Thursday, 7 June 2018

Towards a new model of academic publishing



            © Yould Publications Ltd 2018
The academic publishing landscape has changed beyond recognition since it began as a ‘gentlemanly’ exchange of ideas in journals like Proceedings of the Royal Academy. Yet the essential model of writing lengthy manuscripts, submission, peer review, editing and publication persists. But for how much longer?

The current situation
Some aspects of academic publishing have changed greatly in the past twenty years, but these have been used to improve and perpetuate the traditional model. The model is based on submission of lengthy manuscripts which are peer reviewed and then either published – after amendment – or rejected. They are then subject to the same process, often repeatedly, until acceptance.


Problems with the current situation
The process is lengthy and, while it is designed to produce high quality published work, it is no guarantee of quality or honesty, as the number of retracted manuscripts suggests. Producing lengthy manuscripts is time-consuming and then the peer review process can last from a few weeks to over a year for some journals. The final product, usually a lengthy article, is then published and in many cases never read. Even if the article is read, few people read every section – especially the introductory and discussion sections – focusing rather on the methods and the results. Introductions and discussion probably serve a useful function for editors and reviewers, but they are often a hard read for other academics, students and the public who simply want to know what was done and what was discovered.

Improved processes
The development of the Internet and the World Wide Web have been significant influences on the academic publishing industry. However, they have mainly been used to accelerate and automate the publishing process, they have not been used to change the process and outcomes. Online platforms are now almost universally used by academic publishers for submitting manuscripts, organising the peer review process and then publishing articles. On the back of this has arisen the open access movement which has put pressure on publishers to initiate processes whereby their content can be made available to read without restriction. Two main models operate known and ‘gold’ and ‘green’ routes whereby, respectively, the author either pays to make the final published article available or they may make the final accepted version of the manuscript available on a repository following an embargo period imposed by the publisher. Even more recently, the ‘diamond’ route to open access has developed as evidenced by, for example, the WikiJournal of Medicine. Here, no money changes hands either to publish or to read articles.
Following publication, and as another result of the Internet, the rise of social media – Twitter® in particular – has had a profound influence on the way publications are disseminated. Publishers, long used to disseminating journals contents lists and occasionally other information about their journals, now have journal websites and these are used to disseminate content and to highlight specific articles. A journal Twitter® site often spearheads a range of other social media such as blogs, podcasts or YouTube®. Again, this is not a variation on the publishing process, merely an alternative way of advertising content. However, it is quicker, it can be automated and is easier to access by a range of platforms and in almost any location.

An unintended consequence of online publishing and open access is the rapid growth of predatory publishers. According to the eponymous--and now withdrawn--Beall’s list there were approximately 100 in 2013; now there are over 1000. Predatory publishers cast suspicion over the legitimate aspects of the business run by publishers such as PLOS ONE and Biomed Central, and by publishers offering hybrid journals – where articles may be both pay to view or open access.

New processes
While the above factors have changed the process in terms of speed and efficiency and have also made the process virtually paperless, other changes resulting from the Internet have the potential to alter the process more substantively. These changes are described below.

Data repositories
From the days when data were jealously guarded, and researchers considered it a virtue – often enforced by research ethics committees – to destroy data within a few years of completing a project, it is now almost de rigueur both to maintain data in perpetuity and to make it publicly available. Large databases are now available for secondary analysis, combining with other databases and, increasingly, subject to big data analysis and data-mining. A range of repositories is available for storing and sharing data and these include ones that are hosted by universities and academic publishers and specialised project databases such as SHARE and BioBank.

Supplementary material
Related to the issue of making data available is the use of supplementary material by journals as this is how they tend to make data available. However, it is possible to make other information available such as additional tables, figures, appendices and even reference lists if these aspects are restricted by the journal guidelines. Supplementary material can be made available online and accessible via hyperlinks embedded in the online version of the published article or via hyperlinks on the webpages of the journals. In fact, some publishers provide hyperlinks for many cited articles taking readers directly to the online versions where these are available.

