Mohamed_hassan, Pixabay-Lizenz

Research Integrity Conflicts and Research Data

In the preamble of the German Research Foundation's (DFG) Code of Conduct Guidelines for Safeguarding Good Research Practice it is specified that “[s]cientific integrity forms the basis for trustworthy research”. In practice, however, conflicts or violations of good research practice may occasionally occur. The range of research integrity (RI) breaches is very broad, and the consequences for science and society vary accordingly. In theory, a general distinction is made between research misconduct and questionable research practices (QRP).[1] Which practices or behaviours are to be classified as misconduct and which, on the other hand, count as questionable research practices varies from country to country and (research) culture to (research) culture. In addition, there are RI conflicts that can still be remedied. In this article, the different types of RI violations will be explained with a special attention to research data. This article focusses on the situation in Germany. There is a separate entry on research ethics in Switzerland.

Research Misconduct

According to the explanatory note to Guideline 19 of the DFG Code of Conduct, “only intentional or grossly negligent infringements defined in a set of regulations are considered scientific misconduct”. These can be the examples listed in Section 21 (1) of the Model Statutes of the German Rectors' Conference (German only) on the Implementation of the Code of Conduct at Universities: if a person working in science at the university intentionally or through gross negligence makes false statements in a scientifically relevant context, appropriates the scientific work of others without authorisation or impairs the research activities of others. All research institutions in Germany have their own RI statutes, which implement the guidelines of the DFG Code of Conduct but may vary slightly. These include definitions of good research practice and scientific misconduct.

Misconduct in relation to data includes, in particular, the fabrication and falsification of data and the unauthorised removal of primary data. It can also include the infringement of intellectual property (e.g. unauthorised use or disclosure of research data), claiming authorship despite not fulfilling the criteria and sabotaging research activities (e.g. manipulation of third-party research data or its documentation). Internationally, the term research misconduct is often defined more narrowly and is based on a definition issued by the US National Academy of Sciences in 1992.[2] This definition only comprises fabrication and falsification of data as well as plagiarism (often referred to by the acronym FFP).

Questionable Research Practices

Apart from research misconduct, there are so-called questionable research practices. QRP are usually located in the grey area between good research practice on the one hand and serious misconduct on the other. There is, however, disagreement as to which practices fall under the term. In relation to data, practices like cherry picking, p-hacking and HARKing are often mentioned. Even if they do not constitute research misconduct, QRP can impair knowledge production in science and thus have far-reaching effects on politics and society. The National Academy of Sciences therefore suggested in 2017 that we should rather speak of detrimental research practices.[3] Particularly in the wake of the reproducibility crisis, questionable research practices have increasingly become a subject of discussion in Germany.

Info box: Common QRP in relation to research data

Cherry picking: deliberate selection or de-selection of data to prove a chosen hypothesis

p-hacking: manipulating data (e.g. by deliberate selection or de-selection of data, repeat collection of data or testing of multiple hypotheses after data collection) to present the results as statistically significant (indicated by a p-value of 0,05 or less). The p-value is not an indication of the observed effect being true but about the likeliness that the observed effect is not a product of chance.

HARKing: Hypothesizing After the Results are Known: Adjusting / setting up hypotheses after data has been collected.

Info box: reproducibility crisis

Reproducibility is an important aspect of empirical research. Since the 2000s, there has been an ongoing discourse on the lack of reproducibility of research results, particularly in psychology, medicine, the social sciences and the natural sciences. In consequence, the causes of the reproducibility crisis as well as potential solutions have increasingly become the focus of meta research.

Remediable RI conflicts relating to research data

There is a third category of RI conflicts one could distinguish from misconduct and QRP.  Concerning research data these often comprise questions of access, data usage and data sharing as well as the publication of data. Sometimes these conflicts are also subsumed under the label QRP. However, in contrast to the data QRP mentioned above as well as research misconduct these types of RI conflicts can be remedied with the help of ombudspersons. For example, action can be taken if

  • a person is denied access to data.
  • data access needs to be regulated during or after the departure of a researcher (e.g. due to change of jobs, work groups or institutions, graduation or expiring contracts).
  • disputes about data occur among researchers.
  • questions about data sharing arise (for example with whom and when data need to be shared)

If you are involved in a RI conflict or have observed potential RI violations, you can find more information on how to proceed in the article Procedure in Case of Conflict.

Sources

[1] Steneck, N. H. (2006). Fostering integrity in research: Definitions, current knowledge, and future directions. Science and Engineering Ethics, 12(1), 53–74. https://doi.org/10.1007/s11948-006-0006-y

[2] National Academy of Sciences, National Academy of Engineering, & Institute of Medicine. (1992). Responsible Science: Ensuring the Integrity of the Research Process: Volume I. National Academies Press. https://www.ncbi.nlm.nih.gov/books/NBK234525/ (accessed April 30th 2024)

[3] Committee on Responsible Science, Committee on Science, Engineering, Medicine and Public Policy, & Policy and Global Affairs. (o. J.). Fostering Integrity in Research. The National Academies Press. https://nap.nationalacademies.org/read/21896/chapter/1

Further reading on "reproducibility crisis"

Paper: "1,500 lift the lid on reproducibility"
Author: Monya Baker
Year of publication: 2016
Link to paper

Blog: "Data Colada. Thinking about evidence, and vice versa"
Authors: Uri Simonsohn, Leif Nelson und Joe Simmons
Year of publication: aktiv
Link to blog

Project: "Reproducibility Project: Psychology"
Authors: Various
Duration: 2011-2016
Link to project

Project: "Reproducibility Project: Cancer Biology"
Authors: Various
Year of publication: ca. 2013-2021
Link to project

How To Cite This Page (Chicago)

forschungsdaten.info editorial board. „Research Integrity Conflicts and Research Data“. forschungsdaten.info, May 2nd 2024. https://forschungsdaten.info/praxis-kompakt/english-pages/research-integrity-conflicts-and-research-data/.