Tag Archives: publishers

Press embargoes – a threat from the shadows

Uncategorized, , , , ,

Something has been rumbling under the surface in the repository world recently, at least in the UK. Over the past six months or so, the Office of Scholarly Communication has had some fraught conversations with researchers who are terrified that their papers will be ‘pulled’ from publication by the journal. The reason is because some information about the upcoming paper is publicly available.

The HEFCE policy asks us to deposit the Author’s Accepted Manuscript into a repository  “as soon after the point of acceptance as possible” and to present the manuscript “in a way that allows it to be discovered by readers and by automated tools such as search engine”.  So when a researcher deposits a paper we check the publisher copyright restrictions and deposit the work into the repository – shutting down the article, but making the record public. What this means is that the metadata about an article is publicly available before it is published.

Similarly, researchers also share their research data in the repository and make the metadata publicly available. Funders and universities require researchers to include a statement in their publications about the availability of research data supporting their findings. What this means in practice is that researchers deposit research data supporting their manuscripts into data repositories before the corresponding manuscripts are published and more and more frequently, even before these are accepted (to ensure that peer reviewers have access to all supporting materials).

Terminology

Before I go into detail about our challenge, let’s get a few terms straight here. When we talk about ‘data’ in this space we mean the information generated during a research project from which observations and conclusions are made in academic papers. The ‘metadata’ is the information about that data. In the case of an article, the metadata includes the authors, the title, the journal and the abstract. The metadata for datasets is less well defined, although there are data citation principles, so metadata in this case is information about what the data is, how it was generated, how to access and interpret it.

The second term that matters here is ’embargo’ – the word of the moment for me, given my recent participation in the Open Scholarship Initiative Embargo Workgroup.  Our resolution was that we need to fund a research piece to resolve the reasoning behind embargoes (bearing in mind there is no link between half life of article usage and subscription rates) and what is a reasonable period of time for them.

Some publishers impose publication embargoes on the release of the Author’s Accepted Manuscript for a period of time after publication. The business of managing embargoes falls into the laps of repository managers. Indeed I have written before about how the complexity of different publisher agreements means it is almost impossible for an individual researcher to navigate the rules. We manage embargoes by putting the article under indefinite embargo when we deposit it and then check back on a rotational basis to see if the work has been published. Once it has, we can set the embargo date. And this is time consuming – to give an idea of scale we currently have over 1700 papers in the ‘checking’ pile.

There is a second type of embargo – a press embargo. This is an embargo which prohibits anyone actively discussing the content of accepted papers with the media prior to publication. The exception is that a few days before publication the journal allows journalists access to the published papers so they can prepare news stories to coincide with the publication of the work. It is this second type of embargo that is causing confusion with the research community.

The perceived problem

Our researchers are concerned that having the metadata about an article available means that publishers will consider this a breach of embargo and will pull the publication. Note that the Author’s Accepted Manuscript of the article itself (or the data files, in case of datasets) is locked down and the information about the volume, issue and pages are missing as the work is not yet published.

The researchers are worried because there is a need for publication in high profile journals such as Nature for their careers and if a work was to be pulled from publication this would have huge implications for them. This has caused a challenge for us – clearly we do not wish to threaten our researchers’ publication prospects, but we are also bound by the requirements of the HEFCE policy.

In November last year I put a query out on a couple of mail lists about this which generated a great deal of discussion. As it happens some UK repository managers are locking down the metadata about ALL articles until publication.

The actual problem

So we have decided to go to the source and are now in discussion with various publishers. A few things have come to light. First, generally publishers understand the distinction between these embargoes (details about responses are below). However there appears to be confusion at the editor level about this – and our researchers are in contact with editors.

The second issue is the level of bullying that researchers appear to be subjected to by the publishing industry. They are petrified of doing the ‘wrong thing’, and that they will be punished by having their article pulled. There is of course a question about whether any articles ever have actually been pulled because of metadata being available prior to publication or if this is yet another ghost we are boxing. No researchers to date have given us a concrete example of this happening.

Below is some of the correspondence we have had with publishers to date on this issue. The responses have been varied – from helpful and encouraging to restrictive and uninformative.

No issue – BMJ

The University of York had some information back from BMJ, which noted that the self archiving policy says that the post-print – “Final draft of manuscript: post peer-review, before the article is copyedited, typeset and published” – can be made available without embargo.  They did note that the press (not the author) is subject to the rule “All material accepted for publication in any BMJ journal is under embargo until it is published online.”

Permission and concern – Nature Publishing Group

Late last year I contacted Nature Publishing Group (NPG) to ask their position on this issue and they were adamant that they do not pull papers from publication if the metadata is available prior to publication. They did: “ask that our two embargoes – self-archiving embargo and press embargo – are respected”. They clarified that “NPG deposition to all PMC repositories allows the deposits to be fully discoverable as soon as processed by the repository, and the manuscripts’ full text become accessible 6 months after publication. In practice, this also means metadata can be available upon acceptance“.

