[This is the unexpected fourth part of PPPPredatory Article Counts: An Investigation. Before you read this, you should read the earlier posts—Part 1, Part 2 and Part 3—and, of course, the December 2014 Cites & Insights.]
Yes, I know, it’s hard to call it lagniappe when it’s free in any case, I did spend some time doing a first-cut version of the third bullet just above: That is, did I find clear, cogent, convincing explanations as to why publishers were questionable?
I only looked at 223 multijournal publishers responsible for 6,429 journals and “journals” (3,529 of them actual gold OA journals actually publishing articles at some point 2011-2014) from my trimmed dataset (excluding DOAJ journals and some others). I did not look at the singleton journals; that would have more than doubled the time spent on this.
Basically, I searched Scholarly Open Access for each publisher’s name and read the commentary carefully—if there was a commentary. It there was one, I gauged whether it constituted a reasonable case for considering all of that publisher’s journals sketchy at the time the commentary was written, or if it fell short of being conclusive but made a semi-plausible case. (Note the second italicized clause above: journals and publishers do change, but they’re only removed from the list after a mysterious appeals process.)
But I also looked at my own annotations for publishers—did I flag them as definitely sketchy or somewhat questionable, independently of Beall’s comments? I’m fairly tough: if a publisher doesn’t state its APCs or its policy or makes clearly-false statements or promises absurdly short peer review turnaround, those are all red flags.
For an astonishing 65% of the publishers checked there was no commentary. The only occurrences of the publishers’ names were in the lists themselves.
The reason for this is fairly clear. Beall’s blog changed platforms in January 2012, and Beall did not choose to migrate earlier posts. These publishers—which account for 41% of the journals and “journals” in my analysis and 38% of the active Gold OA journals—were presumably earlier additions to the list.
This puts the lie to the claims of some Beall fans that he clearly explains why each publisher or journal is on the list, including comments from those who might disagree. That claim is simply not true for most of the publishers I looked at, representing 38% of the active journals, 23% of the 2014 articles, and 20% of the projected 2014 revenues.
My guess is that it’s worse than this. I didn’t attempt to find individual journals, but although those journals only represent 5% of the active journals I studied, they’re extremely prolific journals, accounting for 38% of 2014 articles (and 13% of 2014 potential revenue).
If Beall was serious about his list being a legitimate tool rather than a personal hobbyhorse, of course, there would be two ongoing lists (one for publishers, one for authors) rather than an annual compilation—and each entry would have two portions: the publisher or journal name (with hyperlink), and a “Rationale” tab linking to Beall’s explanation of why the publisher or journal is there. (Those lists should be pages on the blog, not posts, and I think the latest ones are.) Adding such links, linking to posts would be relatively trivial compared to the overall effort of evaluating publishers, and it would add considerable accountability.
In another 7% of cases, I couldn’t locate the rationale but can’t be sure there isn’t one: some publishers have names composed of such generic words that I could never be quite sure whether I’d missed a post. (The search box doesn’t appear to support phrase searches.) That 7% represents 4% of active journals in the Beall survey, 4% of 2014 articles, but only 1.7% of potential 2014 revenue.
Then there are the others—cases where Beall’s rationale is available. As I read the rationales, I conclude that Beall made a sufficiently strong case for 9% of the publishers, a questionable but plausible case for 11%–and, in my opinion, no real case for 9% of the publishers.
Those figures break out to active journals, articles and revenues as follows:
Case made—definitely questionable publishers: 22% of active journals, 11% of 2014 articles, 41% of 2014 potential revenues. (That final figure is particularly interesting.)
Questionable—possibly questionable publishers: 16% of active journals, 16% of 2014 articles, 18% of 2014 potential revenues.
No case: 14% of active journals, 7% of 2014 articles, 6% of 2014 potential revenues.
If I wanted to suggest an extreme version, I could say that I was able to establish a strong case for definitely questionable publishing for fewer than 12,000 published articles in 2014—in other words, less than 3% of the activity in DOAJ-listed journals.
But that’s an extreme version and, in my opinion, dead wrong, even without noting that it doesn’t allow for any of the independent journals (which accounted for nearly 40,000 articles in 2014) being demonstrably sketchy.
Here’s what I find when I combine Beall’s rationales with my own findings when looking at publishers, ignoring independent journals:
Definitely questionable publishers: Roughly 19% of 2014 articles, or about 19,000 within the subset studied, and 44% of potential 2014 revenue, or about $11.4 million. (Note that the article count is still only about 4% of serious OA activity—but if you add in all independent journals, that could go as high as 59,000, or 12%.) Putting it another way, about 31% of articles from multijournal publishers in Beall’s list were in questionable journals.
Possibly questionable publishers: Roughly 21% of 2014 articles (34% excluding independent journals) and 21% of 2014 potential revenues.
Case not made: Roughly 22% of 2014 articles (36% excluding independent journals) and 22% of 2014 potential revenues.
It’s possible that some portion of that 22% is sketchy but in ways that I didn’t catch—but note that the combined score is the worst of Beall’s rationale or my independent observations.
I’ve said before that the worst thing about the Shen/Björk study is that it’s based on a fatally flawed foundation, a junk list of one man’s opinions—a man who, it’s increasingly clear, dislikes all open access.
My attempts to determine Beall’s cases confirmed that opinion. In far too many cases, the only available case is “trust me: I’m Jeffrey Beall and I say this is ppppredatory.” Now, of course, I’ve agreed that every journal is ppppredatory, so it’s hard to argue with that—but easy to argue with his advice to avoid all such journals, except as a call to abandon journal publishing entirely.
Which, if you look at it that way, makes Jeffrey Bell a compatriot to Björn Brembs. Well, why not? In his opposition to all Gold OA, he’s already a compatriot to Stevan Harnad: the politics of access makes strange alliances.
Otherwise, I think I’d conclude that perhaps a quarter of articles in non-DOAJ journals are from publishers that are just…not in DOAJ. The journals may be serious OA, but the publishers haven’t taken the necessary steps to validate that seriousness. They’re in a gray area.
Monitoring the Field
Maybe this also says something about the desirability of ongoing independent monitoring of the state of gold OA publishing. When it comes to DOAJ-listed journals, my approach has been “trust but verify”: I checked to make sure the journals actually did make APC policies and levels clear, for example, and that they really were gold OA journals. When it comes to Beall’s lists, my approach was “doubt but verify”: I didn’t automatically assume the worst, but I’ll admit that I started out with a somewhat jaundiced eye when looking at these publishers and journals.
I also think this exercise says something about the need for full monitoring, rather than sampling. The differences between even well-done sampling (and I believe Shen/Björk did a proper job) and full monitoring, in a field so wildly heterogeneous as scholarly journals, is just too large: about three to one, as far as I can tell.
As I’ve made clear, I’d be delighted to continue such monitoring of serious gold OA (as represented by DOAJ), but only if there’s at least a modest level of fiscal support. The door’s still open, either for hired consultation, part-time employment, direct grants or indirect support through buying my books (at this writing, sales are still in single digits) or contributing to Cites & Insights. But I won’t begin another cycle on spec: that
single-digit figure [barely two-digit figure, namely 10 copies] after two full months, with no apparent likelihood of any other support, makes it foolhardy to do so. (firstname.lastname@example.org)
As for the rest of gold OA, the gray area and the questionable publishers, this might be worth monitoring, but I’ve said above that I’m not willing to sign up for another round based on Beall’s lists, and I don’t know of any other good way to do this.