As part of the BioDATEN Science Data Center associated with DataPLANT, an assessment of research software was developed together with the other Science Data Centers in the FDM working group in Baden-Württemberg and has just been published on "Bausteine FDM" online journal:
In review of the replication crisis in science and the associated fundamental questions about the validation and reproducibility of research results, research software is increasingly becoming important in the focus of research data management and open science initiatives. Against this background, the published paper is dedicated to the responsible use of research software with a focus on good scientific practice and FAIR principles (Findable, Accessible, Interoperable, Reusable). Research data can only be meaningfully interpreted and reused in the long run in it's respective context. Accordingly, the sustainable use of software should be implicitly seen as part of FAIR data management.
However, when communicating with researchers in particular, but also with other stakeholders, it becomes clear to consider research software as a topic in its own right and to take into account its special requirements with regard to successful long-term access to research data. Against this background, the published paper will first explain the importance of software in the research context and clarify which forms of research software exist. This is followed by an examination of the risk posed by research software to the reproducibility of research results. The paper deliberately does not consider the topics of version control and versioning of software and the provision of the necessary hardware basis, which have already been discussed in more detail for the reproducibility of data. Established quality assurance procedures such as test-driven development and continuous integration are also not addressed. A risk assessment is also carried out. On this basis, comprehensive recommendations are made as to how different institutional levels can contribute to ensuring that research software remains available and reusable in the long term and sustainably.