NPG also looked at the question of articles being “pulled” ahead of publication and expressed concern that this idea was being propagated. They said that “to our knowledge there are no cases in which this has happened by putting metadata in a university repository. If you have information about any cases in which this is claimed to be the case, we would be very grateful if they could be sent through to us so that we can investigate them further.”

Poor practice – Science

We recently had a Science paper where we had already deposited the data associated with that paper into our repository, (we had shut the data down for release on publication date) and had generated a link that the researchers were able to include in their paper. However the issue of having information in the public domain prior to publication was raised by our researchers so we wrote to Science for clarification. Their response was:

“I discussed your question with my editorial colleagues, and our provisional response is that we would prefer you to take the cautious option of keeping the metadata in the ‘dark archive’ until the date of publication. We appreciate, though, that this is non-ideal from your point of view, and will now be discussing the question with our Office of Public Programs in Washington, to see whether we can accommodate your preferred procedure.”

Remember that this is the publisher telling us to suppress the metadata about data that is published by us in our repository.

The big problem we have (apart from the principle of this issue) is that while we can automate the turning off of an embargo in the repository we cannot automate the movement of an item from a dark archive to an open one – this must be done manually. The paper was published on Good Friday. There was no-one physically at work (indeed the Library was closed) until the following Tuesday. So for the first four days of this article being in the public domain there was a dead link to data in it. This is not just ‘non-ideal’ – it is contrary to the idea of effective and complete publication.

Communication breakdown – The Lancet

An example of the problem we are encountering  happened only last week. We were contacted by a researcher who demanded we take down the details of their accepted paper that was to be published in The Lancet because an assistant editor had taken issue with us posting the metadata prior to publication. Again we contacted the publisher and asked them their position.

When we spoke to The Lancet they were helpful and positive, reassuring us they were “happy to permit the release of article metadata at this stage”. They also said they had been in contact with the assistant editor of this article who has directly contacted the researcher. The explanation was “it looks like there was a misunderstanding regarding the wording of one of our policies”.

Right back at ya – Elsevier

Our correspondence with Elsevier was particularly unhelpful. While acknowledging that press embargoes are “editorial policy rather than open access policy” and that they are designed to create press interest, they concluded that this “isn’t a matter for a corporate policy at company level”.

This means it is apparently up to us to find out individual journal positions rather than the organisation taking responsibility for what is becoming a major problem for us. We contacted Cell Press (a subsidiary of Elsevier). They replied:

“You’ll notice that our policy is that we discourage release of the metadata and abstract prior to official publication. The reason is that this would be considered breaking any press embargoes on the article. So while, like Nature, we would not prohibit release of metadata, it means that it would be unlikely that the article would be considered for a press release. If you think that your article is likely to be appropriate for a press release to top-tier media outlets, then we recommend delaying deposit to the institutional repository until the article is officially published.

So not only do Elsevier not give us an answer, and we need to contacts journals directly, we now have to make editorial decisions on individual Cell Press articles to determine if these might be potentially worthy of a press release. Now, as it happens I have a science journalism background* and could possibly do this – but are we seriously required in the UK to employ repository managers with news reporting skills so they can concurrently meet HEFCE requirements and also ensure that the public profile of their institution’s research is protected? (*This means, ironically, that I have many years of direct experience with press embargoes.)

There are two problems here. There is the practical embargo issue repository managers face, and there is the bullying of researchers problem.

Protecting researchers

In the case with The Lancet, despite us explaining that this was a considerably larger problem for us than this particular example, the researcher was unwilling to give us any further details as to the communication they had received from the journal other than it had been an ‘assistant editor’. This ‘protection’ of journals by researchers is not uncommon in our experience. indeed, a recent article noted the “oppression” by editors stating for example that researchers are “afraid to say anything about the New England Journal because they’re afraid they won’t get something published there”.

Bear in mind that Nature specifically stated that they were unaware of any examples of a paper being pulled because the metadata was available prior to publication. We have not been able to uncover a single concrete example of this actually occurring. Yet we continue to have distressed and frightened researchers contacting us because of this threat.

I don’t have an answer to this bullying problem other than our need to move away from having publication in a high impact journal as the be-all and end-all for research careers. In case you are in any doubt of how destructive this situation is, consider ‘Your Right Arm for a Publication in AER?‘ where researchers indicated they would be prepared to lose half a thumb for a high profile publication. But reconfiguring the reward and recognition processes in research is a long way off. Meanwhile the least we can do is acknowledge this is happening and bring it into the open.

Clarity on press embargoes

In relation to the embargo issue, while we can continue to contact each publisher in turn and try and negotiate these issues with them, this is hardly time efficient given it is a sector-wide problem. The ‘solution’ from some (well resourced and staffed) publishers is that it is up to individual repository managers  to contact each journal in turn for clarification. This is clearly not possible given that our staff time is already spent checking and complying with the myriad of different publisher requirements in place.

A different solution could be that publishers shorten the time between acceptance and publication. When journals have a quick turnover time, the period when an article’s metadata is publicly available before publication is limited to a matter of days. It is different when there is a delay of months or even years between acceptance and publication.

