General remarks

Research misconduct means “fabrication, falsification or plagiarism that has been committed intentionally or with gross negligence in the planning, execution or reporting of research”, cf. section 3 (1), no. 1, of the Act on Research Misconduct, etc.

The three mentioned forms of research misconduct, often collectively referred to as “FFP”, are further defined in section 3 (1) nos. 2-4 of the Act:

• Fabrication: Undisclosed construction of data or substitution with fictitious data.
• Falsification: Manipulation of research material, equipment or processes, or the alteration or omission of data or results, through which research results are presented in a misleading manner.
• Plagiarism: Purloining the ideas, processes, results, text or special concepts of other people without providing proper attribution.

At the same time, section 3 (2) of the Act determines that research misconduct does not encompass:

• Instances of fabrication, falsification or plagiarism that have had little bearing on the planning, execution or reporting of the research,
• Questions relating to the validity of scientific theories, or
• Questions relating to the research quality of a scientific product.

A number of cases of research misconduct may be brought before the Danish Committee on Research Misconduct.

The further definition of the concept of research misconduct is discussed below. It should be noted that the definition of the concept of research misconduct contained in the Act is narrower than the definition that applied before the Act came into force and was the basis for the practice of the Danish Committees on Scientific Dishonesty (DCSD). This earlier practice is discussed in the Research Portal’s historical article on research misconduct under the Research Council Act. However, a significant part of the definition remains the same, and consequently, a significant part of the practice of the DCSD remains relevant for understanding the concept of research misconduct. Part of this practice is described in greater detail below.

The act must have been committed intentionally or through gross negligence

Research misconduct presupposes, as hitherto, that the act was committed intentionally or through gross negligence. The explanatory notes to the Act include the following:

“The Bill implies that a researcher must have committed the act or acts that are considered fabrication, forgery or plagiarism intentionally or through gross negligence. In a judgment handed down on 18 February 2015, 19th dept. No. B-3931-13, the Eastern High Court made a stricter assessment of the imputation requirement in cases of scientific dishonesty, in which the High Court states that the area is characterised by the deliberate or closely associated grossly negligent breach of the fundamental requirements of scientific research. The High Court further found that research misconduct relates to cases in which it must be possible, with the necessary clarity, to establish that the researcher in question acted deliberately or with gross negligence to falsify or distort the scientific message. In a judgment of 30 March 2016, 13th Division no. B-0864-15, the Western High Court expressed similar observations. With this Bill, it is proposed that this stricter perception of the imputation requirement be discontinued. According to the Bill, the imputation of scientific dishonesty must be assessed in accordance with the usual understanding in Danish law of the concepts of intent and gross negligence, without this assessment being further sharpened. Furthermore, the assessment is not proposed to imply a requirement that intent or gross negligence must have existed in connection with a falsification or distortion of the scientific message, as in specific cases research misconduct may be found in methodological sections or the like which does not in itself affect the product’s scientific message, but which may nevertheless have a detrimental effect on the integrity of the research.”

The remarks that the High Courts have applied “a stricter assessment of the imputation requirement” in the two judgments are difficult to follow and probably based on a legal misunderstanding. The judgment of the Eastern High Court may be read here, while the judgment of the Western High Court may be read here.

Fabrication and falsification

As mentioned above, fabrication is defined as the “undisclosed construction of data or substitution with fictitious data”. The explanatory notes to the Act include the following:

“The definition of fabrication is expected to encompass cases in which a researcher uses constructed or fictitious data in research without disclosing this, which may, for example, result in the research not being based on the presented foundation. ‘Data’ should be understood here in a broad sense, i.e. primarily comprising material, measurements, records and results. It is proposed that fabrication be defined without further qualification in relation to the undisclosed construction of data or substitution with fictitious data, other than what follows from the proposed de minimis limit in section 3 (2) no. 1 of the Bill, i.e. that all cases of undisclosed construction or substitution that are not below the de minimis limit will, according to the Bill, be covered by the concept.”

