Thing 17: Open Access and Open Research

Richard White, Manager, Copyright & Open Access, University of Otago

(Art: John Megahan)

Why begin this Thing with a picture of the Haast’s Eagle, the largest known eagle to have ever existed? Perhaps it’s a metaphor for sharp-taloned mega-publishers preying on poor, flightless Moa-researchers? No! First, it’s a cool picture and, second, it was first published in an open access research article in PLoS in 2005. A quick reverse image search on Google shows its footprint (talon-print?) on around 1000 pages on the Internet, from Wikipedia, to Science education sites, to museum sites, to online newspapers, to hundreds of references on Twitter, Pinterest and Reddit, and National Geographic. This doesn’t include the countless science education videos I’ve personally seen it in, like an adapted version in a video by PBS Eons with 1 million views.

On one level, Open access (OA) is about making work available with licences that are not “all rights reserved” so that anyone can access, copy, adapt, translate, build upon it, etc. This is the legal basis that enables something broader, more of a philosophy about how to think about the way we work in academia: not open access but open research.

Increasingly we are seeing funding agencies and governments mandate that the work they fund is open access. This is a policy-driven approach based on the principle that the public paid for it so they should be able to read about the results and use them. At that policy level things are developing fast but in this Thing we will keep our focus on you as individual researchers.

Thinking about it from this point of view, at its heart open research is about:

• making my work inclusive rather than exclusive. When my work is not behind a paywall, it can be used by government agencies, policy makers, the media, businesses, innovators, the general public, practitioners, and teachers and researchers from schools or institutions that can’t afford subscriptions;

• including marginalised communities or groups or people in developing countries;

• being transparent, so my work can be freely critiqued and tested for reproducibility; and

• increasing and broadening the people who read my work, and thus enhancing the potential for impact and to be cited, not to mention potentially improving it and developing it in ways I did not envisage.

It’s not rocket science to suggest that making your work accessible will mean more people read it; more people reading it will mean more researchers using and citing it; making it reusable will increase its impact even more. The research backs this up and it’s generally accepted that, all things being equal, making your work open leads to increased citations (see this article for example).

It’s online so it’s open access! …right??

Making your work free-to-read is one step on the way to open research but it’s just a start. Let’s look at a couple of examples that demonstrate different approaches.

1. The Johns Hopkins Coronavirus Resource Centre has been a central repository of authoritative data in the Covid-19 pandemic. If you look at the footer of their pages you will see “JHU.edu Copyright © 2021 by Johns Hopkins University & Medicine. All rights reserved.” This means I can access the site through my browser; I can use their stats or maps for my own research; I can potentially reproduce them if I am reporting or critiquing them in some way; but because it is all rights reserved I couldn’t copy their resources in other ways. I don’t have access to the underlying data to analyse it myself for things they haven’t been interested in or to combine with other datasets.

2. Contrast that with this work from 2020 about governments’ responses to COVID-19 all around the world. It’s an article that has a Creative Commons Attribution-only licence and that describes an open dataset hosted in a GitHub repository, reusable by anyone. I can use any element of the article or data in my own work without permission; I can use the data and write up my own analyses and reports. (Copyright nerd note: if you’re wondering how news agencies use the Johns Hopkins charts and such, fair dealing provides for news reporting as a permitted use; it’s not because it’s free-to-read on the web).

These are large-scale examples. But the same applies to anything: if I was writing about my research on a massive, extinct New Zealand eagle that was big enough to take down one of the largest birds that has ever existed then I could illustrate my writing with the awesome picture published in an OA journal at the top of this Thing without seeking permission. I couldn’t use this picture because it is commercially-licensed. They’re both online but only one can be used without seeking permission or paying a fee.

It’s important to say both approaches are fine: any artist or creator has the right to choose how their work may be accessed and used. We will all know examples within our own research organisations where intellectual property is regarded as very valuable and tightly controlled, whether that’s a patentable drug formula or years of data collected by a major study. In the context of research, however, the point is that open approaches are designed to speed up the use of resources by others, which in turn increases readership, citations and the advancement of knowledge.

A physical paywall? (Photo by Kenny Eliason on Unsplash)

The machinery behind OA: open access licences

There are many different flavours of open access licence but the most common ones you will see in the context of research are those provided by Creative Commons.