I should note that our team debated about whether writing this piece would actually do more harm than good – triggering publishers to suddenly introduce further restrictions on metadata. Managing repositories and juggling embargoes and funder policies is already complicated enough. Material available in institutional repositories is only a small percentage of material that is publicly available – according to the Monitoring the Transition to Open Access report, institutional repositories in the UK hold 7.9% of Author Accepted Manuscripts, compared to 56.6% in subject repositories and 19.7% in social sharing sites. Yet we are generally the ones actually adhering to the complex sets of embargo rules.

We consider this to be a sector problem – and one that should be addressed by sector-wide resources, such as the funding bodies. I should note Jisc have expressed interest in supporting us with this issue. The increased crack-down on our activities is frustrating, unhelpful and feeds the beast that is the publishing industry stranglehold on researcher behaviour.

Published 20 May 2016
Written by Dr Danny Kingsley
Creative Commons License

Watch this space – the first OSI workshop

Uncategorized, , , , ,

It was always an ambitious project – trying to gather 250 high level delegates from all aspects of the scholarly communication process with the goal of better communication and idea sharing between sectors of the ecosystem. The first meeting of the Open Scholarship Initiative (OSI) happened in Fairfax, Virginia last week. Kudos to the National Science Communication Institute for managing the astonishing logistics of an exercise like this – and basically pulling it off.

This was billed as a ‘meeting between global, high-level stakeholders in research’ with a goal to ‘lay the groundwork for creating a global collaborative framework to manage the future of scholarly publishing and everything these practices impact’. The OSI is being supported by UNESCO who have committed to the full 10 year life of the project. As things currently stand, the plan is to repeat the meeting annually for a decade.

Structure of the event

The process began in July last year with emailed invitations from Glenn Hampson, the project director. For those who accepted the invitation, a series of emails from Glenn started with tutorials attached to try and ensure the delegates were prepared and up to speed. The emails gathered momentum with online discussions between participants. Indeed much was made of the (many) hundreds of emails the event had generated.

The overall areas the Open Scholarship Initiative hopes to cover include research funding policies, interdisciplinary collaboration efforts, library budgets, tenure evaluation criteria, global institutional repository efforts, open access plans, peer review practices, postdoc workload, public policy formulation, global research access and participation, information visibility, and others. Before arriving delegates had chosen their workgroup topic from the following list:

  • Embargos
  • Evolving open solutions (1)
  • Evolving open solutions (2)
  • Information overload & underload
  • Open impacts
  • Peer review
  • Usage dimensions of open
  • What is publishing? (1)
  • What is publishing? (2)
  • Impact factors
  • Moral dimensions of open
  • Participation in the current system
  • Repositories & preservation
  • What is open?
  • Who decides?

The 190+ delegates from 180+ institutions, 11 countries and 15 stakeholder groups gathered together at George Mason University (GMU), and after preliminary introductions and welcomes the work began immediately with everyone splitting into their workgroups. We spent the first day and a half working through our topics and preparing a short presentation for feedback on the second afternoon. There was then another working session to finalise the presentations before the live-streamed final presentations on the Friday morning. These presentations are all available in Figshare (thanks to Micah Vandegrift).

The event is trying to address some heady and complex questions and it was clear from the first set of presentations that in some instances it had been difficult to come to a consensus, let alone a plan for action. My group had the relative luxury of a topic that is fairly well defined – embargoes. It might be useful for the next event to focus on specific topics and move from the esoteric to the practical.

In addition the meeting had a team of ‘at large’ people who floated between groups to try and identify themes. Unsurprisingly, the ‘Primacy of Promotion and Tenure’ was a recurring theme throughout many of the presentations. It has been clear for some time that until we can achieve some reform of the promotion and tenure process, many of the ideas and innovations in scholarly communication won’t take hold. I would suggest that the different aspects of the reward/incentive system would be a rich vein to mine at OSI2017.

Closed versus open

In terms of outcomes there was some disquiet beforehand, by people who were not attending, about the workshop effectively being ‘closed’. This was because there was a Chatham House Rule for the workgroups to allow people to speak freely about their own experiences.

There was also some disquiet by those people who were attending about a request that the workgroups remain device-free. This was to try and discourage people checking emails and not participating. However people revert to type – in our group we all used our devices to collaborate on our documents. In the end we didn’t have much of a choice, the incredibly high tech room we were using in the modern GMU library flummoxed us and we were unable to get the projector to work.

That all said, there is every intention to disseminate the findings of the workshops widely and openly. During the feedback and presentations sessions there was considerable Twitter discussion at #OSI2016 – there is a downloadable list of all tweets in figshare – note there were enough to make the conference trend on Twitter at one point. This networked graphic shows the interrelationships across Twitter (thanks to Micah and his colleague). In addition there will be a report published by George Mason University Press incorporating the summary reports from each of the groups.

