FAIR Data in Social Sciences and Humanities: a Triple Open Science Training Session
The FAIR principles have become a key element in the context of Open Science more broadly, and of EOSC building more specifically. On September 15th 2021, a Triple Project Training Session1 was dedicated to FAIR data in SSH. Elena Giglia (Università di Torino) presented the session.
The presentation by Elena Giglia was organized around four main topics: the definition of data for Social Sciences and Humanities, the definition of FAIR principles, the existing tools to make data FAIR, and how to create a Data Management Plan (DMP). The session turned into a real crash course on how to make your data FAIR in SSH.
Before exploring the topics, some relevant references in the field were highlighted:
- The ALLEA Report, Sustainable and FAIR data Sharing in the Humanities, a practical guide that provides key recommendations to make digital data in the humanities in line with the FAIR principles.
- CO-OPERAS reports on a series of workshops held in national languages that were aimed at gathering the views and needs of SSH researchers regarding the FAIRification of their data and publications.
- The Report of Work Package 6 from OPERAS-P, “Future of Scholarly Communication . Forging an inclusive and innovative research infrastructure for scholarly communication in Social Sciences and Humanities”, especially the section 6.3 that focuses on quality assessment.
- The DARIAH Pathfinder to Data Management Best Practices in the Humanities, by Erzsébet Tóth-Czifra, that presents an inventory of different resources such as tools, videos, short articles and other training materials to make research and data management workflows as open as possible.
- PARTHENOS Guidelines to FAIRify data management and make data reusable, a guide that focuses on emerging trends and best practice in data management, quality assessment.
1. The definition of data for Social Sciences and Humanities
The section started with the question: how to define data in the context of SSH?
CO-OPERAS workshops reached the SSH community aiming at, among other things, defining data in the context of this domain. It was possible to agree on a common understanding of some aspects. The most important of them is that the idea that the definition of data depends on a given context is also key. In other words, to define data it is necessary to take into account the context of data production. In the SSH, besides the discipline, the specific theory, methodology, and technical apparatus are significant if not determinant to the production of data. Furthermore, as data are usually seen as a construct in SSH, they are never raw and the notions of interpretation and subjectivity often come into play. Figure 1 sums up some definitions of data.
Figure 1. Definition of data in the SSH
Author: Karla Avanço
According to the ALLEA report, “we could then define data in the humanities broadly as all materials and assets scholars collect, generate and use during all stages of the research cycle”.
The speaker also emphasized that in the framework of open as possible, as closed as necessary being open is not enough because to be useful, data should be FAIR from the beginning. If data are not FAIR from the start, they could be misused or misinterpreted. Also, it can be time-consuming to make them FAIR afterwards. Finally, besides being FAIR, data should as well be adequately managed, which would benefit primarily the researcher as the whole research process results more effective.
According to this article by Ultike Wuttke (2019), “the responsible handling of research data and FAIR publication of research data is becoming an integral part of Good scientific practice.” Regarding the definition of data, he states that “(…) humanities researchers tend to be ambivalent about the concept of ‘data’. (…) In short, they just don’t use the word “data”, but talk about “sources”, “research materials” etc., which leads to the fact that the whole “data talk” doesn’t appeal to them.” Also, research data in humanities are rather heterogeneous, idiosyncratic, and complex.
2. The definition of FAIR principles
The second topic explored during the training session was the definition of the FAIR principles themselves, which this very blog has also done!
Figure 2. FAIR means machine readable.
Source: Elena Giglia(2021)
This section, however, did not focus on simply introducing the principles. The speaker not only presented some helpful resources, but also introduced important aspects regarding the implementation of the FAIR principles: the distinction between FAIR and open, the importance of having a data steward, and the difficulty of FAIRifying SSH.
About FAIR and Open
Elena Giglia signaled the lasting overlap between FAIR and Open; FAIR data, however, can be not-open. As an illustration, she presented the case of a researcher working with refugees. If the data are open, the participants of the research could be in danger. On the other hand, if the data are FAIR, another researcher can know the data exist, where they can be found, and under what conditions they could be reused (such as under a non-disclosure agreement). This blog has already talked about the relation between Open and FAIR here.
