According to UNESCO's recommendation draft, the term ''Open Science'' is an umbrella concept for several practices that aim at making scientific knowledge publicly available to everyone. Researchers can increase the transparency, verifiability, and reusability of their research work by sharing the collected data as well as the applied methods. This practice facilitates new collaborations with peers and stakeholders outside science (e.g., citizens). The recommendations further encourage collaboration by also paying attention to diversity and inclusiveness. Below you find a collection of definitions for the different practices as suggested by UNESCO. Note: Remarks and comments added by the author of this website are highlighted in italics.
The picture was taken from the UNESCO Open Science brochure.
Open Access is granted if readers receive full and immediate access to scientific outputs, including research articles, data, software, source code, and protocols. Access to these outputs is provided free of charge and in a re-usable way. By giving credit to the original producers, users are given the right to copy, use, and distribute the work publicly as well as to create and re-publish derivative outputs in any medium. The research article and all materials should be published in an online repository that is maintained by an academic institution, scholarly society, government agency, or another non-profit organization. Furthermore, the repository should ensure open access, unrestricted distribution, and long-term archiving of the hosted materials.
Data can be termed Open Data if others can, by giving appropriate credit, re-use as well as re-distribute it. Consequently, the data and databases need to be publicly accessible and made available under an open license. Open Data should be published in a human- and machine-readable format which excludes most proprietary and particularly binary data formats. A further criterion is the publication of the data following the FAIR principles, meaning that the data is
- findable by using a globally unique identifier and describing appropriate and complete metadata,
- accessible with the help of open and standardized communication protocols,
- interoperable using a formal, accessible, shared, and broadly applicable language for knowledge representation, and
- re-usable, which can be achieved by assigning a clear and accessible data usage license.
FAIR data is not necessarily Open Data since FAIR data can meet the requirements while being accessible only to a small group of researchers (see blogpost).
The software can be called Open Source Software if it is made publicly available under an open license allowing others to access, modify, and distribute the software as well as the source code. Similarly, hardware can be indicated as Open Hardware if the design specification of the hardware is shared under an open license allowing others to study, modify, and distribute it.
An Open Science Infrastructures is any kind of digital infrastructure that supports the realization of the Open Science principles. Such infrastructures are essential means to ensure accessibility and re-usability of research outputs by humans and machines. Open Science Infrastructures should be non-profit and ensure long-term and permanent accessibility of the stored materials. Strictly speaking, the popular platforms ResearchGate and Academia.edu are both not in line with the Open Science principles since they are for-profit organizations. It remains unclear what 'long-term' or 'permanent' actually means. Research data management regulations often ask to store the raw data for at least 5-10 years after the project expired. Zenodo retains stored materials for the next 20 years.
Open Evaluation, which is also known as Open Peer Review, means that research articles are assessed in a transparent manner allowing a broader community to comment on the scientific article. While the reviews are publicly accessible, the referees can usually decide whether they would like to reveal their identity or not. Open Peer Review was realized, for example, by Copernicus Publications, who often additionally publish the authors' answers to the reviews.
Learning and teaching materials are open if they are released under an open license giving others permission to access, re-use, adapt, and re-distribute them. Open Education can be also achieved by integrating the teaching of Open Science practices in the curriculum. A further possibility is to make use of Open Source tools instead of licensed software.
To address the needs of the public, science should be open to contributors outside the university, for example, policymakers, practitioners, entrepreneurs, activists, and citizens. In Citizen and participatory science, non-professional scientists frequently conduct research as part of formal research projects. The output of such projects should adhere to Open Science principles as well to ensure effective re-use and benefit to all.
Science should recognize the diversity of knowledge systems, epistemologies, holders, and producers. Key to this practice is non-discrimination, adherence to international human rights, respect for knowledge sovereignty and governance, and the recognition of the rights of knowledge holders. Open Science thus promotes openness to indigenous knowledge systems and all scholarly knowledge and inquiry while also considering principles such as non-discrimination.
Reproducible Research refers to achieving exactly the same results (e.g., tables, figures, numbers) as reported in the paper by using the same source code and data. In Open Reproducible Research, these materials are publicly accessible. Replicable Research refers to coming to similar conclusions based on newly collected data or a newly implemented analysis. Replicability and Reproducibility are essential for scientific work.