HEFCE announced further clarification on the rules for REF2021 yesterday (PDF here) and this included some decisions on outputs and open access (on p.8). Key points are:
- The policy is felt to be working in that more papers than ever before have been made open access.
- The original policy stands with respect to the timeframe for researchers to act: papers must be uploaded to RIS within 3 months of acceptance.
- However, there will be an exception to cover papers that miss this deadline but are uploaded within 3 months of publication.
The rules around this crucial timing issue have varied during the REF period so a summary is given below:
Papers accepted for publication before 1 April 2016 do not need to comply with the REF open access policy in order to be submitted to the REF (however they are encouraged to be made open access).
Papers accepted for publication between 1 April 2016 and 1 April 2018 need to comply with the REF Open Access policy to be submitted: for this period, the full text of the article needed to be uploaded into RIS within 3 months of the date of online publication (unless the paper was published with Gold Open Access on the publisher site or one of the other exceptions can be applied). Papers that were not uploaded to RIS within 3 months of publication (or which failed to meet other conditions of the policy e.g. minimum embargo period) cannot be submitted to the REF.
Paper accepted for publication after 1 April 2018 will need to comply with the REF Open Access policy to be submitted: the full text of the paper must be uploaded into RIS within 3 months of acceptance for publication (unless the paper was published with Gold Open Access on the publisher site or one of the other exceptions can be applied). It will be possible to claim an exception for papers that miss this deadline but which are uploaded within 3 months of publication. Papers that are not uploaded to RIS within 3 months of publication (or which fail to meet other conditions of the policy e.g. minimum embargo period) cannot be submitted to the REF.
The Library Research Support team checks for REF compliance and reports on this to Research Directors & REF staff. We are currently seeing around 85-90% of papers complying with the REF Open Access policy. We will be doing further publicity in 2018 to ensure all researchers are aware of what they need to do and know that we will be glad to help with any questions: email@example.com.
Filed under Open Access, REF
Saturday 1st April 2017 marked the one year anniversary for the REF Open Access policy: this covers all journal articles and (some) conference proceedings accepted for publication after 1st April 2016. These papers must comply with HEFCE’s Open Access policy or they cannot be submitted to the REF.
Green Open Access on Cronfa
We have seen much progress around open access at Swansea University in the last year. Unsurprisingly, there has been a marked increase in papers made open access on our repository Cronfa: the REF policy is all about encouraging researchers to take advantage of publisher copyright policies that allow the accepted version of an article to be made public on a repository, known as “Green Open Access”.
The home page of Cronfa shows our latest full text additions, plus the most-downloaded articles of all time and the last week/month. Whilst Cronfa documents feed into the Core repository search portal and will turn up in Google search results, they are still not appearing regularly on Google Scholar. Reasons for this are not clear but we continue to investigate and are not alone in having this issue with our repository content. If you have an open access version of a paper on Cronfa circulating the URL to the open version ensures maximum impact.
Gold Open Access
We are also seeing many articles published with “Gold” (paid-for) open access. This is not essential for the REF Open Access policy (unless the chosen journal does not permit self-archiving to comply with the HEFCE policy). Whilst we do have money available for RCUK-funded publications, most of the Gold Open Access papers are paid for from research funding or other sources; Swansea University does not have an institutional fund for open access.
We have also seen 27 Swansea University authors take advantage of the excellent Springer deal for free open access in selected journals. This option is available to any staff or student who is corresponding author on a paper submitted to certain Springer journals.
REF Open Access Policy Compliance
For various reasons, it is difficult to give precise figures for REF Open Access compliance at this stage but our estimates suggest we are seeing strong levels of compliance (very rough estimate = 85-90%) for all papers that have been added to RIS (not just those that may be submitted to a future REF). There may be papers published by Swansea University authors but not yet added to RIS which would alter this estimate. The university’s own open access policy means that ALL publications should be made open access where possible, not just those that may be submitted to the REF.
