Designing a project with Open Science principles from the start

Blogger picture

Planning a new research or restoration project with openness in-mind from the start ensures you’ll have the funding and time to broadly share your results.

Open Science is a multi-headed hydra of concepts and jargon that can mean many different things to different people and organizations. There are basically two main pillars of Open Science: reducing barriers to access and increasing transparency. For a quick starter to the many facets of Open Science, the Swiss National Science Foundation’s Horizons magazine covered this topic in their September 2016 issue (PDF).

This blog will address some of the core tenets of Open Science, why they are important, and how to plan a project with Open Science principles from the start.

TL;DR Summary

  • Plan for how much time and money it will cost you to publish your results and data such that it is freely available right from the start
  • Ask for these costs in your grant proposals
  • Publishing both your data and final research products openly increases trust
  • Provide good documentation and metadata so that others may reuse and remix your data, further increasing your citation rate
  • Preregister your research project to avoid “scooping” and to further increase trust

Reducing barriers to access

Open Access is perhaps the best-known aspect of Open Science: not only should published scientific works be freely available to read, share, and remix – but the underlying data, too. The vast majority of the planet does not have access to pay-walled or subscription-only journals: especially users in developing countries and nonprofit organizations (like MARE, which runs OpenChannels). This is especially true for ocean conservation, where nonprofits, NGOs, and developing nations typically do most of the actual conservation work – groups that overall are not paying to read pay-walled academic journal articles.

Publishing Open Access (OA) often gets a bad reputation in the academic world for being cost-prohibitive. Rather than relying on university subscriptions and per-user download fees to subsidize the cost of publishing, OA journals typically require the author to pay the full “true cost” of publishing their work upfront. But publishing behind a pay-wall has many hidden costs that authors largely ignore.

Remember: once you publish in a traditional pay-walled journal, you give away your intellectual property (IP) rights and copyright for that publication. As noted on Elsevier’s website, “authors transfer copyright to the publisher as part of a journal publishing agreement.” In short, if you do not publish Open Access, that means that you cannot do any of the following without permission from the publisher:

Even if you are not publishing OA, there are clearly costs to share your work, should you or a non-author colleague need to pay royalties to do any of the above. (Note: Elsevier and many other publishers use the Copyright Clearance Center’s RightsLink system to collect royalties and distribute rights. Prices vary depending on the type of rights and use requested. For example, a nonprofit wishing to reuse a single figure in a presentation would need to pay US $44.85 to Elsevier to do so. A for-profit entity would need to pay $431.25 for the same rights.)

Archiving and sharing preprints of pay-walled journal articles is a common way to reduce barriers to access. However, publishers often place restrictions on where you can share the submitted manuscript. Many publishers require that preprints only be shared on the author’s personal website or blog for a period of 12-24 months: though, after that time, the preprint may be shared in other repositories. Be sure to check SHERPA/RoMEO for the restrictions placed by your publisher. If you can only share preprints on your personal website or blog, be sure to calculate the time and cost to create and/or update your website. You may be able to include small costs like this in your grant application.

Planning from the start will make sure you have both the funds and the time you need to share your work properly.

Increasing transparency

Transparency is key, especially for government-funded science. Making your data publicly accessible online can increase trust with both the data and derived results. Furthermore, public data allows anyone in the world to run additional analyses on your work. There are hordes of graduate and undergraduate students out there looking for data to analyze. Your raw data could help plenty of students get their feet wet in marine conservation! There are plenty of other benefits, too, least of all is that you will not have to keep responding to email requests for your data if it is publicly available online.

Say you want to reduce barriers for the ocean conservation community to access your published work and data, and you want to be transparent about what you are doing: where do you start?

1. Ask for the costs of publishing your results when seeking funding

The first step on your planning process path is to determine the costs of publishing your results. What journals do you want to publish to? Do you want, or need, to publish in a journal in the first place? Are there extra costs for making your work OA if you publish to a traditional pay-walled journal?

Fees vary widely from journal-to-journal for publishing Open Access. For a list of Elsevier’s Open Access publication charges as of 1 April 2017, click here (PDF). Elsevier’s Marine Policy journal charges US $2,300 to publish a paper Open Access. Wiley, another publisher, charges between $1,500 and $5,200. PLOS ONE charges $1,495. PeerJ charges $1,095; though, PeerJ Preprints is free of charge.

If you want you results to be used (especially in marine conservation, a field dominated by NGOs without subscription access to pay-walled journals) $1,000 to $2,500 is hardly an obstacle to include in a grant application.