Falsification is, as mentioned, defined as the “manipulation of research material, equipment or processes, or the alteration or omission of data or results, through which research results are presented in a misleading manner.” The explanatory notes to the Act include the following:

“The definition of falsification is expected to encompass cases in which an adaptation, change or omission of data in a broad sense may, for example, cause the presented research to deviate from the actual research foundation. The requirement that the manipulation, change or omission must result in the research appearing misleading should be understood in relation to the presentation of the research in its entirety. The provision is thus not intended to require that the research results specifically must be misleading, but rather that the presentation of, for example, data, methods, interpretations or results in the research must be misleading before the case can be considered falsification under the Bill.”

From the practice of the Danish Committee on Research Misconduct (DCRM), reference may be made to the following rulings:

• DCRM ruling of 25 October 2019: The case concerned the manipulation of research material whereby the graphic material in several figures had been duplicated and altered to support the conclusions. Images had also been manipulated.

From the practice of the former Danish Committees on Scientific Dishonesty (DCSD), reference may be made to the following rulings:

• DCSD ruling of 18 December 2013 on image manipulation (case no. 1): Reuse of the same figures and the use of image manipulation to attempt to obscure this were equated with the undisclosed construction of data or substitution with fictitious data.
• DCSD ruling of 28 October 2013 (Helmuth Nyborg case): A reference to UN data was deemed misleading to the reader of the article, as this reference was used to support data that was not immediately apparent from the source to which the reference referred. This was a significant source for the results in the article. The Committee found that this was a serious breach of good scientific practice which must be equated with “undisclosed construction of data or substitution with fictitious data”. The DCSD found that the respondent had acted with, at the least, gross negligence, and added that the respondent had failed to provide another reference where the figures in question could be found.
  • This DCSD decision was brought before the Western High Court, which ruled in a judgment of 30 March 2016 that Helmuth Nyborg had not committed research misconduct as stated in the ruling of 28 October 2013.
• DCSD ruling of 29 August 2012 (figures): The method which, according to the respondent’s information to the DCSD, had actually been used did not appear in the method description provided in the submitted manuscript. The DCSD found that the respondent had thereby provided misleading and incorrect information about the methods she had used. The DCSD stated that this – together with incorrect information about the contributions of two students to the respondent’s manuscript – constituted serious breaches of good scientific practice committed with gross negligence, and that research misconduct had thereby been committed.
• DCSD ruling of 29 August 2012 (mice): The DCSD found that the description of the groups of mice used in a health science article was “unusually imprecise” and that parts of the description could not be correct. The DCSD assessed that not all of the mice described in the article had actually existed. The respondent had thereby provided misleading and incorrect information about the material included in the study, the type of groups of mice included in the study, and the subsequent amalgamation of the groups. The DCSD found that this was a serious breach of good scientific practice that had been committed intentionally and that this thus constituted research misconduct.
• DCSD ruling of 7 June 2012 (health economics): In an article, the authors had used a cost group that was not present in the source material mentioned. The DCSD found that this could not be equated with the construction of data in the sense of the Executive Order on Scientific Dishonesty, but that objectively viewed, it was a deficient and misleading account of the cost groups used in the study. However, this was not considered a serious breach of good scientific practice, but rather a question of the quality of the research, on which the DCSD has no authority to take a position.

Plagiarism

On this subject, reference is made to the following articles:

• Plagiarism
• Overlapping publications and self-plagiarism

Triviality limit (de minimis)

Research misconduct does not encompass “instances of fabrication, falsification or plagiarism that have had little bearing on the planning, execution or reporting of the research”, cf. section 3 (2) no. 1 of the Act.

The explanatory notes to the Act include the following:

“Cases of minor importance are expected to include, for example, the plagiarism of insignificant passages in methodology sections and the like, and completely peripheral fabrications or falsifications of minor scope, which have no impact on the research carried out. […]

It is proposed that the de minimis limit shall apply not only to the scientific message in the scientific product to which the case relates, as fabrication, falsification or plagiarism may exist in, for example, the methodological basis or analyses in the product that do not necessarily affect the scientific message. Similarly, the definition of research misconduct … shall not be limited to matters affecting the scientific message of the scientific product to which the case relates.”

