Good Scientific Practice under the DFG Code: What It Requires on Documentation, Data Access, Archiving and Authorship

In Germany, good scientific practice (gute wissenschaftliche Praxis) is defined by the Deutsche Forschungsgemeinschaft (DFG) in its 2019 Code of Conduct, Guidelines for Safeguarding Good Research Practice [1]. Universities and research institutions had to implement the Code in legally binding form by 31 July 2021 to remain eligible for DFG funding.

This article explains what the DFG Code of Conduct requires on authorship and publication, and how those requirements connect with international journal standards [2, 3].

The Code contains 19 guidelines, organised around principles, the research process and the handling of non-compliance. Those most relevant to publication are [1]:

Guideline 14 defines an author as someone who has made a genuine, identifiable contribution to the content of a research publication involving text, data or software. A leadership or supervisory function does not itself justify authorship, honorary authorship is not permitted, and all authors agree on the final version and share responsibility for it [1]. These rules are consistent with the ICMJE criteria used by biomedical journals, which add explicit requirements for final approval and accountability [3]. We cover Guideline 14 in detail in our guide to the DFG authorship and publication guidelines, and compare it with ICMJE and CRediT in authorship criteria for research papers.

Authors choose the publication medium carefully, considering its quality and visibility in the relevant field [1]. The scientific quality of a contribution does not depend on where it is published. In practice, authors should check whether a journal follows recognised standards, such as COPE membership and transparent peer review, before submitting [4].

Practical test: verify the journal’s editorial board, peer review policy, indexing and fees. Invitations that promise very fast acceptance are a warning sign of predatory publishing.

Researchers decide whether, how and where to make results public, and there can be legitimate reasons not to publish, such as patent applications [1]. When results are made public, they are described in full and transparent detail. Wherever possible and reasonable, the underlying research data, materials and software are made accessible following the FAIR principles, and self-programmed software is made available with its source code where appropriate.

All information relevant to producing a result is documented so that the result can be reviewed and assessed, including negative results that do not support the hypothesis [1]. Research data, materials and software underlying a publication are archived for an appropriate period, normally ten years, at the institution where the data were produced or in cross-institutional repositories.

Under Guideline 6, every institution must appoint independent ombudspersons whom researchers can consult in confidence [1]. The German Research Ombudsman (Ombudsman für die Wissenschaft) provides an alternative, cross-institutional contact point and an online portal with the Code and implementation guidance [2].

Before submission, confirm that [1, 3, 4]:

• Each author made a genuine, identifiable contribution.

• No honorary authorship is included.

• All authors have approved the final version.

• The journal has been checked for quality and visibility.

• Data, materials and software are documented and archived for ten years.

• Data availability matches both the journal policy and the FAIR principles where reasonable.

See how these checks play out in a DFG-funded group in our publication support case studies.

The DFG Code of Conduct turns good scientific practice into concrete obligations for German researchers. Its authorship and publication guidelines require genuine contributions, careful choice of publication medium, transparent results and long-term archiving, which together protect research integrity and align with international journal standards.