Increased Support for Open Access
The Library Research Support Team expanded with 2 new posts in 2016 which reflects the additional reporting and compliance work around open access: Caroline Rauter is the Scholarly Communications Officer (firstname.lastname@example.org) and Penny Lauder is the Scholarly Communications Assistant (email@example.com). The team is managed by Annette Linton, Head of Library Content and Scholarly Communications (firstname.lastname@example.org). We can be contacted on email@example.com and more information on the support we offer can be found on our web pages. We work closely with staff in the Colleges on advocacy and support for open access.
The College of Engineering have also appointed Rebecca Kelleher as REF Officer (firstname.lastname@example.org), who offers a REF compliant mediated deposit service to RIS/Cronfa for all staff in Engineering and Sport & Exercise Sciences and she also reports on Open Access compliance for the College.
Filed under Open Access, REF
Short answer: no. The accepted version of a paper needs to be uploaded to RIS as well.
The HEFCE REF Open Access policy includes “subject based repositories” as a suitable home for open access papers. The associated FAQ state that HEFCE do not stipulate which repositories meet their requirements. One of the best known subject repositories is Arxiv, hosted at Cornell University Library: “Open access to 1,225,076 e-prints in Physics, Mathematics, Computer Science, Quantitative Biology, Quantitative Finance and Statistics” (at time of writing). These subject areas have a longstanding tradition of open science: scholars publish pre-prints on Arxiv for review which then later may get published in scholarly journals.
Traditionally Arxiv is a “pre-print” server rather than the accepted version (or post-print) that is needed for the REF policy, although sometimes these will also be on the site. Unfortunately at the moment files uploaded to Arxiv cannot be used to satisfy the REF Open Access policy because Arxiv does not record the date of acceptance for a paper and the version of the paper to satisfy the REF OA policy’s technical and audit requirements. We believe discussions are ongoing to try and resolve this but Swansea University researchers need to ensure they follow the guidance for compliance with the open access policy and upload the accepted version of all papers to RIS at acceptance for publication. The university’s own open access policy requires that researchers upload the accepted version of a paper into RIS so this remains the position even when an open version of the paper is on Arxiv.
Guidance on the open access policies is on the web here or SU researchers are welcome to contact the Library Research Team at email@example.com.
Filed under Open Access, REF
As you may have seen in the press the Stern review has just issued a report of its recommendations. Some of the main ones are:
- That all research active staff should be included in the REF.
- Outputs should be submitted at Unit of Assessment Level. There will be an average number of submissions but with flexibility for some people to submit more and others less than the average.
- Outputs will not be portable between institutions – this is to avoid universities “buying in” researchers with a good portfolio just before the REF.
- Peer review will still be the main means of assessment although metrics will be provided for panels.
- There will be institutional level impact case studies to encourage universities to show their interdisciplinary impacts. Support for interdisciplinary and cross-institutional initiatives are also mentioned under environment.
As all researchers are now likely to be included in the REF it is even more important that everyone submits their work to the Research Information System.
If you want to know more try:
The Stern Review
Stern aims for lower REF workload, more “game-changing” research. THES
Will REF “portability” plans hobble early career academics. THES
Being Stern about portability / Athene Donald’s blog
Stern review deserves a co-operative response / Universities UK blog
Universities cautiously welcome “sensible” Stern review / Research Professional
HEFCE have updated their FAQs for the REF Open Access policy (no.7.3) with a small but significant statement concerning books (in bold below):
“As noted in FAQ 2.1, it is not a requirement that books be made available in an open-access form to be admissible as outputs to the next REF.
As we stated more broadly in our circular letter of July 2015 (20/2015), we will expect institutions to include a short description in each submission on progress towards delivering open access, including their overall approach to open access strategy and infrastructure. While we have not developed detailed guidance, we may be particularly interested in any quantitative evidence of the proportion of published books that are made available as OA, as part of this statement.“
We have already blogged on developments around the open access academic monograph and the open access options available, most of which have a fairly substantial cost. It will be interesting to see how any future requirements for books to be made open access (as seems likely for the REF beyond the next one) can be achieved without substantial resourcing. This blog post gives an interesting overview and highlights the potential role of the university press in this.