Team Embargo

Our workgroup, like all of them, represented a wide mix of interest groups. We were:

  • Ann Riley – President, Association of College and Research Libraries
  • Audrey McCulloch, Chief Executive, Association of Learned and Professional Societies
  • Danny Kingsley – Head of Scholarly Communication, Cambridge University
  • Eric Massant, Senior Director of Government and Industry Affairs, RELX Group
  • Gail McMillan, Director of Scholarly Communication, Virginia Tech
  • Glenorchy Campbell, Managing Director, British Medical Journal North America
  • Gregg Gordon, President, Social Science Research Network
  • Keith Webster, Dean of Libraries, Carnegie Mellon University
  • Laura Helmuth, incoming president, National Association of Science Writers
  • Tony Peatfield, Director of Corporate Affairs, Medical Research Council, Research Councils, UK
  • Will Schweitzer, Director of Product Development, AAAS/Science

It might be worth noting here that our workgroup was naughty and did not agree beforehand on who would facilitate, so therefore no-one had attended the facilitation pre-workshop webinar. This meant our group was gloriously facilitator and post-it note free – we just got on with it.

Banishing ghosts

We began with some definitions about what embargoes are, noting that press embargoes, publication embargoes and what we called ‘security’ embargoes (like classified documents) all serve different purposes.

Embargoes are not ‘all bad’. In the instance of press embargoes they allow journalists early access to the publication in order for them to be able to investigate and write/present informed pieces in the media. This benefits society because it allows for stronger press coverage. In terms of security embargoes they protect information that is not meant to be in the public domain. However embargoes on Author’s Accepted Manuscripts in repositories are more contentious, with qualified acceptance that these are a transitional mechanism in a shift to full open access.

The causal link of green open access resulting in subscription loss is not yet proven. The September 2013 UK Business, Innovation and Skills Committee Fifth Report: Open Access stated “There is no available evidence base to indicate that short or even zero embargoes cause cancellation of subscriptions”. In 2012 the Committee for Economic Development Digital Connections Council in The Future of Taxpayer-Funded Research: Who Will Control Access to the Results? concluded that “No persuasive evidence exists that greater public access as provided by the NIH policy has substantially harmed subscription-supported STM publishers over the last four years or threatens the sustainability of their journals”.

However there is no argument that traffic on websites for journals that rely on advertising dollars (such as medical journals) suffer when the attention is pulled to another place. This clearly potentially affects advertising revenue which in turn can impact on the financial model of those publication.

During our discussions about the differences between press embargoes and publication embargoes I mentioned some recent experiences in Cambridge. The HEFCE Open Access Policy requires us to collect Author’s Accepted Manuscripts at the time of acceptance and make the metadata about them available, ideally before publication. We respect publishers’ embargoes and keep the document itself locked down until these have passed post-publication. However we have been managing calls from sometimes distressed members of our research community who are worried that making the metadata available prior to publication will result in the paper being ‘pulled’ by the journal. Whether this has ever actually happened I do not know – and indeed would be happy to hear from anyone who has a concrete example so we can start managing reality instead of rumour. The problem in these instances is the researchers are confusing the press embargo with the publication embargo.

And that is what this whole embargo discussion comes down to. Much of the discourse and arguments about embargoes are not evidence based. There is precious little evidence to support the tenet that sits behind embargoes – which is that if publishers allow researchers to make copies of their work available open access then they will lose subscriptions. The lack of evidence does not prevent the possibility it is true however – and that is why we need to settle the situation once and for all. If there is a sustainability issue for journals because of wider green open access then we need to put some longer term management in place and work towards full open access.

It is possible the problem is not repositories, institutional or subject-based. Many authors are making the final version of their published work available in contravention of their Copyright Transfer Agreement in ResearchGate or Academia.edu. It might be that this availability of work is having an impact on researcher’s usage of work on the publishers’ sites. Given that in institutional repositories repository managers make huge efforts to comply with complicated embargoes it is quite possible that repositories are not the problem. Indeed, only a small proportion of work is made available through repositories according to the August 2015 Monitoring the Transition to Open Access report (look at ‘Figure 9. Location of online postings (including illicit postings)’ on page 38).  If this is the case, requiring institutions to embargo the Author’s Accepted Manuscripts they hold in their repositories for long periods will not make any difference. They are not the solution.

Our conclusion from our preliminary discussions was that there needs to be some concrete, rigorous research into the rationale behind embargoes to inform publishers, researchers and funders.

Our proposal – research questions

In response to this the Embargo workgroup decided that the most effective solution was to collaborate on an agreed research process that will have the buy-in of all stakeholders. The overarching question that we want to try and answer is ‘What are the impacts of embargoes on scholarly communication?’ with the goal to create an evidence base for informed discussion on embargoes .

In order to answer that question we have broken the big issue into a series of smaller questions:

  • How are embargoes determined?
  • How do researchers/students find research articles?
  • Who needs access?
  • Impact of embargoes on researchers/students?
  • Effect of embargoes on other stakeholders?

We decided that if the research found there was a case for publication embargoes then agreement on the metrics that should be used to determine the length of an embargo would be helpful. We are hoping that this research will allow standards to be introduced in the area of embargoes.

