sociological science v. plos one
A few days ago, the Sociological Science editors released a report that discusses their journal’s performance over the last three years. I was also reading an interview with the editor of PLoS One, Joerg Heber, These two items show how these journals operate in different ways and the long term results of their editorial policy choices. Before I move on, I want to thank each journal for making their work transparent. Sociological Science and PLoS One have shown how to do scientific publishing in ways that make editorial decisions more transparent.
PLoS One: The idea is here is simple. PLoS One will only evaluate papers based on technical criteria and ethical standards. In other words, they only thing that is judged is whether the evidence in the paper actually matches the claim of the paper. No judgment is made about whether it is “high impact.” Basically, if it is competent, it gets published, assuming the authors are willing to pay the fees. Papers are blind reviewed, but authors are given many, many chances to fix flaws until either (a) the author gives up or (b) all flaws are addressed.
Long term impact? PLoS One now publishes about 20,000 papers a year. Acceptance rate? 50% in 2016, down from about 66% in earlier years. PLoS has published fewer papers than before, probably due to the rise of Science Advances (the open access branch of Science). Also, PLoS One has a decent impact factor (2.8 in 2016) given that, by design, they published a lot of marginal materials.
Sociological Science: Also a simple idea – send us a paper, they peer review fast and give you a “yes or no.” There are no revisions. Then, after you pay the publication fee, it goes open access. The result? They get 100-200 papers a year and publish about 20-25% of them. The impact factor is not reported (I may have missed it).
Perhaps the most interesting thing that I saw in the Sociological Science report was an analysis of the “most senior co-author.” They find that 47% of the top co-authors are full professors. This is insane, given that full professors, by design, a small fraction of the population of sociologists and many of them no longer publish because they are deadwood or administrators. Post-docs should be all over Sociological Science since they are desperate for jobs and have a lot of new work. This fits my impression, expressed on Facebook, that Sociological Science tilts towards research that is more established. It makes sense given the editorial model. If you are shooting for well done articles but only give “up or down” decisions with no revision, you select out for older authors and more established work.
A comparison of both journals shows that open access publishing is successful. If you want a public repository of peer reviewed work, the PLoS One is clearly a winner. Sociological Science seems to have taken the position of a well regarded specialty journal, with an emphasis on more established authors. That is good too.
Readers know that I am a “journal pluralist.” I am very happy that we have both of these publications. Three cheers for Sociological Science and three cheers for PLoS One.
50+ chapters of grad skool advice goodness: Grad Skool Rulz ($4.44 – cheap!!!!)/Theory for the Working Sociologist (discount code: ROJAS – 30% off!!)/From Black Power/Party in the Street / Read Contexts Magazine– It’s Awesome
orgtheory.net 
Is the issue with full professors publishing in Sociological Science the ones you identified, or is that people who have more on the line (like postdocs and Assistant Professors) are worried about the impacts on their career of publishing in a less well-established, open-access journal? If they’d publish data on the rank of who is submitting to them (regardless of whether they get published), we might be able to nail down that question a little better.
LikeLike
Justin
October 5, 2017 at 3:21 pm
It may also be the case that Sociological Science is leaning toward full professors in their own editorial decisions on what to publish in order to bolster their legitimacy as a new journal using a relatively novel publishing model.
LikeLike
JPD
October 5, 2017 at 4:03 pm
What proportion of research faculty in sociology are full professors? I don’t think 47% is that far off, given the greying of academia and the fact that assistant and associate are, almost by definition, time-limited states in the career. (Associates less so than assistants, but stalled associates don’t produce much research. After all, if they did, they wouldn’t be stalled.)
I tried to find info on ASA membership by rank, but gave up after about 15 minutes of fruitless searching of the web site, and IPEDs doesn’t publish “human resource” data by discipline. Maybe someone with more patience, or with access to the ASA member data base, could weigh in.
LikeLike
anon34
October 5, 2017 at 5:01 pm
Thanks for the kind words about Sociological Science, and the endorsement of the idea of a plurality of models for academic journals. I can’t help but respond whenever I see the word “insane” in close proximity to the name of our journal, so just a couple of quick thoughts.