As stated above, the current REF Open Access policy does not include books and book chapters. Regarding journal articles and conference proceedings, it notes that “institutions can achieve full compliance without incurring any additional publication costs through article processing charges” i.e. by taking the green route to open access.
The countdown is over – the REF policy is now active
Any papers accepted for publication after 1st April 2016 must now comply with the REF’s open access policy. The university expects all papers to be compliant even if you do not anticipate that they will form part of the REF – the university’s own open access policy applies to all publications produced by our researchers.
To ensure you comply with the policy:
- Take action when you have a paper accepted for publication
- Enter details of the publication onto RIS including the accepted date
- Upload your accepted manuscript into RIS
- Check the copyright for the journal using the Sherpa Romeo database
- Use the “Publish to Cronfa” link to make your file open access, setting an embargo period if required. The system will delay publication until the date you specify.
The REF policy limits embargo periods to 12 or 24 months so contact us if you find the journal exceeds this.
We are here to help: firstname.lastname@example.org. Please send us your accepted manuscript and details of the publication including the date of acceptance.
Quick guide to uploading files to RIS (PDF)
Guide to using RIS (PDF)
(We have some briefing sessions next week if you want to learn more or ask questions.)
Filed under Open Access, REF
The REF open access policy (which comes into effect on 1st April 2016) was not designed to compel authors to pay for gold open access. In the original policy they state: “Institutions can achieve full compliance without incurring any additional publication costs through article processing charges”. Instead the focus of HEFCE’s policy is encouraging researchers to take the free “green” route of uploading their papers to institutional repositories.
Do you need to pay for gold open access for the REF?
No, only in one situation: HEFCE have a limit on the embargo periods permitted for the green self-deposit route (12 months for Panels A & B, 24 months for Panels C & D). If you have had an article accepted for publication in a journal which has an embargo period longer than HEFCE allows then you will need to pay for gold open access so that your paper is eligible for the REF unless you can justify an exception. We have a central fund available for RCUK researchers; an agreement is also in place that covers many Springer journals.
What do you need to do for the REF if you paid for gold open access?
If you have paid to publish open access you don’t have to upload a version of your paper to RIS to comply with the open access policy for the REF but it is still strongly encouraged (see below).
What you DO need to do is to set the drop-down box in RIS for “Processing Charge” to “Paid” and add details of the funder(s) in the following box (separate multiple funders with a comma):
This will flag in the system that this is a Gold OA paper and we will not then chase you up for your accepted manuscript! ISS staff may upload a copy of the paper where permitted to the repository, particularly if you have used our RCUK fund.
The policy background
If you have paid to publish a paper via the “Gold” route, then you are covered by the exception stated in policy section 38f: “The output was published as ‘gold’ open access (for example, RCUK-funded projects where an open access article processing charge has been paid)”. However they do still “strongly encourage these outputs to be deposited in a repository to facilitate preservation, aggregation and text-mining”.
HEFCE adjusted the policy back in 2015 and this is the wording of their section on “Gold open-access outputs”:
“We further recognise that many papers will be published as ‘gold’ open access, and will therefore be available as the final published version-of-record2. We believe that there are significant benefits to the deposit of gold OA outputs – repositories support the effective preservation, aggregation and text-mining of research material. However, we recognise that when publishing as gold OA, authors typically prefer to deposit the final published version instead of the accepted manuscript, and that in some cases this will not be available within three months of acceptance. In light of this, we have decided to introduce an exception to the deposit requirements for outputs published via the gold route. This may be used in cases where depositing the output on acceptance is not felt to deliver significant additional benefit. We would strongly encourage these outputs to be deposited as soon as possible after publication, ideally via automated arrangements, but this will not be a requirement of the policy.”
Filed under Open Access, REF