Twitter turbulences and their impact on Altmetric scores
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.
So while we continue to look into this issue around Almetric scores – the score for this EJHG paper has climbed to 22 between the writing and the posting of these paragraphs – it's worth bearing in mind these scores may no longer be as reliable a proxy for research impact as they once where (a statement that many will dispute anyway).