But do you need to publish in a traditional journal in the first place? For authors without archaic tenure committees to worry about, publishing in a journal is hardly necessary. Preprint services like PeerJ Preprints, OSF Preprints, or your institution’s publication repository may be all that is needed. Any of those services would be completely free to use, both for you (the author) and your audience. We (OpenChannels) can also publish your paper for free in our Literature Library, if you wish.

2. Publish your work in a pay-walled journal? Make sure you share the pre- or post-print online

Even if you publish your work in a pay-walled journal, you retain some limited rights on the submitted manuscript, also known as the “preprint.” Akin to the preprint is the “postprint” or the peer-reviewed and edited manuscript that has not undergone full typesetting.

SHERPA/RoMEO is an excellent resource for determining what rights you have for sharing your work if you have otherwise given away the copyright to the publisher. For quick reference, here are Elsevier’s default policies for non-Open Access content:

  • The authors may share the preprint on any website.
  • The postprint may be shared on the author’s personal website immediately.
  • The postprint may only be shared in Open Access repositories after a 12- to 48-month embargo period.
  • The publisher’s version/PDF may not be shared anywhere, ever.
  • The author most link both preprints and postprints to the publisher’s pay-walled copy with a DOI link.
  • The author must release their postprint with a Creative Commons Attribution Non-Commercial No Derivatives license.

Many universities and funding institutions have Open Access mandates. Be sure you know the rules for your own institution and funders! Depending on those policies, you may be prevented from publishing in journals which place restrictions on archiving postprints. Know your responsibilities and the rights you sign away when publishing. And know that publishing is meant to be a discussion, not a one-way street to complying with default publisher-policies. If your institution requires that your peer-reviewed work be archived on its repository, but the publisher traditionally disallows such use, be sure to let the publisher know that this policy will prevent you from publishing with them. Remember that the journals require your work to make a profit! If they’re going to lose you, and thus lose the opportunity to profit off your intellectual property, they’re likely going to make exceptions to their default policies. (I’ve heard of this happening plenty of times.)

If you don’t have a personal website or blog to share your preprints/postprints, it’s not a bad idea to factor in the costs to create and host one in your grant proposal – particularly if that website will be hosting information about the project for the long-term. If you are based at a university, chances are you can create a website for free using the university’s IT infrastructure. There are plenty of free or cheap, simple to complex, website-hosting platforms out there. OpenChannels is hosted by Pantheon, which allows both Drupal and WordPress-based sites. WordPress lets you create free websites, given that you must use a subdomain. There is also SquareSpace, which every podcast listener knows quite well. Regardless of what service or type of website you decide to create, planning ahead to know how much it will cost to run and create will make sharing your results much easier when the time comes!

3. Share your results and your data

There are a number repositories for sharing data, some free and others not. Some of the most common are integrated with journals: Figshare and Dryad. The Open Science Framework (hosted by the Center for Open Science, the creators of OSF Preprints) is a relative newcomer to the field. The Open Science Framework, for one, allows you to create free public Project sites that integrate file- and data-sharing and a wiki, and gives you a free DOI link to easily share and cite your project.

Sharing your data requires that your data is understandable to users other than yourself. Good metadata must be understandable both by humans, but also computers; therefore, metadata has a certain structure depending on the data you are describing.

Having good documentation and metadata is key. Ideally, you should post your data alongside your submitted manuscript or preprint. Be sure to plan for creating a clean, well-organized, well-described dataset to share with the public from the onset of your project. If your data is a mess, it will not get used and could even undermine trust in your results.

Sharing your data will obviously take time: make sure you include time estimates for these tasks, as well as any estimated archiving or consulting fees for doing so in your research proposal.

4. Working on a public project? You should especially share your hypothesis and research goals upfront

Preregistration is another facet of Open Science designed to increase transparency and trust. The Center for Open Science defines preregistration as “[separating] hypothesis testing from hypothesis generating research.” By publicly stating your hypothesis before gathering data, analyzing it, and publishing it, you can increase trust by showing the world that you did not alter your research goals to get a publishable result.

There’s another often-overlooked aspect of preregistration: avoiding getting “scooped.” Researchers doing a simple Google search can find preregistered projects, thus knowing upfront if there is another team already working on the same idea. Taking the pessimistic viewpoint, this will allow “good scientists” who preregister their results to protect themselves from “bad scientists” who try to win the race to publication. The “good scientists” in this case can simply cite the date they registered their project to provide support that they were “scooped.”

If you are looking for a specific service to preregister your project, beyond simply posting this information on your own website or blog, the Center for Open Science is currently running a Preregistration Challenge.