Study registration
Largely due to the AllTrials campaign for transparency in the conduct and reporting of clinical trials there has been a significant move towards registering the protocols of studies in advance of conducting them. A range of registries exist across the world – ClinicalTrials.gov being a typical example – and these are, essentially, databases where researchers intending to conduct a clinical trial can publish the protocol. These databases permit the public to see what trials have been registered and there is provision for publishing deviations from protocols, an indication of completion of the trial and then a summary of the results. These registries are not confined to trials. It is more common to see other designs being registered and, indeed, it is now common for journals to publish protocol articles. In addition, it is becoming more common to register systematic reviews and the Prospero website exists specifically for this – although it is confined to systematic reviews of clinical studies. The Cochrane library also publishes the protocols of reviews that are being conducted under their auspices. The situation regarding study registration is being levered further by academic journals and publishers as they strive to be at the forefront of setting and maintaining rigour and transparency in academic publishing.

Preprints
Another phenomenon, which began in the 1990s but has accelerated recently, has been the publication of preprints. Described by Wikipedia as: ‘…a version of a scholarly or scientific paper that precedes publication in a peer-reviewed scholarly or scientific journal. The preprint may be available, often as a non-typeset version available free, before and/or after a paper is published in a journal’. Preprints have gathered credibility with research funders and most publishers. Preprints arose because of frustration with the length of time taken to review manuscripts and the need to share scientific results early. Academic publishers have largely accommodated the existence of preprints and the need to accelerate the publishing process. The increasingly common phenomenon of publishers facilitating the publication of final accepted copies of manuscripts in addition to the early view of corrected proofs acknowledges this. The Times Higher Education recently described preprints as being ‘largely indistinguishable’ from the final published articles.

Post-publication review
Post-publication review, as opposed to the usual pre-publication system of review, is comment on publications once they have been published. This has always been possible either by correspondence with authors or in the correspondence pages of journals. But the advent of the internet and social media has facilitated this and led to some consideration of whether this is a serious contender to pre-publication review. A process could be imagined whereby manuscripts are posted online – possibly as pre-prints – to receive comment and then be altered accordingly, thereby to change and evolve as comments are made. Alternatively, published articles – either peer reviewed or not – could receive comments linked to the article for readers to take into consideration. Potentially several models exists and there has been the rise of platforms, eg PubPeer, which exist for this purpose, and facilities for comment within other platforms such as ReseachGate. Recently publons, a site for recording reviewing and editing activity which is run by Clarivate, has championed post-publication review.

Wikipedia
Wikipedia is an online open access encyclopaedia which may be freely edited. There are strict guidelines about what may be entered on Wikipedia and edits are closely monitored to ensure that all entries are useful and demonstrably linked to reliable sources. One extension of Wikipedia is Wikiversity, a Wikimedia project, which aims to support learning through provision of courses and tutorials. Unlike Wikipedia – where entries are not peer reviewed – Wikiversity allows the publication of original research through the WikiJournal project and this started with the WikiJournal of Medicine. This journal aims to publish peer reviewed articles which are then available open access. No article processing charge is made to authors; therefore, this is the ‘diamond’ route to open access and, being published as editable Wikipedia pages means that articles may be edited post-publication.

The Conversation 
The Conversation is an independent source of news and views, sourced from the academic and research community and delivered direct to the public. The Conversation ‘motto’ is ‘Academic rigour, journalistic flair’. Only academics or research students with official university email addresses may contribute to The Conversation and potential articles are ‘pitched’ using a structured format to a specialist editor who can decide to accept or reject the article. If accepted, the article must be written to a struct format – usually 700 words – and in language of a very high general readability: that of an educated 16-year old. Jargon and complex sentences are discouraged, and authors may check and adjust the readability of their pieces during the process of submission. While articles are not peer reviewed, the editors will scrutinise them and check the provenance of the evidence being cited. Not every piece in The Conversation reports directly on research – some verge on opinion pieces; but backed by evidence. However, many are very effective abstracts linked to original reports and data.