First, along with anon34, I don’t immediately see that 47% is an insane number, or quite understand the claim that full professors are a “small fraction” of the population of sociologists. I guess it depends on how you define the population of sociologists. I did do a quick count of the Soc department at Indiana, and if I did it right Full Profs account for 46.4% of the tenure-line faculty there. More importantly, I don’t know how that number compares with other generalist sociology journals. I wouldn’t be surprised if the number is higher at Sociological Science than at ASR or AJS, for example, but I suspect it won’t be “insanely” higher. My overall point here is that it is very hard to know what the right baseline is when assessing a number like the 47% we report.
Second, I think it is worth being clear that I think you are right that to the extent that our number is higher, it is probably to a large extent (if not entirely) because of our process, and not because we are insane, or, per JPD’s comment, “leaning toward” Full Professors to establish our legitimacy. We certainly don’t lean that way as a matter of policy, and in fact work hard to protect against it, but (our) editors are (mostly) human.
For example, it seems that one could build a reasonable theoretical model of the process that looked at a) cumulative experience formulating research questions, designing studies and writing papers, and b) incentives to publish in established top tier journals no matter what, and therefore be willing to suffer through multiple rounds of revision, long waiting times etc. Post-docs and junior faculty will be high on b) but lower on a), and vice-versa for senior faculty. The first factor (a) increases acceptance rates conditional on having submitted to Sociological Science, because we explicitly do not engage in developmental reviewing. If you believe that senior faculty, on average, will have higher levels of (a), then they will have higher acceptance rates. Similarly, senior faculty are (in my experience) often burnt out on the long developmental reviewing cycle, and more likely to just want their work out there, and so may be more likely to submit to Sociological Science. In fact the suspicion that all of us had when we launched the journal was that the discipline suffers because smart people who still have something to say despite being Full Professors don’t have the patience to go through extended rounds of trying to please multiple reviewers. Contrast this with PLoS, where anyone who is willing to beat their head against the wall long enough will get published, and you will likely get a very different distribution.
I wish we had data on the flow of submissions, and how the rank composition compares to that of other journals. We have the names, of course, but it would be a lot of work to code ranks, and I don’t think we want to ask at the point of submission. (Currently we ask after acceptance about the senior-most author, so that we can determine the appropriate publication fee. We don’t ask about the rank of the other authors, although in principle I suppose we could.)
I do think a vulnerability in our model is that because we may get disproportionate submissions from senior faculty, we might “crowd out” quality work by junior faculty, post-docs etc. who do submit. Apart from vigilance on this front, which we do exercise, another protection against this bias is that we are not space constrained, since we are on online-only journal.
LikeLike
Jesper Sorensen
October 6, 2017 at 5:57 am
I should add that if you take this simple model seriously, then the very best papers by junior faculty (high on (a) should be the least likely to come to Sociological Science relative to ASR/AJS even though they would be accepted at Sociological Science, because of the career incentives. Or rather, because of the career incentives as long as ASR/AJS are reviewed as higher status the Sociological Science. (Which won’t be much longer :-)) Instead, we will only get those papers when junior faculty get a bad draw in the review process at ASR/AJS.
LikeLike
Jesper Sorensen
October 6, 2017 at 6:06 am
It’s important when interpreting these statistics to understand that Sociological Science tracks only the most senior authors (because that’s how we have chosen to structure the page fees). So, an article with three coauthors, a full professor, an assistant professor, and a graduate student would count only in the full professor category in our statistics (because that’s the most senior person on the paper). If we tallied the ranks of all authors, you would see many more junior authors.
So, how *insane* is it that we would have 47% of most senior authors being full professor?. As a benchmark, it would seem useful to compare it to the distributions in other journals. I calculated the same stats for the last year of AJS and ASR.
AJS: 61% Full, 11% Associate, 17% Assistant, 11% Post-doc or PhD student
ASR: 42% Full, 22% Associate, 25% Assistant, 11% Post-doc or PhD student
It appears that Sociological Science has roughly the same distribution of “most senior” authors as ASR. AJS has an even more senior-heavy distribution.
LikeLiked by 2 people
Olav Sorenson
October 6, 2017 at 9:16 am