#scholarlycomms

Pablo de Castro, Open Access Advocacy Librarian at U Strathclyde

As stated in the previous post on this topic a month and a half ago, we continue to look into this issue of the possible impact of the "twitter instability" on Altmetric scores as a proxy for social impact of research.

The bad news first: the Altmetric score for the March 2023 paper in the European Journal of Human Genetics that the previous post was looking into has barely climbed despite all the media attention. As of May 6th, 2023 the score for this paper is "just" 124. This is of course one order of magnitude higher than it was a month and a half ago, but it's still rather low (see the good news bit below). The silver lining is that all the references in the news seem to have been incorporated to the score – there's 14 of them recorded in the Altmetric page.

The good news is that some publications are still able to grow very high Altmetric scores quite quickly. See for instance this PLoS Genetics paper "Imputed genomes and haplotype-based analyses of the Picts of early medieval Scotland reveal fine-scale relatedness between Iron Age, early medieval and the modern people of the UK" by A Morez (Liverpool John Moores University) et al at.

This paper, also on the topic of paleogenetics/ancient DNA and also with a key input from the University of Aberdeen (by Linus Girdland-Flink and his team), was only published on Apr 27th, 2023, i.e. just over a week ago. Its Altmetric score as of May 6th is 1,363, resulting from 166 identified mentions in news outlets and 152 tweeters. This means first and foremost that Altmetric scores seem to remain a reliable proxy for social impact of research (though the exploration continues).

It also means that the authors of this more recent paper in PLoS Genetics have probably taken a more proactive approach to dissemination within a closely-knit international community in the discipline. The journal of choice also plays an important role here – the Public Library of Science is an Open Access publisher and pioneered the adoption of article-level metrics quite a long time ago. On top of all this, the publication has been picked by AlphaGalileo, who have published and interview with the first author at Liverpool John Moores University.

Pablo de Castro, Open Access Advocacy Librarian at U Strathclyde

This is the first post in a series of informal musings on public ownership of scholarly comms-related infrastructure. This topic is very dear to the staff working in the domain at public research-performing organisations – including at U Strathclyde – and the series may subsequently see input from different members of the institutional Scholarly Publications and Research Data (SPRD) team.

Just a few days ago – on Fri Mar 17th, 2023 – news emerged in mainstream media on a ground-breaking research work conducted in Aberdeen and Edinburgh about a breast cancer-inducing gene (BRCA1) that could be traced back to some distant Orkney ancestor. This was a particularly inspiring piece of news not just because it was coming from Scotland but – mainly – because of its social implications in terms of ‘precision medicine’ application (Iceland came of course instantly to mind).

The news was very good – not the most frequent case these days unfortunately – and subsequently featured in lots of mainstream outlets beyond the usual university websites and niche research forums. One got to hear about it on BBC Radio 4 but there were pieces about it on the BBC and STV websites, The Guardian, the Aberdonian Press and Journal, the National and the Daily Mail among many other sources. Upon learning of the news – not something one should be particularly proud of given this specific genetics research work seems to have been going on for 25 years – the logical next step was to check the (Open Access) journal article in the European Journal of Human Genetics where the research findings were reported that gave rise to the hype.

The article web page contained a link to its Altmetric score – which one automatically assumed would eventually become massive same as for other high-impact research outputs in the recent past. Eventually because at the time of checking (the next day after the widespread media reporting) the score was just 15. So one kept the URL for the Altmetric page for the article in order to revisit it every week or so and witness the expected steep increase.

However, this steep increase has not materialised – or at least it has not five days after the news first broke out. This is bizarre. An Altmetric score of 17 for a paper that has featured essentially everywhere? Very strange. One could argue that it may be the news channels’ fault if they haven’t included the DOI of the article in the text or its associated metadata for their pieces, but this doesn’t look too credible an explanation – media have surely followed standard practice around this and Altmetric’s standard practice is to ‘catch’ these references regardless of whether or not the levers are there.

So the most likely explanation one is able to fathom is that the ‘Twitter crisis’ may be hitting services largely based on social media impact. It’s not just that the desertion of large swathes of very active communicators in the scholarly comms domain to Mastodon has dried up the references (tweets) that should be there for Altmetric to catch. It’s – presumably – also that such a hit to the information-gathering workflow and to its associated business model has somehow rendered the Altmetric snapshot unreliable. The impact of this paper a week after release has surely been higher than 17 (as of Mar 21st, screenshot posted at the top).

Reflections on the domino-effect impact of the recent Twitter ‘turbulences’ may or may not have been published already – one hasn’t been able to identify much in the way of analysis so far beyond the internal conversations held behind closed doors at institutions. But if there haven’t yet been any, it’s reasonable to expect a good number of them to come out sooner rather than later.