How should the academic publishing industry respond?
The academic publishing industry, which has proved itself to be adaptable over the past twenty years, will continue to adapt. It must, resistance to the changes already taking place would be futile. The changes taking place should be seen both as a sign that the traditional model of academic publishing is under further threat but also as an opportunity to increase their utility to the academic community.

Assimilated
While all the above developments have been challenges for the academic publishing industry, some have been assimilated by the industry and are now considered good practice. Indeed, some have enhanced academic publishing, including:
  • data repositories 
  • supplementary material 
  • study registration 
Threatening
Other issues, currently tolerated by the academic publishing industry, are threats to the ‘traditional’ model because they either pre-empt or by-pass the established systems of peer review. Other more recent changes could be viewed as threats which have not yet been addressed and which, if they take hold, may erode the foothold of the academic publishing industry. This is not likely to be very rapid, given the extent of the established academic publishing industry in most fields, and there is time for the industry to adapt; these threats are:
  • preprints 
  • post-publication review 
  • Wikipedia 
  • The Conversation 
Issues to overcome
The issues to be overcome are:
  • Quality control 
  • Bibliometrics 
It is also questionable what the role of academic publishers would be if all the developments outlined above became established in academic publishing.

Publishers may have to reinvent their role but, of course, none of this is going to happen immediately. Therefore, publishers may be reluctant to change, significantly, the way they work. All will monitor competitors to see how far they are willing to go and to see if any new approaches, pioneered by others, are working.

Publishers could and, indeed, should carve out a role for themselves in this new situation. They already have vast experience, resources and staff and – of prime importance – reputation. While individual initiatives such as the WikiJournal of Medicine may multiply they are unlikely and probably do not intend to become major competitors to any of the established academic publishers. Academic publishers operate as competitors but there is also considerable cooperation and cross-fertilisation of ideas and movement of staff between companies. If a major threat was perceived there would probably be some coordinated and collaborative action.

Academic publishers should be capable of providing gateways into academic publishing which provide a ‘one-stop shop’ to features such as online registration of studies, data repositories, quality assurance through peer review (at one or several stages of the process) and allocation of Digital Object Identifiers. Publication of streamlined – The Conversation type – summaries of research and scholarship which provided all the necessary links to data, registration and features such as tables and figures could be one model. They could also provide Wikipedia like platforms for updating and amending articles.

In terms of peer review, the established publishers can offer platforms that individual diamond routes such as WikiJournal of Medicine are unlikely to be able to afford. Reviewing and manuscript processing platforms such as ScholarOne® and the Elvise® system are very expensive and require regular maintenance, customisation and upgrading. Therefore, operating such platforms is likely to remain solely in the domain of the established academic publishing companies. Thus, while any journal of whatever size may operate a system of peer review, the established publishers will do it better.

Bibliometrics
In this context is meant, specifically, bibliometrics related to journal and author performance in terms of citations. Currently, there are several systems such as Clarivate, Scopus and Google Scholar which record these and, based on citations, there are several ways of calculating bibliometrics. But, for journal performance the impact factor remains the most commonly understood and accepted method and the impact factor published by Clarivate is considered the gold standard. To measure author performance – and again Clarivate provides the gold standard measure – the h-index is a widely accepted measure.

If the academic publishing industry is driven purely by these bibliometrics then the many aspects of the model proposed above will not happen. However, there is some pressure from within academia to stop using bibliometric to measure performance and this was best articulated in the DORA (Declaration on Research Assessment). A new model of publication may give impetus to the end of the use of citation-based metrics. Nevertheless, an alternative, more suitable to the potential changes in academic publishing does exist in the form of alternative metrics – the Altmetric score – based on online mentions of articles across a range of platforms such as online newspapers and social media. Some academic publishers are already providing Altmetric scores on the landing pages of articles.

