Generally speaking, research data can be understood as any data generated in the conduct of academic research projects or made available, reused and processed for research purposes. Research data include measurement data, laboratory values, audio-visual information, texts, survey data, objects from collections or samples that are created, developed or evaluated in the course of scientific work; methodological test procedures such as questionnaires, software and simulations can also be considered results of academic research. (As defined in the DFG Guidelines on the Handling of Research Data (2015) [German])
The definition of what constitutes research data varies among academic disciplines, as does the way research data are handled. Data need not necessarily be digital to meet the definition of research data if they form the basis on which academic knowledge is generated. However, with progressing digitalisation a narrowing of the definition of research data can be observed, referring more specifically to digital or digitised data. This is, at least in part, the result of the digital transformation across academic disciplines and the respective discipline-specific research data management needs.
The understanding that research data have an intrinsic value underlies the need for research data management. The documentation and reusability of this data not only contributes to the transparency and reproducibility of academic research and thus to the verification and validation of research findings, but also facilitates – in the spirit of the Open Science movement – further scientific progress and quality assurance. This value of research data is expressed, among other things, in the latest research and funding policy efforts to establish a national research data infrastructure. In addition, research funding organisations have moved from merely encouraging researchers to manage their research data in line with the above to expecting that they do so as a matter of course: this is increasingly becoming a factor in funding decisions if not an outright condition for funding (see, e.g. Mauer and Recker, 2019 [content in German]).
Research data management is a building block of good research practice and was included in the German Research Foundation (DFG)'s Guidelines for Good Research (Guidelines 7, 10–15, 17) in 2019. Consequently, this means that – irrespective of external (third-party) funding – it is important to observe the relevant discipline-specific standards and methods when collecting, processing and analysing research data; moreover, the handling of research data throughout the entire research process should be documented in a transparent and comprehensible manner. Research data management covers the entire data curation lifecycle from the initial project planning stages to the review, collection, preparation, processing and analysis, and finally to the storage, publication or deletion of the research data.
In the absence of legal or ethical concerns that would prevent this, research data – like other results of a research project – should be made available for subsequent use in accordance with the so-called FAIR principles.
The so-called data curation lifecycle is mapped with a different number of stages, depending on complexity.
A more detailed description and further information can also be found on the fdm-bayern.org [German] website.
The Data Management Plan (DMP) is an essential instrument for research data management. It describes how data are handled throughout their lifecycle, which facilitates the structuring, documentation, organisation and reuse of research data. For German and international research funding bodies, information on research data management as well as formal DMPs are increasingly becoming a prerequisite for successful funding applications.
DMPs are dynamic, as they are continually updated throughout the research process and do not have a rigid structure. At its most basic, a DMP should generally include the following information:
For details, read the guidelines of the BMBF-funded WissGrid Project by Ludwig and Enke (2013) [German].
There are a variety of free tools available online for creating data management plans (DMPs). The Research Data Management Organiser (RDMO), is one example of a functional tool. A good overview and further links can be found on the interdisciplinary information pages of forschungsdaten.info.
Examples:
Please also take a look at the handout from the Working Group Research Data Management on RDM in DFG proposals [content in German].
Depending on the funding body, the requirements for research data management differ, as do the details that researchers have to provide in the context of their funding applications. In most cases, a brief summary on the nature and scope of the research data, on the underlying research data policy or on discipline-specific guidelines and on storage and publication intentions will suffice. Increasingly, however, applications are expected to come with a data management plan. The Open-Access publication of research data after the completion of an externally (third-party) funded research project has not yet been made compulsory by German funding bodies but is expressly desired, for example by the DFG. Project-specific costs for research data management can be requested from most funding bodies.
You can find a detailed overview of the key guidelines and policies of various third-party funding bodies on the information pages of forschungsdaten.info [German].
| Funding institution | BMBF | DFG | EU Horizon 2020 | ERC |
|---|---|---|---|---|
| What to archive? | Research data, irrespective of the research result | Research data | Research data, unpublished data, program code | Research data, unpublished data, program code |
| Where to archive? | Data storage in a data centre/repository | Free choice of repository | Free choice of repository | Free choice of repository, GenBank and PDB suggested |
| When to archive? | After project completion | Without undue delay | At the earliest possible time | Within 6 months of project completion |
| General conditions | Making data available to the academic community and for long-term preservation/reuse; Open Access if possible; adherence to FAIR principles wherever possible | Primary data must be stored securely for 10 years at the originating institution | In the Open Access pilot programme, a data management plan must be provided within the first six months of the project. These projects must take measures to ensure the research data can be used and duplicated free of charge (CCA or CCO licence). The scope of the disciplines involved is constantly being expanded. | Data deposition (nucleotide/protein sequences, macromolecular atomic coordinates, anonymised epidemiological data) should take place immediately after publication of the results |
Source: https://www.forschungsdaten.info/themen/informieren-und-planen/foerderrichtlinien/, last modified: 1 July 2020; own abbreviated presentation.