Discoverability and the issue of searching behaviour is extremely relevant in this space. Our hypothesis is if people are following publishers’ journal pages to find material then the fact that some of the same information is disbursed amongst lots of repositories means that the publisher arguments that embargoes threaten their finances are weakened. However if people are primarily using centralised search engines such as Google Scholar (which favours open versions of articles over paid ones) then that strengthens the publisher argument that they need embargoes to protect revenue.

The other question is whether access really is an issue for researchers. The March 2015 STM Report looked at the research in this area which indicate that well over 90% of researchers surveyed in separate studies said research papers were easy or fairly easy to access which appears to suggests on the face of it little problem in the way of access (look for the ‘Researchers’ access to journals’ section starting p83). Rather than repeating these surveys indicators for how much embargoes restrict access to researchers could include:

  • The usage of Request a Copy buttons in repositories
  • The number of ‘turn-aways’ from publishers platforms
  • The take-up level of Pay Per View options on publisher sites
  • The level of usage of ‘Get it Now’ – where the library obtains a copy through interlibrary loan or document delivery and absorbs the cost.

Our proposal – Research structure

The project will begin with a Literature Review and an investigation into the feasibility of running some Case Studies.

Two clear Case Studies could provide direct evidence if the publishers were willing to share what they have learned. In both cases, there has been a move from an embargo period for green OA to removing embargoes completely. In the first instance, Taylor and Francis began a trial in 2011 to allow immediate green OA for their library and information science journals, meaning that authors published in 35 library and information science journals have the right to deposit their Accepted Manuscript into their institutional repository and make it immediately available. Authors who choose to publish in these journals are no longer asked to assign copyright. They now sign a license to publish, which allows Taylor & Francis to publish the Version of Record. Additionally, authors can choose to make their work green open access with no embargoes applied. In 2014 the pilot was extended for ‘at least a further year’.

As part of the pilot, Taylor and Francis say a survey was conducted by Routledge to canvas opinions on the Library & Information Science Author Rights initiative and also investigated author and researcher behaviour and views on author rights policies, embargoes and posting work to repositories. The survey elicited over 500 responses, including: “Having the option to upload their work to a repository directly after publication is very important to these authors: more than 2/3 of respondents rated the ability to upload their work to repositories at 8, 9, or 10 out of 10, with the vast majority saying they feel strongly that authors should have this right”. There are no links to this survey that I have been able to uncover. It would be useful to include this survey in the Literature Review and possibly build on it for other stakeholders.

The second Case Study is Sage that, in 2013, decided to move to an immediate green policy. Both examples would have enough data by now to indicate if these decisions have resulted in subscription cancellations. I have proposed this type of study before, to no end. Hopefully we might now have more traction.

The Literature Review and Case Studies will then inform the development of a Survey of different stakeholders – which may have to be slightly altered depending on the audience being surveyed.  This is an ambitious goal – because the intention is to have at least preliminary findings available for discussion at the next OSI in 2017.

There was some lively Twitter discussion in the room about our proposal to do the study. Some were saying that the issue is resolved. I would argue that anyone who is negotiating the embargo landscape at the moment (such as repository managers) would strongly disagree with the position. Others referred to research already done in this space, for example the Publishing and Ecology of European Research (PEER) project. This study does discuss embargoes but approached the question with a position that embargoes are valid. The study we are proposing is asking specifically if there is any evidence base for embargoes.

Next steps

We will be preparing a project brief and our report for the OSI publication over the next couple of weeks.

The biggest issue for the project will be for us to gather funding. We have done a preliminary assessment of the time required to do the work so we could work out a ballpark figure for the fundraising goal. Note that our estimation of the number of workdays required for the project was deemed as ‘ludicrously low’ by a consultant in discussion later.

It was noted by a funder in casual discussions that because publishers have a vested interest in embargoes they should fund research that investigates their validity. Indeed Elsevier have already offered to assist financially for which we are grateful, but for this work to be considered robust and for it to be widely accepted it will need to be funded from a variety of sources. To that end we intend to ‘crowd fund’ the research in batches of $5000. The number of those batches will depend on the level of our underestimation of the time required to undertake the work (!).

In terms of governance, Team Embargo (perhaps we might need a better name…) will be working together as the steering committee to develop the brief, organise funding and choose the research team to do the work. We will need to engage an independent researcher or research group to ensure impartiality.

Wrap up summary of the workshop

There were a few issues relating to the organisation of the workshop. Much was made of the many hundreds of emails that were sent both from the organising group and also amongst the delegates before-hand. This level of preliminary discussion was beneficial but using another tool might help. It was noted that the level of email was potentially the reason why some of the delegates who were invited did not attend.

There was a logistic issue in having 190+ delegates staying in a hotel situated in the middle of a set of highways that was a 30 minute bus ride away from the conference location at George Mason University (also situated in an isolated location). The solution was a series of buses to ferry us each way each day, and to and from the airport. We ate breakfast, lunch and dinner together at the workshop location. This combined with the lack of alcohol because we were at an undergraduate American campus (where the legal drinking age is 21) gave the experience something of a school camp feel. Coming from another planned capital city (Canberra, Australia) I am sure that Washington is a beautiful and interesting place. This was not the visit to find that out.