Is there a role for editors?
Currently, with support from publishers, editors manage and operate the systems that are involved with submitting and publishing an academic article. Models and titles change across the industry but, currently, it is hard to envisage the system working without editors who interact with authors and reviewers. Moreover, it is editors who make decisions at every step following submission through to acceptance. Unless any significant changes in academic publishing lead to a ‘free for all’ in terms of what can be submitted and published, it is hard to envisage any model working without editors. Even if there is a significant move towards post-publication review someone must still decide if post-published comments are acceptable and if changes ought to be made on their basis to article. This is, for example, how Wikpedia works where editors check entries and updates for utility and sourcing – having already decided about the worth of the entry in general. This is one possible model for academic editors who, in addition to scrutinising suggested amendments for utility and sourcing could also judge the scientific aspects of reviews and, if necessary, draw on further experts to assist. They would also have to investigate claims of scientific misconduct. Therefore, it is unlikely that the role of editors can be dispensed with; but editors would have to learn to work differently.

Conclusion
The academic publishing industry continues to change and we can only assume further change will come. In parallel, the established processes of publishing are being eroded and alternative models of publication are developing. Particularly, the traditional academic article where all the aspects of a study are gathered in one long piece of writing, is surely on the verge of extinction. The purpose of such articles was to gather together information, with references, that was otherwise inaccessible couched within a structured argument arranged into relatively standard sections. Often, the essential information contained in such articles tells the reader little more than the abstract. In the age of social media, information overload and with the hundreds of journals available in most fields, publishers are going to have to find new ways of packaging their products and the product and the processes of production must also change.

The issues of quality and metrics are not insurmountable, and the role of editors is unlikely to become redundant. The traditional reliance on citations alone and derivative calculations may be threatened; this will be welcomed by many. Alternative metrics, more tuned to the way information is now obtained, already exist and these could be used in conjunction with or even as an alternative to metrics based on citations.

****

This is a position paper reflecting the views of the author and aims to stimulate discussion amongst academics, editors and publishers. Roger Watson declares the following interests: Editor-in-Chief, Journal of Advanced Nursing; Editor, Nursing Open; Editorial Board member, WikiJournal of Medicine.

Roger Watson



Tuesday, 13 December 2011

Peer review under the spotlight in the UK

Last year, the UK government put the peer review process for scientific publication under scrutiny in the shape of a House of Commons (HoC) Science and Technology Committee report entitled Peer review in scientific publications (HoC 2011a).  The report was produced on the basis of written and oral evidence to the committee and those with sufficient interest can actually watch video footage of the proceedings on the internet (http://www.parliamentlive.tv/Main/Player.aspx?meetingId=8301; retrieved 11 December 2011).  Perhaps not the greatest draw for the majority of the public, but for someone deeply engaged with the academic publishing industry, I found the live proceedings fascinating and the ensuing report very informative.  It is always of great interest to see how others view one of the fundamental pillars of our industry.  The evidence was given by publishers, editors and organisations with an interest in the process of scientific publishing, but the questions were asked my Members of Parliament with little or no knowledge of the scientific publishing industry.

The UK government has previously considered peer review (HoC 2011b) and several organisations have also published on the process in recent years (Council of Science Editors 2006, The British Academy 2007, Thomson Reuters 2011, Ware 2008, Ware & Monkman 2008).  The immediate reason for some of the recent interest were doubts about the process related to some peer reviewed information, later questioned in terms of its accuracy, that was published about climate science (HoC 2011b) and the—somewhat naive—assumption by the press, the public and politicians that the peer review process protected against this kind of thing that, in the words of the ‘Climategate’ report, that the peer review process was a ‘firewall’ between truth and falsehood (http://www.scribd.com/doc/34003747/Muir-Russell-Final; retrieved 13 December 2011) which, clearly, it was not.