Task 1: Watch a 5-minute video on how Creative Commons licences work, “Creative Commons Kiwi”

This 23 Things site has a Creative Commons licence – and in fact this blog contains elements based on a previous 23 Things that had an open licence made by completely different people. We don’t have to work out who to ask for permission or if any of our uses would be covered by copyright exceptions [link to last year’s copyright Thing]: permission has been given ahead of time by the creators using an open access licence. Read more on this in last year’s Thing on Copyright.

There are other licences you might come across, especially if you’re looking at data (e.g. CC0) or software (e.g. GNU Public Licence) but there’s too many to go into here. Whichever one you come across there is likely to be very good information available on the web as to what they mean.

“Open access is a good thing but it’s so expensive”

When open access journals first started becoming more prevalent in the 2000s – and yes Physicists and Mathematicians, I know you’ve been doing open access since the advent of the Internet via arxiv.org – open access attracted a lot of negative attention due to the unethical practices of predatory publishers. These were outfits that solicited submissions from authors and pretended to have peer review and editorial standards but in fact just charged you a fee to publish your article pretty much as you’d sent it to them in the first place. As hopefully everyone has realised by 2023, the quality of a publication has nothing to do with its business model. There are OA journals whose publications are of lower quality and/or impact, just as there are the same for non-open journals. There are some brilliant journals that charge nothing for publication or for access. Put this idea of Open Access=Low Quality out of your mind and just assess any publication venue on its merits. Think, Check, Submit provides a simple checklist that’s useful for this.

Cost can genuinely be an issue for researchers. APCs (Article Processing Charge) were introduced by publishers as a way of making work open access. Of course, publishing has always cost money, it’s just that researchers started noticing this because they were being asked to pay, instead of publishers charging subscriptions to libraries. If you do want to make your publication open and that will involve an APC being paid, don’t forget to check out a publisher’s policy on APC waivers.

As mentioned above we’re seeing more funders mandating open access outputs from work they fund but also offering financial support for this; we’re also seeing publishers and universities moving to a different kind of subscription deal where libraries still pay the licence fees but this comes with a certain number of ‘free’ APCs (i.e. free to researchers, not actually free). If you work at a big university you’ve likely heard people talking about these as “Read and publish” or “Transformative” deals with publishers.

You don’t have to pay to make your work open

APCs are not the only way to make your work open. There are thousands of open access journals that don’t charge APCs: the Directory of Open Access Journals lists nearly 13,000 as I write. Many people are using pre-prints (this has exploded in health disciplines due to the pandemic). And more and more we are seeing people share more than just chapters or articles, making data, software, notes, tools, videos, etc. available on sites like Zenodo, Figshare, and Github. In some publications peer review is open now, in the interests of transparency and rewarding people for their review work.

And if you take one thing away from this Thing it should be that most publishers allow you to deposit an accepted manuscript version of research outputs in your institutional research repository. This usually has an embargo period but not always.

Task 2: if you have published in a journal that is paywalled, go to the Sherpa/Romeo database to check on the publisher’s policy. Type in the name of the journal and it will tell you if/when you can deposit an accepted manuscript. (They have help videos).

At my own organisation we have found that, about half of our research is free-to-read, if we deposited everything we could put in our research repository then 90% of our research outputs could be open). Many of you can probably make your prior publications open access – check out Sherpa/Romeo and talk to your library!

The future of open access

For the first 20 years or so, debates about open access have focused pretty much on how traditional means of publication could be transposed into a digital, free-to-read medium. As you’re hopefully grasping from the above things are changing fast in the broader sense of how researchers communicate their work to others. We’re also seeing more thought being put into how to align indigenous knowledge systems – where ideas about knowledge ownership and sharing tend to be very different – into the push for open research. In terms of openness, there will always be things that need to be protected but a current catch-cry that rings true for many is that we should be striving to make our work “as open as possible but as closed as necessary.”

Task 3: read this opinion piece Open science: after the COVID-19 pandemic there can be no return to closed working

Exploring further

• Some would say we have already moved past “open access” as being a useful concept and that we should be thinking more along the lines proposed by the F.A.I.R. principles, that is research should be Findable, Accessible, Interoperable and Reusable. Read more here: www.go-fair.org/fair-principles/

• Go to www.leidenranking.com/ranking/2020/list Change the indicator drop down on the left to ‘Open Access’. This gives you a list of universities with the proportion of work that is open access for each. Find your own organisation and see how it compares. (Mine is ranked 561 in the world for openness at 47%. Some have over 90%.)