I was extremely pleased to contribute an article on Open Acess to the latest issue of Phenotype Journal, the magazine of the Oxford University Biochemical Society. The full issue is available online and looks gorgeous! Thanks to the absence of length and formatting constraints on this blog, I added a few more references to the version of the article posted here.
Open research encompasses a multitude of views on how to make science a more collaborative, transparent and resource-efficient process. Open Access (OA), open data and open education have become widely used expressions, each referring to a specific aspect of the wider debate on good research practice and knowledge transfer. This article focuses on the accessibility of research data in the physical and life sciences, although the paradigm of open research is equally relevant to the humanities and social sciences.
In the traditional model of scientific publishing, authors are not charged to publish their article in journals (although colour pages can be very expensive). Publishers instead source their revenue from institutional subscriptions, views and downloads from non-subscribers, and third-party advertisements. In this scheme, commercial publishers report higher profits compared to their counterparts managed by learned societies and academic institutions. The ‘figures of the business’ can be tricky to obtain: while some publishers refuse to disclose how profitable their journals are, others admit that they lack a clear idea of the publishing cost of individual articles (1).
Despite this uncertainty, some numbers do appear in the publishing literature. A survey by the Association of Research Libraries in the US reported average yearly subscription prices for individual journals well above $1000 – for instance, $4215 in chemistry and $2520 in biology (2). These figures raise the question of how lucrative publishing really is. Most publishers stress the added value brought by their services, which ensure a high-quality finished product and successful dissemination of a research team’s achievements. However, available data demonstrates an increase in journal subscription costs over time that is not justified by inflation.
If traditional publishers have broken the delicate balance of ‘editorial service and economic return’, what is the alternative? OA can be regarded as one of the sector’s major recent innovations; its growing success suggests that it is a viable alternative to subscription-only publishers. What is clear is that it has already opened a debate into publishing practices involving scientists, librarians, journalists and research funders. Indeed, there is an increasing demand from funding bodies for the results of research that they supported to be made publicly available, which is particularly important for projects funded by the taxpayer.
Broadly, OA sets out to challenge and overcome an issue that is all too familiar to university researchers, teachers, pupils and curious individuals – being unable to read past the abstract of a paper because full access requires an institutional subscription. This situation is described by many OA advocates as ‘hitting the paywall’. If you have ever found yourself in this scenario, you may wish to make sure that your own articles will not face the same fate by publishing your results in an OA journal (also called the ‘gold route’ to OA). The Directory of Open Access Journals (DOAJ) currently lists more than 10,000 OA periodicals. Inclusion in this directory is based on a list of criteria established by the Open Access Scholarly Publishers Association (OASPA) and the overall figure testifies to the success of this alternative publishing model. Authors are usually asked to pay an article processing charge (APC) upon acceptance and prior to publication; from then on, their paper is freely accessible and reusable to the extent dictated by the article license.
An alternative option to OA publishers, already popular among some disciplines, is to self-archive papers, dissertations and theses in one of over 3,000 repositories. This is sometimes referred to as the ‘green route’ to OA. It is commonly adopted in physics, astronomy, earth sciences and mathematics, but still appears to be poorly acknowledged in the medical and life sciences (3).
Given this picture of the publishing landscape, why does the debate around OA become outright incendiary at times? Some researchers are doubtful of the validity of gold OA due to high APCs; defenders of the OA model reply by recalling that there exist advantageous membership schemes to cut off such charges, and that partial to full waivers are also available. Some scientists are also persuaded that only subscription periodicals ensure high standards of review and copy-editing for their papers. However, there is no fundamental reason why an OA journal should not be able to provide the same quality of services as its traditional counterparts; furthermore, article retractions can be observed in both publication categories.
Besides matters of personal taste and beliefs, constructive criticism may well benefit the publishing sector as a whole. This already holds true for the OA ecosystem where new ideas, journals and editorial platforms are implemented and reviewed relentlessly – from open peer reviews, which in OA journals such as eLife can be read alongside the main paper, to the possibility of publishing negative results (4) or partial findings followed by updates (5). The DOAJ, OASPA and similar initiatives aim to constantly monitor and improve the quality and reputation of OA publishers: their approach relies on a white list of trustworthy OA periodicals, in a strategy that opposes black lists of ‘predatory publishers’ such as those compiled by librarian Jeffrey Beall (6).
A closely related initiative to OA is that of open data. Interestingly, stored in the about 3000 web repositories one can find journal articles and theses, but hardly any datasets. The idea of making research code and data available is not always greeted with enthusiasm within the scientific community: the fear of losing the exclusive on a piece of research and the prospect of hours spent reformatting data can discourage even the most open-minded scientists (7). However, more and more journals require authors to include datasets and analysis algorithms with their papers; willingness to set up new collaborations or accelerate a project may work in favour of the open-data movement. The publishing giant Springer Nature made a significant step in this direction by launching the journal Scientific Data, which revolves primarily around datasets.
OA has generated a greater awareness that publishing should not and cannot remain unchanged over time. If it is to guarantee the dissemination of ideas, it must mirror the evolution of scientific enquiry over the centuries, taking into account the increasing importance of interdisciplinary collaboration, large-scale projects producing terabytes of data, and a growing role for computational approaches. It is possible to read a positive message even amid the arguments: be informed, be aware of your rights and duties as a scientist, and do not take for granted models and practices. After all, open-mindedness may lead to transparent, open research.
(1) R. Van Noorden. Open access: The true cost of science publishing. Nature 495:426–429 (2013) (available here).
(2) S. Bosch and K. Henderson. Steps Down the Evolutionary Road: Periodicals Price Survey 2014 (available here).
(3) The Right to Research Coalition. The State of Open Access – pre OpenCon 2015 Webcast, https://www.youtube.com/watch?v=hxGko0czXEo (17:18 for relevant slide).
(7) In 2014, a professor of tropical ecology and Latin American studies quantified the ‘costs’ of his committment to open science: you can read his blog post here.
OA in Oxford, UK
- Open Access Oxford website (a project of the University of Oxford).
- The UK headquarters of CrossRef are based in Oxford.
Collaborative progress: the OA Button
The Open Access Button offers a tool against the ‘paywall’ as well as a way to support and contribute to the development of wide open-access policies.