About the different players and the role of data stewards
FAIR principles work as guidelines that should be adapted to different domains, disciplines, and communities. Besides, different actors are implicated in the implementation of FAIR principles, and different FAIR data skills are necessary, according to the role of the player (research funders, institutions, researchers, policy-makers, data services providers, publishers, data steward, data scientist, educators and trainers, etc.). Among these roles, the data steward is crucial because it is the player that can act in three different axes: policies, research, and infrastructure. A data steward should have strong domain-specific knowledge, understand information technology systems and structures, and ensure quality and compliance with rules and regulations. Some of the expected skills of a data steward were presented at the webinar Challenges in professionalising data stewardship.
The speaker emphasized the need for funders to invest in data stewardship, how urgent it is to hire them, and that they have to be part of a research infrastructure, not only part of a project.
and education, by Mijke Jetten (2021).
“This report was commissioned by ZonMw on behalf of the National Programme Open Science (NPOS) to provide arguments for urgent decisions and activities to ensure adequate data steward capacity in the Netherlands to realise its ambitions with respect to Open Science. (…) The report gives an analysis of the current situation in the Netherlands with regard to data stewardship competences, education and training, and draws attention to the urgent need for a nationally coordinated action by the main (…) The recommendations presented in this report tackle the challenges and needs that are experienced at present, both at local and at national level, and thus meet the ambitions that are anticipated in the transition to Open Science.”
About FAIR and SSH
Regarding SSH, the speaker asked a couple of relevant questions. In which steps should the principles be applied? In which language should the metadata be, in national languages or in English? These questions are important because the application of the FAIR Principles must preserve the specificities of the ways we do research.
Besides, awareness and willingness is not enough. There are other aspects that we have to take into account:
- Services and tools must be sustainable;
- The SSH still need a registry of existing tools;
- The issue of incentives and rewards for the FAIRification must be addressed (making FAIR data represents a huge effort and it is time-consuming)
- There is still a lack of data and metadata skills among researchers
The speaker also pointed out that the difficulty of FAIRifying data in SSH comes from the fact that data are collected, they are not produced. When data come from cultural heritage, the ownership is not known. Moreover, the level of metadata and the description of the research is still uneven.
2.1 Detailing the FAIR principles
During the section, the speaker also detailed the FAIR Principles, specifying what each one of them implies and suggesting resources to better understand and implement them.
For the data to be findable, we need metadata, we need to describe resources in the richest possible way.
Accessible is not a synonym of open. Accessible data can be reserved, restricted, or embargoed.
- Access categories on the CESSDA Data Expert Guide
- Data repositories: Zenodo (is free and allows the creation of communities), Dataverse, Dryad, Figshare
- Data repository specific for the SSH: Humanity Commons, a community for the humanities that has a repository
To find the most suitable repository for a specific discipline, one can check the global registry of research repositories, re3data. If possible, one should try to converge to the repository the community is using.
To make data accessible, having an open format is also crucial. In general terms, the best formats to assure long-term sustainability and accessibility are frequently used, have open specifications, are independent of specific software, developers, and vendors.
For instance, Excel is not accessible as it is a proprietary software. If your data is in this format, you should convert it to a preferred format such as csv. The Data Archiving and Networked Services (DANS) prepared a list of preferred formats.
- The National Archives of the United States also have a list of preferred formats for data.
- The FAIR Cookbook gives the recipe of how to convert from proprietary to open formats.
Interoperability requires standards. To understand the notion of standards, there are some resources that can help.
- ALLEA recommendations: normalise, use authority files
- What are standards?, by Parthenos.
- Huma-Num’s Standardisation Survival Kit.
- FAIRsharing: you can search by discipline, by domain, community driven
- Existing Formal Ontologies, by Pathenos
An example of open methodologie is OntoME,an ontology management environment for historians, developed by Francesco Beretta. The methodology presented in the article “A challenge for historical research: Making data FAIR using a collaborative ontology management environment (OntoME)” (Beretta, 2021) can be used by other disciplines.
To be used, the data must be well documented, the processes used to collect and process data must be described. Documentation is necessary to avoid misinterpretation and misuse of the data.
The speaker suggested the use of Notebooks for the documentation.
