is to be careful about how we "value" different journals. One example is in giving feedback to younger colleagues - how do we perceive the value of a paper in Nature Communications vs. Journal of Chemical Physics, for example.

  4. the problem is that people are publishing too many papers.. Yes , it results in incremental publications and not significant ones.

  5. It is popular to take pot shots at Nature publications, but I always wonder how much of this is driven by hurt egos, rather than the actual content of the journal. I think it is also important to distinguish between Nature Communications / Scientific Reports (cash-cow journals, where admittedly plenty of incorrect or badly-written stuff ends up) versus the Nature Physics / Materials, which have a pretty rigorous reviewing process (usually 3 referees), and relatively few really rotten papers. I also disagree with the comment about "real scientists"; the professional editors tend to have a pretty good overview of the field, and if you can't convince them to print your work, that's your problem.

    1. 1. The number of referees does not mean the reviewing process is rigorous;
      2. Professional editors are mostly failed scientists. What do you mean they have a pretty good overview of the field? You seriously believe that good science be trusted in their hands rather than those of active fellow workers? Editors tend to look for fads, not good science. Their overview is superficial rather than based on a deep appreciation of the field.
      3. I believe the real problem is not about the existence of whatever journals. Rather, it is about how we perceive the value of individual researchers and their science. Nowadays, scientists have sold out themselves to government and businessmen and the latter judges everything by superficial means like impact factors and citation counts. So, the deep problem rests with the scientists themselves. Removing certain journals only targets at the symptoms. Scientists must correct themselves first !

    2. 1. agreed.
      2. strongly disagreed; everyone I know agrees that PRB is a solidly good journal. The APS uses professional editors. Some of those switch to that job after significant time as an active scientist. Moreover, an editor whose job it is to have a good overview of the field may have a better overview than a scientist whose job it is to do science and thus, by definition, does this job as a secondary task.
      3. Agreed with the first two and the last three sentences.

    3. 1. Although it's true that the number of referees does not necessarily imply a rigorous review, in fact Nature and PRB will have pretty much the same list of referees to select from, so I don't see how you can describe one as "solidly good" while implying that the other is not rigorous.

      In practice, the reason that you think that PRB is a solidly good journal is that the self-selecting scientists who submit there have have the time and space to explain their results and feel no compulsion to oversell their data. But given the sheer volume of papers in PRB (including a fair number of bad / insignificant papers), I think there's a good argument for having "higher" journals for particularly significant results. If I publish a string of PRBs, and then get a notably better result which I would like to show off, it's natural to look for a higher-profile journal.

      I think that a lot of the anger is actually due to a small number of relatively high-profile groups (mostly experimentalists) who consistently publish results in e.g. Nature Communications which are either poor quality or do not stand the test of time.
      Sure, they are gaming the system which annoys us all, but we largely know who they are and I wouldn't let this stop good groups getting kudos they deserve for actually strong results by publishing at Nature.

    4. There is no use blaming experimentalists or theoreticians. The next line you have written gaming. This gaming is due to metrics. Metric should be abolished. Even if they tighten metrics the following will happen "Changes to tighten
      up metrics may happen but whatever method is chosen it is obvious that academics will learn to manipulate it as effectually as any laboratory rat would". These lines cited from the article "The h-index, the citation rating, impact factors and the aspiring researcher" Journal of Mental Health Volume 22, 2013 - Issue 6. It is available in open access.

  6. " Make Science great again" by abolishing all indexes.

  7. I'm in New Orleans for the APS meeting, and at the joint DCMP/DMP executive committee meeting tomorrow we are going to talk w the Phys Rev editors. I will specifically ask about PR Materials, its target purpose, and its relationship to PRB.

    1. Thanks Doug. Let us know what they say. The fact that this discussion will take place and the associated accountability highlights the superiority of APS journals over commercial ones.

  8. I'll probably write a separate post about this. The short version: Some materials-physicists (particularly those who report the synthesis and growth of new materials, and some quantum chemistry types who do electronic structure calculations) were apparently unhappy with their treatment by PRB, and pushed for a new venue. The claim is that this is not supposed to be like PRX or N Mat. That being said, the description of the suggested topics sure overlaps a lot with PRB, and there is definitely some concern that this will cannibalize some of PRB's possible content.