The Journalology newsletter helps editors and publishing professionals keep up to date with scholarly publishing, and guides them on how to build influential scholarly journals.
Journalology #55: Financing mechanisms
Hello fellow journalologists, It’s been a relatively slow news week, which will probably come as a relief to many of you as we enter the final stretch of 2023. Mind you, the latest 43 page OSTP report should keep you busy for a while. NewsReport to the U.S. Congress on Financing Mechanisms for Open Access Publishing of Federally Funded ResearchCross Subsidies exist when a firm uses money from one product or source to pay the costs of a different product. These make it especially hard to estimate fixed or marginal costs accurately, even for firms with access to their own proprietary data. Cross-subsidization is common both in large for-profits and small non-profit businesses. While non-profits can make a profit, whether from conference fees, membership fees, or other revenues, they have to re-invest excess earnings to further the organization’s purpose rather than simply distributing them to shareholders. The US Office of Science and Technology Policy (report) JB: I haven’t read the report in detail yet, but this extract gives me hope that perhaps the OSTP has listened to feedback from publishers. Plan S has been pushing for transparency on costs, which is a meaningless request since shared allocations make up a decent chunk of any cost calculation. Digital Science acquires Writefull to empower researchers and publishersDigital Science has today announced it has fully acquired the AI-based academic language service Writefull, which assists users worldwide with all aspects of their scholarly writing. Writefull’s AI language models are trained on billions of sentences taken from millions of journal articles. Matched with a firm commitment to data privacy, this means its models offer unparalleled assistance to users in academic writing, paraphrasing, copy editing and revisions. Digital Science (press release) ChatGPT generates fake data set to support scientific hypothesisThe ability of AI to fabricate convincing data adds to concern among researchers and journal editors about research integrity. “It was one thing that generative AI could be used to generate texts that would not be detectable using plagiarism software, but the capacity to create fake but realistic data sets is a next level of worry,” says Elisabeth Bik, a microbiologist and independent research-integrity consultant in San Francisco, California. “It will make it very easy for any researcher or group of researchers to create fake measurements on non-existent patients, fake answers to questionnaires or to generate a large data set on animal experiments.” Nature (Miryam Naddaf) JB: This news story was pegged to the paper in JAMA Ophthalmology, which was mentioned in last week’s issue. ResearchGate and MDPI partner to boost the visibility of open access journals through Journal HomeAround 210,000 version-of-record articles from these 10 titles will be readily available on ResearchGate, including the full archive material and all new articles as they are published. These journals also benefit from enhanced brand visibility with dedicated journal profiles, prominent representation on all associated article pages and all relevant touchpoints across the ResearchGate network – keeping the journals top of mind with their reader and author audiences. All articles covered by the new partnership will automatically be added to the authors' publication records in ResearchGate. This not only reduces MDPI authors' needs for direct management but also offers them valuable insights in to the impact of their work, including data about readership and citations. ResearchGate (press release) JB: 10 of MDPI’s journals are included in this pilot, but it's not clear from the press release which journals made the cut. ResearchGate announced a similar partnership with Pensoft last week. As scientists face a flood of papers, AI developers aim to helpBut for now, scientists using AI tools need to maintain a healthy level of skepticism, says Hamed Zamani of the University of Massachusetts, Amherst, who studies interactive information-access systems. LLMs “will definitely get better. But right now, they have a lot of limitations. They provide wrong information. So scientists should be very aware of that, and double check their output.” Science (Jeffrey Brainard) There is a scientific fraud epidemic — and we are ignoring the cureImpropriety also ranges wider and deeper than blatant deception. It can be sloppy analysis or cherry-picked data. It could be the well-meaning conviction of an academic that, if only he can recruit the right patients and tailor his trials in the right way, he will crack the terrible problem of dementia. Factor in that medical regulators seem willing to revisit the cost-benefit trade-off in hard-to-treat conditions, and it becomes an environment in which the artful presentation of trial outcomes can potentially make or break a billion-dollar drug. And that, really, is the problem: the lack of due diligence means the rewards for bending or breaking the scientific rules tend to outweigh the incentives to observe them. Financial Times (Anjana Ahuja) OpinionThe persistence of eugenics in mainstream journals highlights major gaps in research integrityWould it surprise you that papers are regularly published in the academic literature that promote the idea that people in sub-Saharan Africa have cognitive ability so low that they are, on average, on the verge of intellectual impairment? Not just papers in ‘fringe’ journals, with little credibility, but in journals published by major academic publishers, Elsevier and Springer, and with influential and respected scientists on their editorial boards. This blog tells the story of the national IQ dataset. This dataset claims to provide data on average cognitive ability in nation-states, yet in reality is scientifically flawed and has been used to explicitly argue for a genetically-determined racial hierarchy of intelligence. Impact of Social Sciences (Rebecca Sear) Female researchers are less influenced by journal prestige, will it hold back their careers?There is a growing body of (economic) research on gender differences in academia. Notably, female economists tend to be held to higher standards in leading journals as measured in citations. The results are in line with recent work by Erin Hengel, who shows how papers by female economists spend longer in peer review. Based on the assumption of equal ability among men and women (on average), women (have to) devote more time to writing their papers. This, in turn, implies a lower output relative to their male colleagues and ties in with the finding that among professors in psychology, women benefit more from their papers than men. On the other hand, it may simply reflect that the work of females has to be at a higher standard. Impact of Social Sciences (W. Benedikt Schmal) Observations from the Charleston Vendor ShowcaseSubscribe to Open is on the rise. Learned societies and university presses in particular are leaning in to Subscribe to Open with the recently announced support for the model by Project MUSE, which already has 50 titles committed. There seems to be a sense that this is a no-fail model, which at some level it is, in the sense of preserving the paywall-publishing pathway. But, more than one person was surprised