The value of the UK HoC report is that it is, surely, one of the most comprehensive and authoritative documents on peer review that exists.  As such, it should be compulsory reading—at least the first 100 pages that contain the main body of the report—for editors, publishers, reviewers and, indeed, scientific authors.  The report is wide-ranging, taking in definitions of the peer review process and what it aims to achieve; a review of current practices and discussion on their efficacy.  The role of editors, authors and reviewers is discussed and new approaches to reviewing such as post-publication approaches and publishing ethics.  In addition, the review of data in the reviewing process, not something I have encountered in nursing, is covered.

The peer review process
It is worth reiterating the terms of reference of the HoC committee in full to give a clear idea of what the purpose of the investigation was and what the report addresses, these were:
1. the strengths and weaknesses of peer review as a quality control mechanism for
scientists, publishers and the public;
2. measures to strengthen peer review;
3. the value and use of peer-reviewed science on advancing and testing scientific
knowledge;
4. the value and use of peer-reviewed science in informing public debate;
5. the extent to which peer review varies between scientific disciplines and between
countries across the world;
6. the processes by which reviewers with the requisite skills and knowledge are
identified, in particular as the volume of multi-disciplinary research increases;
7. the impact of IT and greater use of online resources on the peer-review process;
and
8. possible alternatives to peer review.
The key features of the peer review process were summarised and included acknowledgement of the fact that many journals are under extreme pressure for the limited space they offer due to the very high level of submissions they receive.  This means that there is often a ‘triage’ stage where some kind of ‘in-house’ evaluation of manuscripts takes place whereby editors make a decision about whether items go forward for review.  A great many manuscripts are rejected at this stage.  The different types of review: single-blind; double-blind; and open, were described but, whatever the process, the variation in quality of reviews was acknowledged.  New models of publication—including open access—were included and it was indicated that these newer, less traditional, forms of publishing, which are gaining popularity, are not devoid of a peer review process.

With increasing submissions to scientific journals, approximately three million manuscripts annually, and the increasing complexity of research—inter- and multidisciplinary—the need for more reviewers, in some cases per paper, and the increasing frequency with which they were asked to review was acknowledged as a burden.  Nevertheless, even without concrete evidence of its efficacy, the peer review process is widely considered to be essential to the scientific publishing industry. Moreover, participating in peer review is also considered ‘integral’ to the career of a researcher.

Participants
Editors, authors and reviewers participate in the peer review process and, of course, many people play all three roles as only active and successful academics are likely to be asked to review manuscripts and be appointed as editors.  However, the activities related to the publishing industry remain largely the domain of the gifted amateur and, while there is training available for editors and peer reviewers, this is far from the norm.  Of course, the issue of who would fund such training was raised and there was some indication that this might be in the domain of the research councils.  Surprisingly, the publishing industry—which benefits financially from the peer review process—was not identified as a source of funding.

Returning to the issue of burden, the difficulty in and processes for finding reviewers was covered as were mechanisms for cutting down on the burden such as cutting out re-review by putting the responsibility for ensuring the validity of papers on the authors following an initial review and then comprehensive instructions from an editor on what needed to be done to ensure publication.  However, not acknowledged in the report, depending on authors who are ambitious and dependent on their publications for promotions and research grants could lead to some malpractice and, of course, increases the burden on editors to provide more detailed feedback and to oversee this process.

The general lack of recognition that academics who carry out peer review receive was acknowledged and, speaking as a reviewer and an editor, it is clear that those who willingly and efficiently review manuscripts are those who are repeatedly asked to do it and the burden can be significant.  This is highly legitimate academic work but, apart from a general expectation that this is part of the duties of an academic, it is rarely if ever precisely measured and taken into account in staff review and promotion rounds.

Impact and integrity
The perceived importance of having well published research, implying quality and importance of the published finding, is the main reason why the peer review process was being considered.  It was acknowledged that publication in a peer reviewed journal, especially one of high impact, tended to signal something about the standing of the published work.  While publication metrics were discounted in the UK Research Assessment Exercises and would be in the forthcoming Research Excellence Framework, it is clear that the vast majority of ‘outputs’, which include books, reports and patents, are papers in refereed high impact factor journals.