- “Open laboratory notebooks: good for science, good for society, good for scientists”, a paper by Schapira et al. (2019), “highlight the benefits of open lab notebooks to science, society and scientists, and discuss the challenges that this nascent movement is facing”.
- Here there is an example of using ELN, Electronic Laboratory Notebook, in the Humanities
The use of open notebooks raises the question: Do we still need journals to share our research if we can use notebooks?
To be reusable, the data also need an adequate license. The speaker reminded the public that raw data is not protected by copyright, that protects creativity.
The speaker also highlighted the difference between copyright and Sui generis. They last different amounts of time but they are both an economic way to protect the effort. Copyright protects the structure, selection or arrangement of the content, not the data itself, whereas the Sui generis database right protects the substantial effort in obtaining data.
The speaker claimed that the only possible license for data is CC0: “By applying CC0 to your data you enable everyone to freely reuse your data as they see fit by waiving (giving up) your copyright and related rights in that data” (Fact Sheet by Creative Commons & Open Science).
- Licence differentiator: a tool that helps its users understand their own preferences in relation to free and open source software licenses.
To learn more about the legal uncertainty surrounding data in humanities, check the book “Digital Technology and the Practices of Humanities Research”, by Jennifer Edmond (2019).
Access to Scientific Information and knowledge: a matter of democracy, by Ludovica Paseri (2020).
3 guides by OpenAIRE, from the Toolkit for researchers on legal issues, by Margoni and Tsiavos (2018):
- How do I license my research data?
- How do I know if my research data is protected?
- Can I use someone else’s research data?
3. Data Management Plan
Figure 3: The data management plan.
Source: Elena Giglia (2021)
After detailing the FAIR principles and presenting related resources, the speaker raised the following question: where should we put all this information? The answer: a data management plan (DMP). The most important aspect to retain here is that it is not a matter of how to design the DMP, it’s about learning how to manage FAIR data. The DMP is a responsible and powerful tool for researchers. Some practical advice on how to create a DMP:
- be synthetic and specific = use bullet points, tables;
- do not copy paste = every dataset is different, every research has its particularities;
- don’t be generic = it’s useless;
- only state on the DMP what will be done.
Some resources regarding the DMP:
- 10 tips for writing a data management plan
- Practical Guide to the International Alignment of Research Data Management – Extended Edition: It is an extended version of a guide on how to properly organize and preserve research data that tagerts organisations, scientific communities and researchers. The guide presents core requirements for DMPs, criteria for the selection of trustworthy repositories, and guidance for researchers to comply with organisational requirements.This new version features a rubric for the evaluation of a DMP.
- Data Management Plan for the Humanities, by UCLA Library
- Adapt your data management Plan, by CESSDA, works well for people who are inexperienced in making a DMP.
- The Risk of Losing the Thick Description: Data Management Challenges Faced by the Arts and Humanities in the Evolving FAIR Data Ecosystem (Tóth-Czifra, 2019).
- DMPOnline is is a tool that can be customised.It allows you to edit, share, download, copy or remove. They also have a list of useful resources on DMP.
To go further
Turning FAIR into reality: Final report and action plan from the European Commission expert group on FAIR data. The FAIR Data Expert Group addressed a wide range of changes necessary to “turn FAIR data into reality”. The resulting report looked for existing practices in certain fields to learn “from those research areas that have already developed standards, international agreements and infrastructure to enable FAIR”. The report also recommends and specifies actions for different stakeholder groups in order to make possible the necessary changes.
Explanation of the FAIR Data Principles, by the Swiss National Science Foundation (SNSF). The document consists of a table containing the explanation of each FAIR Principle, the researcher’s responsibility regarding each principle, and the requirements to be fulfilled by the repository.
Six Recommendations for implementation of FAIR practice by the FAIR in practice task force of the European open science cloud FAIR working group. The report “analyses the state of FAIR practices within diverse research communities and FAIR-related policies in different countries and offers six practical recommendations on how FAIR can be turned into practice. These recommendations are aimed primarily at decision making entities of the European Open Science Cloud (EOSC), as well as research funders”.