These minor gripes aside, as is often the case, the opportunity to meet people face to face was fantastic. Because there was a heavy American flavour to the attendees, I have now met in person many of the people I ‘know’ well through virtual exchanges. It was also a very good process to work directly with a group of experienced and knowledgeable people who all contributed to a tangible outcome.

OSI is an ambitious project, with plans for annual meetings over the next decade. It will be interesting to see if we really can achieve change.

Published 24 April 2016
Written by Dr Danny Kingsley
Creative Commons License

‘It is all a bit of a mess’ – observations from Researcher to Reader conference

Uncategorized, , , , , , , , ,

“It is all a bit of a mess. It used to be simple. Now it is complicated.” This was the conclusion of Mark Carden, the coordinator of the Researcher to Reader conference after two days of discussion, debate and workshops about scholarly publication..

The conference bills itself as: ‘The premier forum for discussion of the international scholarly content supply chain – bringing knowledge from the Researcher to the Reader.’ It was unusual because it mixed ‘tribes’ who usually go to separate conferences. Publishers made up 47% of the group, Libraries were next with 17%, Technology 14%, Distributors were 9% and there were a small number of academics and others.

In addition to talks and panel discussions there were workshop groups that used the format of smaller groups that met three times and were asked to come up with proposals. In order to keep this blog to a manageable length it does not include the discussions from the workshops.

The talks were filmed and will be available. There was also a very active Twitter discussion at #R2RConf.  This blog is my attempt to summarise the points that emerged from the conference.

Suggestions, ideas and salient points that came up

  • Journals are dead – the publishing future is the platform
  • Journals are not dead – but we don’t need issues any more as they are entirely redundant in an online environment
  • Publishing in a journal benefits the author not the reader
  • Dissemination is no longer the value added offered by publishers. Anyone can have a blog. The value-add is branding
  • The drivers for choosing research areas are what has been recently published, not what is needed by society
  • All research is generated from what was published the year before – and we can prove it
  • Why don’t we disaggregate the APC model and charge for sections of the service separately?
  • You need to provide good service to the free users if you want to build a premium product
  • The most valuable commodity as an editor is your reviewer time
  • Peer review is inconsistent and systematically biased.
  • The greater the novelty of the work the greater likelihood it is to have a negative review
  • Poor academic writing is rewarded

Life After the Death of Science Journals – How the article is the future of scholarly communication

Vitek Tracz, the Chairman of the Science Navigation Group which produces the F1000Research series of publishing platforms was the keynote speaker. He argued that we are coming to the end of journals. One of the issues with journals is that the essence of journals is selection. The referee system is secret – the editors won’t usually tell the author who the referee is because the referee is working for the editor not the author. The main task of peer review is to accept or reject the work – there may be some idea to improve the paper. But that decision is not taken by the referees, but by the editor who has the Impact Factor to consider.

This system allows for information to be published that should not be published – eventually all publications will find somewhere to publish. Even in high level journals many papers cannot be replicated. A survey by PubMed found there was no correlation between impact factor and likelihood of an abstract being looked at on PubMed.

Readers can now get papers they want by themselves and create their own collections that interest them. But authors need journals because IF is so deeply embedded. Placement in a prestigious journal doesn’t increase readership, but it does increase likelihood of getting tenure. So authors need journals, readers don’t.

Vitek noted F1000Research “are not publishers – because we do not own any titles and don’t want to”. Instead they offer tools and services. It is not publishing in the traditional sense because there is no decision to publish or not publish something – that process is completely driven by authors. He predicted this will be the future of science publishing will shift from journals to services (there will be more tools & publishing directly on funder platforms).

In response to a question about impact factor and author motivation change, Vitek said “the only way of stopping impact factors as a thing is to bring the end of journals”. This aligns with the conclusions in a paper I co-authored some years ago. ‘The publishing imperative: the pervasive influence of publication metrics’

Author Behaviours

Vicky Williams, the CEO of research communications company Research Media discussed “Maximising the visibility and impact of research” and talked abut the need to translate complex ideas in research into understandable language.

She noted that the public does want to engage with research. A large percentage of public want to know about research while it is happening. However they see communication about research is poor. There is low trust in science journalism.

Vicki noted the different funding drivers – now funding is very heavily distributed. Research institutions have to look at alternative funding options. Now we have students as consumers – they are mobile and create demand. Traditional content formats are being challenged.

As a result institutions are needing to compete for talent. They need to build relationships with industry – and promotion is a way of achieving that. Most universities have a strong emphasis on outreach and engagement.

This means we need a different language, different tone and a different medium. However academic outputs are written for other academics. Most research is impenetrable for other audiences. This has long been a bugbear of mine (see ‘Express yourself scientists, speaking plainly isn’t beneath you’).

Vicki outlined some steps to showcase research – having a communications plan, network with colleagues, create a lay summary, use visual aids, engage. She argued that this acts as a research CV.

