Today I want to write about how to practice open science.
I have to admit that I was not really into that topic beforehand. I mean, in some occasions I was – as many scientists (and also non-scientists!) – bothered by the strict rules of big publishers like e.g. Wiley or Elsevier, when trying to download some publications.
For instance, currently I give a lecture about “theoretical ecology” at the University of Jena. To prepare my presentations, I try to read most of the big “classic” ecology papers about the topics I present. However, especially in the 70s and 80s A LOT of important papers were printed in “the American Naturalist” – a journal whose papers I may not access without paying a horrendous amount of money for every single paper o_O. Searching the papers elsewhere is tedious and not always successful.
For me it always felt a bit unfair that we scientists seem to have to pay for both: publishing our papers AND reading them. However, as everyone is expected to publish, I – as most other scientists – just played my part in this stupid game without complaining.
Maybe it is time to change the system for good – therefore I will write about the information I received during a meeting with the open science group of the iDiv. There I learned some things every researcher should know about and I will provide here a summary of the information (and useful links) given.
Why open science?
Building new theories and thus deepening and refining our knowledge only works in a discursive way: Our publications and datasets are not meant as a purpose in itself, but as “building blocks” for new theories and ideas and even as “grindstone” to inspire opposite theories.
If we practice open science, we amplify our collective intelligence
Therefore, both, our publications but also the data (or models) our publications are build upon should be made publicly available. This helps to ease the replication of our results and also improves the credibility of our findings. Therefore anyone should have access to our information.
Furthermore, the research money – often provided by taxpayers – should go directly to science and not to publishers.
Things you need to know to make your science more “open”
1. Am I am allowed to archive my papers online?
First of all you should check, if you are allowed by the journal you are publishing in (or are planning to submit to). Journals differ a lot in what they allow you to do and what not. You check this on SHERPA/RoMEO – just type in the name of the journal and see what is possible. I just tried it with “The American Naturalist”:
This tells us, that for an American Naturalist paper we can archive all: pre-, post- and the publisher’s version of the article, for instance on our personal homepage or a pre-print server (in that case the pre-print version of the paper). However, to put the paper on an open access repository we still have to wait for 12 months after publication.
What are pre- and post-print articles?
Many journals distinguish between pre- and post-print versions of your papers, so be careful what to put where:
- A pre-print is the original version of a scholarly or scientific paper that precedes
- A post-print is the journal article version after it has been peer reviewed and before it is sent off to the journal for publication (before final layout).
- A publisher’s version is the version that is published on the publisher’s website.
You should take advantage of your opportunities, so if you are allowed to archive your paper in any way, do it. Why? Because doing so will – additionally to helping the (scientific) community – increase the visibility and readership of your publications.
2. Where can I put my papers?
Check dissem.in with your name or ORCID to identify publications that you can make available for the public. Then self-archive your papers in one of those places:
- your private homepage
- Institutional or subject-based repositories for papers (e.g. arxiv, bioRxiv,…)
- You can also send your author manuscripts and a letter of agreement to email@example.com (they will check the requirements and help you publishing your articles)
By the way: it is also useful to archive your pre-prints to get a kind of “time-stamp” for your ideas.
Many scientists are used to put their papers on ResearchGate or Academia.edu – which both are literally like a researcher’s “facebook”. They are somehow convenient to use, BUT also have the disadvantage of being commercial sites which track your actions (see here for further information). Furthermore, putting your papers there comes with the risk of sudden removal: read more about the removal of papers from ResearchGate or Academia.edu. (See also the comment on this article).
What is a repository?
A destination and archive for collecting, preserving, and disseminating digital copies
of the research output, e.g. data, paper. A comprehensive and authoritative list of
institutional and subject-based repositories is provided by opendoar.
What is ORCID?
The Open Researcher and Contributors ID repository. Academics, researchers and
contributors can register for a unique ID with ORCID to reliably identify
their research. ORCID connects all important platforms and identifiers, e.g., Scopus,
ResearcherID/WoS, Crossref, DataCite etc, see orcid.org/content/collect-connect.
3. What about my data and sourcecode?
Sharing both, lab/fieldwork AND virtual data is an essential tool for replication, validation and extension of empirical results. However, the status quo in empirical research is not to share data (more information about this problem can be found here: paper about data sharing).
Proceedings in the lab should not be written exclusively in the lab book, but also be digitalized, this helps to store and also understand your data for a longer period of time. Data should be stored in one of the several public or institute repositories (see list at the bottom of the article). It is best, if you also provide the scripts you used/created to read and analyze the data.
One disadvantage of many models is that they are often designed to answer one single research question. However, for the purpose of one or two publications alone a thorough model development does not seem cost-efficient. Sharing a model with the scientific community can motivate to execute the tests necessary and thus improve the model quality and the reliability of results. Once online, the model can be re-used for different research questions or even adapted to other systems. Furthermore, modelling decisions and model implementation can be discussed within the community and this
can inspire important modifications.
Therefore it is always a good idea to archive your model source code online. To increase the chances that your model will be understood and used by the community, consider the following things:
- write neat source code with many comments
- use modular blocks of code which can be switched on or off when
- provide a complete description of the model which is easy to understand, like e.g. the ODD protocol or even a TRACE document.
- perform analyses to provide a better understanding of the possibilities and limits (path dependence: “robustness analysis“, parameter decisions: “sensitivity analysis“) of your model and provide also their documentation online
- if helpful, provide also programs developed to analyze the data (to perform sensitivity analysis, to analyze the outcomes statistically, make figures etc.)
All useful links at a glance:
- creativecommons – Information about licenses (-> Do I want to allow commercial use or not? Also: Do I want to allow derivative works or not?).
- https://doaj.org – Directory of Open Access Journals.
- Thinkchecksubmit.org – Check the quality of an open access journal (is it safe to submit my paper there?)
- ORCID – Get your unique researcher’s ID
- altmetric – Check for the impact of your paper (other than just citations), e.g. in social networks.
- Open Science Monitor – General information about open science and sharing data, figures, papers etc. Provided by the European commission.
- SHERPA/RoMEO – Check what I am allowed to do with my paper (e.g. about publishing a pre- or post-print online) before or after publishing.
- dissem.in – This site helps you to check how many of your articles are already available for the public and which articles still need to be archived in a public place.
- arxiv – An open access paper repository for Physics, Mathematics, Computer Science, Quantitative Biology, Quantitative Finance, Statistics, Electrical Engineering and Systems Science and Economics.
- bioRxiv – An open access pre-print server especially for biology.
- git – Free and open source version control for your software.
- github – A web-based git version control repository hosting service.
- See also my former post on git github: git and github
- DataCite – A non profit organization which provides persistent identifiers (DOI) for research data for safer citing.
- Research Data Alliance – Build the social and technical bridges to enable open data sharing.
- re3data.org – A global registry of research data repositories.
- pangaea – A data publisher for Earth & Environmental Science.
Edit (01/2018): There are signs of hope: universities win temporary journal access after refusing to pay fees