Digital skills for FAIR and Open Science: Report from the EOSC Executive Board Skills and Training Working Group “are a cornerstone of the European Open Science Cloud (EOSC)’s operations and future. (…) The EOSC Skills and Training Working Group (WG) was formed in 2020 to identify a framework for building competence and capabilities for EOSC”. The WG focused on four priority areas: 1. Developing the next generation of FAIR and open science professionals; 2. Collaborating to enhance digital skills for FAIR and open science in Europe; 3. Building a trusted and long-lasting and trusted knowledge hub of learning and training resources and related tools; 4. Influencing national open science policy for skills by supporting strategic leaders.
FAIR Data Training, by the Australian National Data Service.
SSHOC Training, a series by the Social Science and Humanities Open Cloud.
You could start with the SSh Training Discovery Toolkit. It is a resource for different actors (data stewards, service providers, researchers and trainers) in the fields of SSF and it provides an inventory of training materials. SSHOC also offers an active and ongoing schedule of online and in-person workshops and webinars.
Manage, Improve And Open Up Your Research And Data, by Parthenos.
Data Management Expert Guide, by CESSDA.
The FAIRification workflow allows to FAIRify data, software, methodology, etc. Every block of the research cycle can be FAIRified and the FAIRness of the data can be assessed. FAIRification is a step by step process. The speaker highlighted once more that the best way to FAIRify the data is by having a multidisciplinary team led by a data steward.
The paper A Generic Workflow for the Data FAIRification Process, by Jacobsen et al (2020), describes “a generic step-by-step FAIRification workflow to be performed in a multidisciplinary team guided by FAIR data stewards”.
The FAIR Cookbook is a tool that “provides recipes for you to learn: how to FAIRify datasets, the levels and indicators of FAIRness, the maturity model, the technologies, the tools and the standards available, as well as the skills required, and the challenges, to achieve and improve FAIRness”.
FAIRdata Forum: the forum was initiated by the FAIRsFAIR “Fostering FAIR Data Practices in Europe” project which has received funding from the European Union’s Horizon 2020 project call. It supports “the FAIR Competence Centre as an open space for users (individuals, institutions, organisations) to post queries, answers, comments and events notifications pertaining to making data FAIR.”
- FAIR assessment and FAIR maturity
To assess the level of FAIRness of a dataset, a software, etc. means to look at the data from a technical point of view. The goal is to be as FAIR as possible since FAIRness is a process.
FAIR data self-assessment tool, by the Australian National Sata Services (ANDS), enables the user to assess the ‘FAIRness’ of a dataset and determine how to enhance it. This tool is useful to start asking the right questions about the data but it is human-readable and hence subjective.
FAIR-Aware is an online tool developed by the FAIRsFAIR project. It is disciplinary-agnostic, but it can be adapted to the use by different scientific communities. It guides the researcher through the assessment process and helps them to better understand the FAIR Principles and how making data FAIR can increase the potential value and impact of research data. This tool is also human-readable. An advantage of this tool is that if the answer for a question is negative, a short and practical fact sheet about the topic pops up.
Evaluating FAIR maturity through a scalable, automated, community-governed framework. The article by Wilkinson et al. (2019), proposes “a scalable, automatable framework to evaluate digital resources that encompasses measurable indicators, open source tools, and participation guidelines, which come together to accommodate domain relevant community-defined FAIR assessments”. The framework includes maturity indicators, compliance tests, and an Evaluator.
The tool FAIR Evaluation Services is a more objective tool. The machine tests if the data are FAIR-compliant.
F-UJI Automated FAIR Data Assessment Tool was developed by FAIRsFAIR. The tool adheres to existing web standards and PID resolution services best practices and utilises external registries and resources. The tool tests the FAIRness of the dataset using the persistent identifier of the resource.
- Tools for SSH
The DARIAH Gateway is a platform that provides access to various digital applications and services for the Arts & Humanities researchers.
OpenMethods: Highlighting Digital Humanities Methods and Tools is a DARIAH initiative created by the Humanities at Scale project in cooperation with OPERAS. Humanities at Scale is a H2020-project of the DARIAH ERIC (Digital Research Infrastructure for the Arts and Humanities). The OpenMethods platform highlights curated content about Digital Humanities Methods and Tools, an important area currently underrepresented in the DH peer-reviewed literature especially, but not exclusively, from a European perspective.