Rick Anderson, the Associate Dean of the University of Utah talked about the Deeply Weird Ecosystem of publishing. Rick noted that publication is deeply weird, with many different players – authors (send papers out), publishers (send out publications), readers (demand subscriptions), libraries (subscribe or cancel). All players send signals out into the school communications ecosystem, when we send signals out we get partial and distorted signals back.

An example is that publishers set prices without knowing the value of the content. The content they control is unique – there are no substitutable products.

He also noted there is a growing provenance of funding with strings. Now funders are imposing conditions on how you want to publish it not just the narrative of the research but the underlying data. In addition the institution you work for might have rules about how to publish in particular ways.

Rick urged authors answer the question ‘what is my main reason for publishing’ – not for writing. In reality it is primarily to have high impact publishing. By choosing to publish in a particular journal an author is casting a vote for their future. ‘Who has power over my future – do they care about where I publish? I should take notice of that’. He said that ‘If publish with Elsevier I turn control over to them, publishing in PLOS turns control over to the world’.

Rick mentioned some journal selection tools. JANE is a system (oriented to biological sciences) where authors can plug in abstract to a search box and it analyses the language and comes up with suggested list of journals. The Committee on Publication Ethics (COPE) member list provides a ‘white list’ of publishers. Journal Guide helps researchers select an appropriate journal for publication.

A tweet noted that “Librarians and researchers are overwhelmed by the range of tools available – we need a curator to help pick out the best”.

Peer review

Alice Ellingham who is Director of Editorial Office Ltd which runs online journal editorial services for publishers and societies discussed ‘Why peer review can never be free (even if your paper is perfect)’. Alice discussed the different processes associated with securing and chasing peer review.

She said the unseen cost of peer review is communication, when they are providing assistance to all participants. She estimated that per submission it takes about 45-50 minutes per paper to manage the peer review. 

Editorial Office tasks include looking for scope of a paper, the submission policy, checking ethics, checking declarations like competing interests and funding requests. Then they organise the review, assist the editors to make a decision, do the copy editing and technical editing.

Alice used an animal analogy – the cheetah representing the speed of peer review that authors would like to see, but a tortoise represented what they experience. This was very interesting given the Nature news piece that was published on 10 February “Does it take too long to publish research?”

Will Frass is a Research Executive at Taylor & Francis and discussed the findings of a T&F study “Peer review in 2015 – A global view”. This is a substantial report and I won’t be able to do his talk justice here, there is some information about the report here, and a news report about it here.

One of the comments that struck me was that researchers in the sciences are generally more comfortable with single blind review than in the humanities. Will noted that because there are small niches in STM, double blind often becomes single blind anyway as they all know each other.

A question from the floor was that reviewers spend eight hours on a paper and their time is more important than publishers’. The question was asking what publishers can do to support peer review? While this was not really answered on the floor* it did cause a bit of a flurry on Twitter with a discussion about whether the time spent is indeed five hours or eight hours – quoting different studies.

*As a general observation, given that half of the participants at the conference were publishers, they were very underrepresented in the comment and discussion. This included the numerous times when a query or challenge was put out to the publishers in the room. As someone who works collaboratively and openly, this was somewhat frustrating.

The Sociology of Research

Professor James Evans, who is a sociologist looking at the science of science at the University of Chicago spoke about How research scientists actually behave as individuals and in groups.

His work focuses on the idea of using data from the publication process that tell rich stories into the process of science. James spoke about some recent research results relating to the reading and writing of science including peer reviews and the publication of science, research and rewarding science.

James compared the effect of writing styles to see what is effective in terms of reward (citations). He pitted ‘clarity’ – using few words and sentences, the present tense, and maintaining the message on point against ‘promotion’ – where the author claims novelty, uses superlatives and active words.

The research found writing with clarity is associated with fewer citations and writing in promotional style is associated with greater citations. So redundancy and length of clauses and mixed metaphors end up enhancing a paper’s search ability. This harks back to the conversation about poor academic writing the day before – bad writing is rewarded.

Scientists write to influence reviewers and editors in the process. Scientists strategically understand the class of people who will review their work and know they will be flattered when they see their own research. They use strategic citation practices.

James noted that even though peer review is the gold standard for evaluating the scientific record. In terms of determining the importance or significance of scientific works his research shows peer review is inconsistent and systematically biased. The greater the reviewer distance results in more positive reviews. This is possibly because if a person is reviewing work close to their speciality, they can see all the criticism. The greater the novelty of the work the greater likelihood it is to have a negative review. It is possible to ‘game’ this by driving the peer review panels. James expressed his dislike of the institution of suggesting reviewers. These provide more positive, influential and worse reviews (according to the editors).

Scientists understand the novelty bias so they downplay the new elements to the old elements. James discussed Thomas Kuhn’s concept of the ‘essential tension’ between the classes of ‘career considerations’ – which result in job security, publication, tenure (following the crowd) and ‘fame’ – which results in Nature papers, and hopefully a Nobel Prize.

This is a challenge because the optimal question for science becomes a problem for the optimal question for a scientific career. We are sacrificing pursuing a diffuse range of research areas for hubs of research areas because of the career issue.