From an institutional blogger’s perspective, the main reflection on this saga is that it provides yet another piece of evidence for the convenience of relying on publicly-owned services and data sources for a sustainable management of the scholarly comms domain. It’s not that public ownership will necessarily provide the desired safeguards to prevent the unravelling of a service stack – the recent discontinuation of the funding for CORE is a case in point – but at least it warrants public institutions a certain margin for action.

Pablo de Castro, Open Access Advocacy Librarian

Five months ago we reported that a no-hybrid Open Access funding policy had been introduced at Strathclyde as of mid-Nov last year following the running out of the block grant that the Research Councils UK (currently UK Research and Innovation) had allocated to the university for 2018/19. In that post we promised "regular updates on the progress around the updated APC funding eligibility policy, ideally including a list of the funded journals since the change in the policy". This is the first one of such follow-up posts.

One caveat to keep in mind is that we're only applying a no-hybrid policy to the UKRI block grant and not to the one allocated by the Charities Open Access Fund (COAF) since this one specifically supports hybrid Open Access for the time being (though not for much longer according to the updated Wellcome Trust Open Access policy).

No-hybrid policy results

1. Lower number of funded APCs

An automatic result from the application of a no-hybrid policy is the decrease in the number of funded APCs per month. This is in fact the main reason for the implementation of such policy, since a lower number of funded APCs will mean a lower aggregate expenditure that will make the 2019/20 UKRI block grant last for the whole period it's intended to be used, i.e. until Mar 31st, 2020.

The extrapolation for the annual figure for 2019 shows we're back onto an annual average of a 100 APCs paid from the library. This is roughly where we were in 2016, but a significant underspend forced us to increase the dissemination for the available Open Access funding in 2017 and 2018, something we did so effectively that it eventually led to an overspend.

2. Much more balanced distribution across publishers

The most relevant impact of the introduction of a no-hybrid policy is by far the immediate shift in the distribution of APC funding by publisher, see the figure below. 

Not only the number of APCs paid to large 'hybrid' publishers has very significantly decreased with regard to the past two years, but the number of funded APCs with fully Open Access publishers like MDPI, Frontiers, PLoS or Copernicus has increased in parallel. 

It's not like fully Open Access titles are terribly popular among Strathclyde researchers beyond the very well-established Scientific Reports or Nature Communications, but we're happy to see our first Copernicus entry ever (not that frequent outside German-speaking countries) and to see MDPI topping the table for 2019.

A parallel table to this one charts the number of rejected Open Access funding applications for manuscripts accepted in hybrid titles and their distribution by publisher. The results are not surprising, and researchers are generally well-inclined to follow the Green OA route for these papers. We have nevertheless introduced a 1-paper-per-author-per-year exception under which they may still be able to exceptionally apply for hybrid Open Access funding for excellent papers of theirs (something for them to judge, not for us at the library, it goes without saying).

3. A parallel increase in the use of the Springer Compact

This is not something we were expecting, and in fact we still consider it to be a potential coincidence until we have more data, but it is a fact that since we introduced the no-hybrid policy, the number of publications funded under the Springer Compact has stabilised and is gradually increasing. It's been a while since we stopped trying to predict this monthly trend, as it showed a seemingly erratic behaviour. However, the number of monthly Springer Compact papers remains unusually easy to predict since Dec'18.

Whether or not this is related to the introduction of the no-hybrid policy it's hard to tell, the samples are too small to be statistically significant. If it were so, though, this would hint at a very preliminary Plan S-aligned transition whereby the APCs are transferred to Read & Publish deals. It's still hybrid titles and there's probably not that much of a correlation after all, but it's something we're keeping an eye on anyway. This is definitely the way it should happen, and we now have an ACS R&P deal to test too.

4. List of fully OA titles funded since Nov'18

Not all funded manuscripts have been published in fully OA journals, partially because of the above-mentioned exception and also because authors had occasionally checked their funding eligibility before the no-hybrid policy had been introduced and had received confirmations upon manuscript submission. The list below includes the fully OA titles we have funded in the past few months since the policy was introduced.

