Editor’s Note: In conversation with our regular Chefs, the idea was raised for us to start offering some different formats of posts, which we’ll be trialing over time. Here, the idea is to post “grab bag” articles, consisting of short pieces on a variety of subjects. There’s a lot happening in our world, and a lot to pay attention to, but much that doesn’t rise to the needs of writing a lengthy 2000-word analysis. And so we present our first “Smorgasbord” post, a buffet offering a variety of dishes. Let us know if it works below in the comments.
A Sh***y Metaphor for Science Publishing
Ask most people what journals do, and they’ll probably reach for a metaphor. The one you’re most likely to hear is ‘gatekeeping’ – the idea that journals are positioned at the gate to an otherwise inaccessible (but desirable) area, and they get to decide who can enter. While it’s not inaccurate, the gatekeeper metaphor has a lot of uncomfortable baggage, and I would argue that this baggage may drive some of the current antipathy towards journals and peer review.
First, the process of gatekeeping instantly divides people into those that can enter and those that can’t. The ones making it into the compound must somehow be special or better than everyone else, leaving those on the outside resentful and suspicious that the rules for entry are unclear or rigged. It’s also much easier to identify with the cause of those outside the gate – they’ve been unfairly excluded by the likely corrupt and anonymous gatekeepers and the only true remedy is to tear down the gates and the walls.
So, while the gatekeeper metaphor does help people understand what journals and publishers do, its widespread use also fuels calls for the destruction of the journal system: anything that can be described as gatekeeping must be bad and should be brought low.
This video got me thinking about a different metaphor: journals as the sewage works standing between pipes full of untreated research and the pristine beaches of public discourse. The ‘sewage works’ metaphor brings home the importance of what we do, filtering out information that really should not become public, and improving the information that does make it out onto the beach. Tearing down the sewage works suddenly doesn’t seem like such a good idea.
Twitter, Musk, and Scholarly Communication
Given how frequently and enthusiastically scholars and scientists (especially in the humanities and social sciences) use Twitter to announce, promote, and argue about their work, I find myself watching the coverage of Elon Musk’s on-again/off-again plans for acquiring that platform with both fascination and trepidation. The invocation of free speech principles in connection with this issue has become something of a trigger – some opponents of Musk (mainly on the political left) characterize appeals to free speech as a cover for fascist ideology, while some of his defenders (mainly on the right) characterize any expression of concern over the impacts of unrestricted speech on Twitter as “radical leftism.” Neither of these postures strikes me as intellectually serious, which is unfortunate given the profound need for intellectually serious discussion around free-speech issues in the context of social media, particularly where scholarly communication is involved. Surely scientific and scholarly inquiry should be unconstrained by political considerations, and yet no reasonable person really believes that there should be no constraints whatsoever on speech. How, then, will we negotiate the tension between the need for free inquiry – a bedrock principle of academic scholarship – and the need for reasonable constraints on expression? Should some research topics be off-limits in and of themselves? Who will police the boundary between what’s debatable and what is not up for debate? I expect these questions to become both more acute and more fraught as the Musk/Twitter relationship continues along its current (i.e., as of 5 October) trajectory – assuming it does continue to do so, of course. One thing is clear: where Musk is concerned, the apparent trajectory today is not necessarily a solid predictor of what will happen tomorrow.
Accessible — Spirit vs. Letter?
When I speak with content providers and their tech partners about accessibility, I realize our industry often addresses compliance with accessible publishing standards from a literal perspective — which has me worried that we miss out on the deeper meanings behind those standards, designed to ensure equitable access to the scholarly record. For instance, the recent OSTP guidance (aka, the Nelson Memo) directs funders to ensure “equitable access” to literature, which calls on a number of concepts implied by the word access — access as in authorization, access as in capability, and access as in comprehension.
Much of the dialogue thus far on the Nelson memo has focused on literal aspects of the first concept: access as in authorization, or the permission to read a document. Shifting the financial models from authorized access to public access is at the heart of how the open access movement has disrupted traditional publishing businesses. So, providers are working out the most sustainable methods to comply with the letter of open access mandates.
Secondarily, I’m hearing assumptions that funder mandates will expect providers to accommodate the capabilities of readers, authors, and other stakeholders. When I discuss interpretations of the Nelson Memo with colleagues, I hear most of them assume that accessible publications are those that comply with screen readers. Here again, we expect more demand to conform with the letter of the law when it comes to meeting the baseline standards of the Web Content Accessibility Guidelines or the Americans with Disabilities Act.
It’s 2022: Are Open Access Citation Advantage Studies Still not Controlling for Selection Bias?
I know — there are hundreds of studies like this one, and I should probably just let it go, but as a former experimental biologist and editor of the first edition of a superb book on understanding experimental design and controls, it’s still maddening every time an author adds to the pile of “evidence” for a phenomenon without accounting for obvious confounding factors. I’ve written about the alleged citation advantage for open access (OA) papers before in these pages (and elsewhere), and The Scholarly Kitchen’s Phil Davis conducted one of the very few randomized controlled trials with proper controls addressing the question. Perhaps it’s just the low-hanging fruit of what seems an easy analysis or a desire by advocates to prove a point that they really, really want to be true. I don’t know the true motivation, but the hits keep coming.
I came across this one, from a 2022 conference on Science and Technology Indicators, via Twitter: “Does Open Access Really Increase Impact? A Large-Scale Randomized Analysis“. Many open access citation advantage (OACA) studies elide the question of selection bias, which was pointed out way back in 2007 as the most likely cause for the results seen here, by referring to it as a potential confounder, then ignoring it. Here, it’s not even mentioned. To make it clear once again, comparing OA articles in a journal (or cohort of journals) to non-OA articles in that same journal (or cohort) is not a fair comparison, because the OA articles may overrepresent well-funded researchers and/or authors may save their precious OA funds and only spend them on their “most important” articles. When one truly randomizes the selection of articles made OA versus those not made OA, all such advantage disappears. And of course, given that pretty much every article is freely available through sites like ResearchGate and Sci-Hub, there is no longer any real distinction between OA and subscription articles, at least in terms of free availability.
There are a few interesting findings in this study — first, caveats aside, it seems to confirm an earlier study suggesting that publishing in fully-OA journals confers a citation disadvantage over articles in subscription journals. The authors wave this finding away suggesting that it is due to fully-OA journals being “newly-created and less attractive to high level researchers”, despite ‘journal impact” allegedly being one of the factors they did control for. Also of interest is the decline seen in the alleged OACA in hybrid journals, which I would instead suspect suggests a decline in selection bias in their samples over time. OA has become increasingly commonplace, and as more and more funders mandate it and provide funds for a broad swathe of authors to pay for article processing charges (and more and more publishers turn to transformative agreements), the differences between the OA-haves and the have-nots have lessened. The biased sample is becoming less biased over time, and the “effect” seen begins to approach the non-effect seen when the sample is truly randomized.