In addition to impact, the peer review process both conveys and requires integrity.  Without question, reviewers are supposed to hold the highest professional standards and, for instance, to declare and conflicts of interest with manuscripts they may be reviewing.  In addition, the highest standards of integrity are expected of authors; unfortunately, they are not always adhered to and reviewers, alongside editors, play their part in detecting fraud, duplication and plagiarism. All editors will report an increase in the incidence of malpractice in publications and the reasons must include the sheer increase in the number of submissions, an increased awareness—thanks to organisations like the Committee on Publications Ethics (http://publicationethics.org/; retrieved 13 December 2011)—of the issues involved and the inception and more widespread use of new technology to detect malpractice.

Conclusions
The report concluded that peer review was crucial and that, while there were different models available, they did not all suit all types of publication.  The role of editors was acknowledged and that, while they often lacked formal training, a great deal was offered ‘on the job’.  Integrity and transparency were essential aspects of the process.  However, none of the above guaranteed the worth of a piece of research; its worth remains in the eyes of the person reading it, for which, as the HoC report (2011 p. 94) stated: ‘there is no substitute’.

Roger Watson
Editor-in-Chief, JAN
Email: jan@wiley.com

References
Council of Science Editors (2006) CSE’s white paper on promoting integrity in scientific journal publishing CSE, Reston VA

House of Commons Science and Technology Committee (2011a) Peer review in scientific publications The Stationery Office Ltd, London

House of Commons Science and Technology Committee (2011b) The reviews into the University of East Anglia’s Climactic Research Unit’s E-mails The Stationery Office Ltd, London

The British Academy (2007) Peer review: the challenges for the humanities and sciences The British Academy, London

Thomson Reuters (2011) Increasing the quality and timeliness of scholarly peer review Thomson Reuters, New York NY

Ware (2008) Peer review: benefits, perceptions and alternatives Publising Research Consortium, London

Ware & Monkman (2008) Peer review in scholarly journals: perspectives of the scholarly community – an international study Publising Research Consortium, London

Sunday, 19 June 2011

What systems of peer review are available?

I'm not the first to enter the publishing blogosphere with questions and views on peer review processes for scholarly publishing...and I won't be the last.  However, in teaching an online unit on Information and Communication in Healthcare, a large proportion of which is about the academic publishing process - especially as it applies to nursing - they students have to compare and contrasts the various methods and, of course, there are several variations on a theme.

Traditionally, the peer review process is one where neither party knows the identity of the other (double-blind peer review), but this is lately being challenged with more open reviewing systems such as in BioMed Central journals where identity is not only know, the reviews are published alongside the papers.  BioMed Central have prodiced a rigorous system that is leading to the publication of high quality papers but, at the extreme end of the online publishing industry, there is post-publication review where manuscripts are published in whatever state they are submitted, and reviews follow online.  This strikes me as a very public way of making your mistakes and, of course, if the reviews do not lead to revision of the articles, then this process - unlike more 'traditional' reviewing routes - does not lead to improvement of the published articles.

Therefore, the routes for reveiwing include:

  • Open reviewing - the identities of reviewers and reviewed are mutually known
  • Single-blind reveiwing - the reviewer knows the identity of the reviewed (usually) but vice versa is theoretically possible
  • Double-blind reviewing - neither the reviewed not the reviewer know the identity of the other.
Of couse, in the double-blind reviewing system the reviewer, ultimately, knows the identity of the reviewed.

Within all of the above the variations include the number of reveiwers - clearly, the bare minimum is one, otherwise review has not taken place, but two is common and some journals request many reviews - I think the record I have heard in a North American psychology journal is around 10.  In addition, some journals ask authors to suggest reviewers - some requesting that they are not pre-warned by the author; others requesting that they are to avoid asking reviewers who will not respond.  Journal may use these reviewers exclusively; others will use them - or some of them - and also some of their own, thereby introducing an element of uncertainty about who the reveiwers are.

If I have not exhausted the possibilities here then please leave a comment.