ACS Omega (American Chemical Society)

Aerosol and Air Quality Research (AAQR)

APL Photonics (AIP)

Applied Network Science (Springer)

Applied Sciences (MDPI)

Biology Open (Company of Biologists)

Biomedical Optics Express (OSA)

Biosensors (MDPI)

BMJ Open (BMJ)

Bone and Joint Research (British Editorial Society of Bone and Joint Surgery)

Bone Reports (Elsevier)

Cell Death Disease (NPG)

Communication Physics (NPG)

Computational and Mathematical Methods in Medicine (Hindawi)

Design Science (Cambridge)

Energies (MDPI)

Entropy (MDPI)

F1000Research

Fibers (MDPI)

Frontiers in Aging Neuroscience (Frontiers)

Frontiers in Bioengineering and Biotechnology (Frontiers)

Frontiers in Cellular and Infection Microbiology (Frontiers)

Frontiers in Immunology (Frontiers)

Frontiers in Neural Circuits (Frontiers)

Frontiers in Neuroscience (Frontiers)

Geofluids (Hindawi)

High Power Laser Science and Engineering (Cambridge)

IEEE Access (IEEE)

International Journal of Molecular Sciences (MDPI)

International Journal for Parasitology: Drugs and Drug Resistance (Elsevier)

International Journal of Naval Architecture and Ocean Engineering (Elsevier)

Journal of Biological Chemistry (American Society for Biochemistry and Molecular Biology)

Journal of Marine Science and Engineering (MDPI)

Letters in Biomathematics (Taylor & Francis)

Machines (MDPI)

Marine Drugs (MDPI)

Materials (MDPI)

Mathematical Biosciences and Engineering (AIMS Press)

mBio (American Society for Microbiology)

Micromachines (MDPI)

Molecules (MDPI)

Nature Communications (Springer Nature)

New Journal of Physics (IOP/DPG)

Nucleic Acids Research (Oxford)

Ocean Science (Copernicus)

Oncotarget (Oncotarget)

Optica (OSA)

Optics Express (OSA)

Optics Materials Express (OSA)

Pharmaceutics (MDPI)

Photonics Research (OSA)

PLoS One (PLoS)

Royal Society Open Science (The Royal Society)

Science Advances (AAAS)

Scientific Data (NPG)

Scientific Reports (Springer Nature)

Sensors (MDPI)

Solid Earth (Copernicus)

Systems Science & Control Engineering (Taylor & Francis)

Pablo de Castro, Open Access Advocacy Librarian

Sometime in mid-November, the situation that most other research-intensive universities in Scotland had already reported when we met last Sep for the Autumn 2018 Open Access Scotland working group meeting in Aberdeen also hit Strathclyde Uni: the block grant awarded to the library by the Research Councils UK (now UK Research and Innovation, UKRI) to cover Open Access publishing fees (APCs) in the period spanning from Apr 1st, 2018 to Mar 31st, 2019 ran out. This typically marks the introduction of APC funding restrictions and a subsequent change in Gold Open Access funding eligibility criteria. Open Access librarians do not enjoy this situation, as institutional researchers are notoriously difficult to be made aware of such sudden changes in policy, but there is not much alternative given the circumstances.

It's not the first time we run out of block grant funding at Strathclyde – this was also the case last year, but back then we managed to complete the funding period by overspending in the knowledge that a new tranche of funding would eventually arrive. This year however, following the release of Plan S and the updated Wellcome Trust Open Access policy, we have decided to take a different course: while we will still overspend and charge the excess expenditure to the next block grant to be awarded later this year, a tightening of the eligibility criteria in line with the above-mentioned recent pieces of policy-making will take place, very much in line with what our colleagues at fellow research-intensive Scottish institutions have been doing for some time. This means enforcing a no-hybrid Open Access funding policy.

It has occasionally been the case in the past few months while we still had some budget that researchers would come to us with a request to fund Gold Open Access for a top-of-the-class accepted manuscript whose leading institution had refused to pay an APC for, instead sticking to the Green Open Access line of action, namely depositing a copy of the accepted author manuscript – typically under a 12-month embargo – into the appropriate institutional systems. Since the manuscript (accepted in a hybrid journal) was technically eligible for Gold Open Access funding, we granted the funding request. Comments from the co-authors of the publication celebrated that "at least some universities still cared about the subsequent increase in research impact". This is an interesting remark. It again proves the fact, well-known by any institutional librarian dealing with APC funding, that every single researcher thinks that their accepted manuscript is the sole one that the library will process.

The remark also hints at the need to better coordinate institutional Open Access funding services across HEIs, something that has already been advocated in the past.

This post is not aimed however to discuss possible improvements in cross-institutional coordination around APC funding, but rather to examine the effects of the sudden introduction of the no-hybrid policy on the APC funding distribution by publishers and on researchers' attitudes. On the former, the impact of the updated eligibility criteria on the distribution of funded journal titles has been immediate. Although it's still too early to assess, early results suggest it could mark a more permanent element in our policy even beyond the arrival of the new tranche of UKRI funding on Apr 1st, 2019. This is mainly because a return to the hybrid funding policy would automatically result in running out of budget in the midst of the funding period again sometime towards the end of 2019 (if not earlier).