The ARIADNEplus project is an extension of the previous ARIADNE Integrating Activity for archaeological data infrastructures. ARIADNEplus extends and supports the research community and further develops the relationships with key stakeholders such as European archaeological associations, researchers, heritage professionals, national heritage and agencies.
As FAIR are principles, not standards, each community should implement the principles in their community, discipline.
FAIR Principles for Library, Archive and Museum Collections: A proposal for standards for reusable collections, by Koster and Woutersen-Windhouwer (2018). This paper proposes a set of guidelines and best practices to facilitate the process of making heritage collections reusable”.
This blog also recommends the publication FAIRifying a scholarly publishing service: Elements for a toolkit, by Avanço and Gingold (2021) . The paper aims at presenting the methodology employed by a SSH-dedicated publishing system to assess the level of FAIRness of its services and the resulting recommendations.
Check past and next Triple Training sessions schedule. It is totally worth it. Do you know any other resource useful in the FAIRification process? Let us know in the comments.
Featured image: Elena Giglia (2021).
Avanço, Karla, and Arnaud Gingold. 2021. “FAIRifying a Scholarly Publishing Service: Elements for a Toolkit,” July. https://doi.org/10.5281/zenodo.5148089.
Beretta, Francesco. 2021. “A Challenge for Historical Research: Making Data FAIR Using a Collaborative Ontology Management Environment (OntoME).” Semantic Web 12 (2): 279–94. https://doi.org/10.3233/SW-200416.
Edmond, Jennifer, ed. 2020. Digital Technology and the Practices of Humanities Research. Open Book Publishers. https://doi.org/10.11647/obp.0192.
Giglia, Elena. 2021. “FAIR Data in the Humanities.” September 15. https://doi.org/10.5281/zenodo.5510388.
Jetten, Mijke, Marjan Grootveld, Annemie Mordant, Mascha Jansen, Margreet Bloemers, Margriet Miedema, and Celia W. G. van Gelder. 2021. “Professionalising Data Stewardship in the Netherlands. Competences, Training and Education. Dutch Roadmap towards National Implementation of FAIR Data Stewardship.” Zenodo. https://doi.org/10.5281/zenodo.4486423.
Koster, Lukas, and Saskia Woutersen-Windhouwer. 2018. “FAIR Principles for Library, Archive and Museum Collections: A Proposal for Standards for Reusable Collections.” The Code4Lib Journal, no. 40 (May). https://journal.code4lib.org/articles/13427.
Margoni, Thomas, and Prodromos Tsiavos. 2018. “Toolkit for Researchers on Legal Issues,” December. https://doi.org/10.5281/zenodo.2574619.
Paseri, Ludovica. 2020. “Access to Scientific Information and Knowledge: A Matter of Democracy.” In IFDaD.
Schapira, Matthieu, The Open Lab Notebook Consortium, and Rachel J. Harding. 2019. “Open Laboratory Notebooks: Good for Science, Good for Society, Good for Scientists.” 8:87. F1000Research. https://f1000research.com/articles/8-87.
Tóth-Czifra, Erzsébet. 2019b. “The Risk of Losing Thick Description: Data Management Challenges Arts and Humanities Face in the Evolving FAIR Data Ecosystem.” https://halshs.archives-ouvertes.fr/halshs-02115505.
Wilkinson, Mark D., Michel Dumontier, Susanna-Assunta Sansone, Luiz Olavo Bonino da Silva Santos, Mario Prieto, Dominique Batista, Peter McQuilton, et al. 2019. “Evaluating FAIR Maturity through a Scalable, Automated, Community-Governed Framework.” Scientific Data 6 (1): 174. https://doi.org/10.1038/s41597-019-0184-5.
Wuttke, Ulrike. n.d. “‘Here Be Dragons’: Open Access to Research Data in the Humanities.” Scientific Blog. Digital Humanities Methods and Tools (blog). Accessed October 21, 2021. https://dhmethods.hypotheses.org/262.
- In the framework of the Triple Project, Work Package 6 “Open Science and EOSC Integration” organises training on EOSC guidelines and Open Science. The training sessions target not only Triple project members, but also the SSH community in order to develop Open Science skills.