when I observed that this was also the year we saw S2O offerings fail to meet the sustainability thresholds. I think we are going to learn a lot about the application and sustainability of the S2O model in the coming two years, particularly as libraries face continued financial pressures of paying for open access publishing through transformative and pure publish agreements in times of declining budgets. The Scholarly Kitchen (Lisa Janicke Hinchliffe) JB: I’ve been surprised by the recent spate of S2O announcements and remain to be convinced that this is a viable business model in the long term. LLMs are not ready for editorial work (paywall)This is not to say that artificially intelligent systems such as LLMs could never achieve the standard of scientific review, but models that are based purely on language statistics will struggle to get there. As scientific publishers fight to find enough reviewers, it may be tempting to prematurely turn to artificial intelligence for editorial work. But this would be a grave mistake, as the ways that the evaluations of LLMs are wrong can sometimes be subtle. Compounding errors over years of artificial intelligence-assisted review would waste countless resources and stall scientific progress. Nature Human Behaviour (Grace W. Lindsay) Defence against the dark arts: a proposal for a new MSc courseIf I’m right, this relaxed attitude to the fraud epidemic is a disaster-in-waiting. There are a number of things that need to be done urgently. One is to change research culture so that rewards go to those whose work is characterised by openness and integrity, rather than those who get large grants and flashy publications. Another is for publishers to act far more promptly to investigate complaints of malpractice and issue retractions where appropriate. Both of these things are beginning to happen, slowly. But there is a third measure that I think should be taken as soon as possible, and that is to train a generation of researchers in fraud busting. We owe a huge debt of gratitude to the data sleuths, but the scale of the problem is such that we need the equivalent of a police force rather than a volunteer band. BishopBlog (Dorothy Bishop) Other opinion articlesKitchen Essentials: An Interview with Jennifer Gibson of Dryad Code sharing in the spotlight Living in a brave new AI era First, novel and paradigm-shifting Journal ClubGender imbalances among top-cited scientists across scientific disciplines over time through the analysis of nearly 5.8 million authorsOur evaluation of a comprehensive science bibliometric database with over 9 million authors who have published at least 5 full papers shows that there have been substantial corrections of the gender imbalance in the scientific workforce over time. However, these corrections are still lagging behind in many scientific subfields and vary extensively across countries. Moreover, while the difference between the number of male and female authors has overall become modest (about 1.3-fold across all scientific authors), the difference in the number of top-cited authors between the 2 genders remains much higher. The overall imbalance in this regard is about 2-fold in high-income countries (and also in the USA specifically) and 3-fold in other countries. PLOS Biology (John P. A. Ioannidis et al) Analyzing the impact of companies on AI research based on publicationsArtificial Intelligence (AI) is one of the most significant technologies of our time. Thus, in various contexts, it becomes vital to understand and quantitatively measure who is exerting the greatest influence on the future of AI. In pursuit of this understanding, we used scientometric data from multiple academic databases to examine and compare the citations and altmetric influence of academic and corporate AI research papers. Although the vast majority of publications are still authored by academics, we found that the citation impact of a paper is significantly higher when it is (co-)authored by a company, confirming previous studies. Similarly, papers (co-)authored by companies receive significantly more attention online, as measured by altmetrics. The robustness of our results across different methods of data selection and metrics indicates that corporate AI research has indeed become more important than purely academic AI research in recent years. Scientometrics (Michael Färber and Lazaros Tampakis) Time to rethink academic publishing: the peer reviewer crisisWe hypothesized that the length of time it takes to publish a scientific article in a microbiology research journal had been increasing, especially since the pandemic, resulting from an increase in the difficulty of finding appropriate expert peers to review the submissions. To test this hypothesis, we acquired data on the time from submission to first decision for 49,052 articles published in nine American Society for Microbiology (ASM) journals between 2016 and 2022. In addition, we acquired data over the same period that represented the number of peer reviewers that were contacted to review these articles during the assessment process. mBio (Carolina Tropini et al) And finally...In Our Time is a BBC radio series that’s been running since 1998 and is hosted by Melvyn Bragg, a national treasure. The winning formula is a recorded conversation, over 45 minutes, in which Bragg quizzes academic specialists about almost any subject of interest in human life, including history, science, philosophy, religion and the arts. The rules are that there must be no book-plugging and the guests must be active in teaching the subject. The series published episode number 1000 in September. This week, Ian Mulvany’s blog pointed me in the direction of a directory of the episodes, classified by subject area. Auditory bliss. Until next time, James |
Journalology
The Journalology newsletter helps editors and publishing professionals keep up to date with scholarly publishing, and guides them on how to build influential scholarly journals.
Hello fellow journalologists, The Journalology newsletter has been rather quiet in recent months; I had surgery at the start of September, which took some time to recover from. I’m in the final stages of migrating the newsletter to Substack, which is designed for writers rather than email marketers. This should help Journalology to reach a wider audience and will allow me to offer a paid subscription option further down the line. Substack is a social media platform and, like all such...
Subscribe to newsletter Hello fellow journalologists, The hottest topic of the moment is publishing integrity in a world being changed (for good and bad) by AI. This email follows a different format to normal. I’ve pulled together the key news stories and announcements that were published over the past month on this theme. I’ve excluded opinion pieces, otherwise this email would be much, much longer. The title and text that follow are extracts from the sources. None of the text is my own; my...
Subscribe to newsletter Hello fellow journalologists, Here’s the gist of what’s happened in scholarly publishing in the past week. The full length version of Journalology will return later this month. Thank you to our sponsor, Digital Science Digital Science is excited to launch the Dimensions Author Check API — a powerful tool that enables publishers to evaluate researchers’ publication and collaboration histories in seconds, directly within existing editorial or submission systems.Built on...