The centre of the research cycle is publication rather than the ‘problems in the world’ that need addressing. Publications bear the seeds of discovery and represent how science as a system thinks. Data from the publication process can be used to tune, critique and reimagine that process.

James demonstrated his research that clearly shows that research today is driven by last year’s publications. Literally. The work takes a given paper and extracts the authors, the diseases, the chemicals etc and then uses a ‘random walk’ program. The result ends up predicting 95% of the combinations of authors and diseases and chemicals in the following year.

However scientists think they are getting their ideas, the actual origin is traceable in the literature. This means that research directions are not driven by global or local health needs for example.

Panel: Show me the Money

I sat on this panel discussion about ‘The financial implications of open access for researchers, intermediaries and readers’ which made it challenging to take notes (!) but two things that struck me in the discussions were:

Rick Andersen suggested that when people talk about ‘percentages’ in terms of research budgets they don’t want you to think about the absolute number, noting that 1% of Wellcome Trust research budget is $7 million and 1% of the NIH research budget is $350 million.

Toby Green, the Head of Publishing for the OECD put out a challenge to the publishers in the audience. He noted that airlines have split up the cost of travel into different components (you pay for food or luggage etc, or can choose not to), and suggested that publishers split APCs to pay for different aspects of the service they offer and allow people to choose different elements. The OECD has moved to a Freemium model where that the payment comes from a small number of premium users – that funds the free side.

As – rather depressingly – is common in these kinds of discussions, the general feeling was that open access is all about compliance and is too expensive. While I am on the record as saying that the way the UK is approaching open access is not financially sustainable, I do tire of the ‘open access is code for compliance’ conversation. This is one of the unexpected consequences of the current UK open access policy landscape. I was forced to yet again remind the group that open access is not about compliance, it is about providing public access to publicly funded research so people who are not in well resourced institutions can also see this research.

Research in Institutions

Graham Stone, the Information Resources Manager, University of Huddersfield talked about work he has done on the life cycle of open access for publishers, researchers and libraries. His slides are available.

Graham discussed how to get open access to work to our advantage, saying we need to get it embedded. OAWAL is trying to get librarians who have had nothing to do with OA into OA.

Graham talked the group through the UK Open Access Life Cycle which maps the research lifecycle for librarians and repository managers, research managers, fo authors (who think magic happens) and publishers.

My talk was titled ‘Getting an Octopus into a String Bag’. This discussed the complexity of communicating with the research community across a higher education institution. The slides are available.

The talk discussed the complex policy landscape, the tribal nature of the academic community, the complexity of the structure in Cambridge and then looked at some of the ways we are trying to reach out to our community.

While there was nothing really new from my perspective – it is well known in research management circles that communicating with the research community – as an independent and autonomous group – is challenging. This is of course further complicated by the structure of Cambridge. But in preliminary discussions about the conference, Mark Carden, the conference organiser, assured me that this would be news to the large number of publishers and others who are not in a higher education institution in the audience.

Summary: What does everybody want?

Mark Carden summarised the conference by talking about the different things different stakeholder in the publishing game want.

Researchers/Authors – mostly they want to be left alone to get on with their research. They want to get promoted and get tenure. They don’t want to follow rules.

Readers – want content to be free or cheap (or really expensive as long as something else is paying). Authors (who are readers) do care about the journals being cancelled if it is one they are published in. They want a nice clear easy interface because they are accessing research on different publisher’s webpages. They don’t think about ‘you get what you pay for.’

Institutions – don’t want to be in trouble with the regulators, want to look good in league tables, don’t want to get into arguments with faculty, don’t want to spend any money on this stuff.

Libraries – Hark back to the good old days. They wanted manageable journal subscriptions, wanted free stuff, expensive subscriptions that justified ERM. Now libraries are reaching out for new roles and asking should we be publishers, or taking over the Office of Research, or a repository or managing APCs?

Politicians – want free public access to publicly funded research. They love free stuff to give away (especially other people’s free stuff).

Funders – want to be confusing, want to be bossy or directive. They want to mandate the output medium and mandate copyright rules. They want possibly to become publishers. Mark noted there are some state controlled issues here.

Publishers – “want to give huge piles of cash to their shareholders and want to be evil” (a joke). Want to keep their business model – there is a conservatism in there. They like to be able to pay their staff. Publishers would like to realise their brand value, attract paying subscribers, and go on doing most of the things they do. They want to avoid Freemium. Publishers could be a platform or a mega journal. They should focus on articles and forget about issues and embrace continuous publishing. They need to manage versioning.

Reviewers – apparently want to do less copy editing, but this is a lot of what they do. Reviewers are conflicted. They want openness and anonymity, slick processes and flexibility, fast turnaround and lax timetables. Mark noted that while reviewers want credit or points or money or something, you would need to pay peer reviewers a lot for it to be worthwhile.

Conference organisers – want the debate to continue. They need publishers and suppliers to stay in business.

Published 18 February 2016
Written by Dr Danny Kingsley
Creative Commons License