The APC funding by publishers resulting from the first weeks of no-hybrid policy shows, as expected, a much more balanced distribution between hybrid and fully Open Access publishers. Publishers like MDPI or PLoS have suddenly become a regular entry behind the titles that are receiving funding, while hybrid publishers like the IEEE, the Optical Society of America or Elsevier still feature on the list thanks to fully Open Access journals of theirs like IEEE Access, Optics Express or Materials & Design. SpringerNature is the sole publisher that sees almost no impact from the updated eligibility criteria, since their hybrid journals are covered by the Springer Compact agreement and – assuming the deal gets renewed in the UK as it has in the Netherlands – are hence not affected by the policy update. Critically, the APC distribution is not only more balanced across publishers, it is also significantly more affordable – and this is one of the main objectives at this stage.

On the latter aspect, i.e. the reactions from institutional researchers whose Gold Open Access funding requests are turned down due to the lack of funds, the findings thus far are that authors understand the situation and do not insist on having their paper published Gold Open Access via the library. Whether this means that they'll choose the Green Open Access route or that they will instead try to find some alternative source of funding is hard to tell at the moment, but the evidence so far tends to suggest the former choice.

One unexpected – if not unwelcome – outcome of the new policy is that it takes longer to address the funding requests than in the past, when they would just be routed into the default payment workflow for the specific publisher. Now it's often necessary to explain to the authors that the current limitations in funding availability prevent the library from accepting their funding request, and this requires a higher level of customisation in the communications than it used to be the rule.

Another potential outcome of this updated eligibility policy might be a stronger case for setting up an Institutional Open Access Fund at Strathclyde. Centralised APC funding via block grant budgets from research funders inevitably introduces a strong bias towards the more intensively UKRI- and COAF-funded departments (in our case, Electronic & Electrical Engineering, Physics and the Strathclyde Institute of Pharmacy & Biomedical Sciences, mostly to the detriment of Social Science and Humanities Schools and more generally of unfunded authors anywhere). An updated funding eligibility policy that resulted in lower expenditures could reinforce the feasibility of serving a significant fraction of unfunded authors with a limited amount of institutional funding.

Pablo de Castro, Open Access Advocacy Librarian

A number of research outputs by Strathclyde researchers ranging from the very valuable to the extraordinary are typically being released without meeting the basic standards of open publishing. This is mainly because these are unusual types of publications – i.e. not the standard journal articles, conference proceedings, books or book chapters, but rather reports or policy papers, which automatically fall under the category of ‘grey literature’. Scholarly publishers face a great deal of often justified criticism from Open Science advocates, but this specific area is one where they definitely prove to be reliable collaborators for academics. Very few publishers these days will not make emphasis on aspects like using the appropriate open licence, minting the appropriate persistent identifiers (DOIs or ISBNs) or ensuring that the impact of their publications on social media is adequately tracked. As opposite to this, and even in cases where they are made openly available, ‘internally’ published reports or books are to a certain extent condemned to ‘digital obscurity’ if they fail to keep these aspects in mind upon their online release.

For anyone aware of how relevant these publications may be for the institution and the wider society, this is very painful to watch. It also points at certain lack of digital science skills among researchers – an issue that furthermore tends to arise in very specific fields.

As with all other aspects that relate to a given scholarly culture, changing this – even gradually – represents quite a mountain to climb. Training activities on social media competence as a critical scholarly communication skill for researchers may be designed and advertised, but it will usually prove difficult to attract very busy researchers to this kind of ‘soft’ skills training. While the HEFCE policy of requiring an early deposit of every institutional research output in Pure may help identifying candidates for providing the adequate support, raising awareness of these basic requirements among the key authors and departments is extremely complex.

To climb a mountain one starts with the first step though, and it’s rewarding to see that in certain occasions the internal publishing workflows do happen to meet the appropriate standards. This has recently been the case for a report arising from the field of Speech and Language Therapy. The lead author of this publication got in contact with the Library while the final version was still being written. The purpose was to ask about Creative Commons licences that might suit the publication.

The report was subsequently released under a CC-BY licence. Moreover, a DOI or Digital Object identifier was also minted for the publication and shared with the researcher so that it would feature on the document cover for citation purposes. The full-text final version of the report was published in the Strathprints repository, which allows to track its centralised usage both via the number of downloads and its Altmetric score for impact on social media. It was not a surprise to find out that just four days after its online release, the figures for the social media impact of this humble user manual on the use of ultrasound to treat speech disorders were far higher than any other research work that their authors have ever published. The lead author herself declared to be delighted about this.