From the practice of the Danish Committee on Research Misconduct (DCRM), reference may be made to the following rulings:

• DCRM ruling of 14 March 2019: With regard to the part of the complaint submitted to the DCRM that concerned plagiarism, the Committee found that the specific cases had little significance to the reporting in the scientific product. In the ruling, the specific cases are mentioned as follows:

“In support of the plagiarism charge, the complainant states that the idea expressed in the scientific product that ‘engagement in learning objectives (regardless of the quality of the engagement or the nature of the objectives) should provide a greater learning outcome than teaching without learning objectives’, is based on a ministerial guide that the authors could and should have cited as a source.”

• DCRM ruling of 24 September 2019 (dissent): A majority in the DCRM ruled that the instances of plagiarism were below the de minimis limit, while a minority (one DCRM member) found that they were above. The plagiarism is described in the ruling as follows:

“In assessing the significance of the plagiarism, the Committee has emphasised in particular that the reported product consists of eight chapters, of which Chapter 1, […], and Chapter 2, […], may be found on the product’s page […] to […]. The sections mentioned may be described as the introductory and methodological sections of the product, and the instances of plagiarism are found only in these sections.”

• DCRM ruling of 12 November 2019: In the scientific product it was stated that the age of 20 of the animals studied was “known”, regardless of the fact that it was estimated. The DCRM found that the scientific message in the product mainly referred to a demographic composition, and that the incorrect information identified had therefore been of little significance in the reporting of the research.
• DCRM ruling of 6 February 2020: The alleged plagiarism was considered to be covered by the de minimis The DCRM stated the following in this connection:

“There is a linguistic coincidence between the three sentences in the respondent’s product and source no. 1. Following an assessment of the meaning of the sentences in the respondent’s product, the Committee finds the use of the text of source no. 1 without proper attribution to have had only slight significance for the reporting of the research on which the respondent’s product is based. The Committee therefore finds that in instance no. 2, research misconduct has not taken place …”

The validity of scientific theories and the research quality of a scientific product

Research misconduct does not encompass questions relating to the validity of scientific theories, pursuant to section 3 (2) no. 2 of the Act. The explanatory notes to the Act include the following:

“The Bill essentially represents a continuation of current law with the linguistic modification that the current Act’s reference to the truth of scientific theories is deleted, since in a research context it is more accurate to speak of the validity of scientific theories than of their truth. The purpose of the rule is to establish that disputes relating to the conclusions of the research or the methods used, etc., cannot be considered research misconduct. The Bill thus proposes that the Danish Committee on Research Misconduct should not address questions of whether the theories and results presented in the scientific product to which the case relates are correct on the basis of a professional research and scientific assessment, but is instead expected to assess whether the research was produced by fabrication, falsification or plagiarism. The professional and scientific assessment of a scientific product is expected, as hitherto, to be carried out through peer review, etc.”

Neither does research misconduct cover questions of the research quality of a scientific product. This corresponds to hitherto applicable law.

From the practice of the DCRM, reference may be made to the following rulings:

• DCRM ruling of 14 March 2019 (case 2018-05): The Committee stated that the allegation related to the validity of scientific theory and the research quality of the scientific product.
• DCRM ruling of 6 February 2020: The Committee stated the following:

“It appears from the above that instances nos. 1 and 3-6 are instances in which the respondent, in the complainant’s opinion, attaches a different perception or significance to a source than that believed by the complainant to be correct. Disputes relating to the conclusions of the research or the methods used, etc., cannot be considered research misconduct. The Committee cannot address the question of whether the theories and results presented in the scientific product to which the case relates are correct on the basis of a professional and scientific assessment, and the Committee cannot undertake a qualitative assessment of the scientific product. Neither is there, on the basis of the information in the case, a foundation for establishing fabrication or falsification. Research misconduct is not therefore deemed to have taken place in instances 1 and 3-6 …”

From the practice of the DCSD, reference may be made to the following rulings:

• DCSD ruling of 22 August 2012: The complaint was found to concern the quality of the research.
• DCSD ruling of 7 June 2012 (health economics): In an article, the authors had used a cost group that was not present in the source material mentioned. The DCSD found that this could not be equated with the construction of data in the sense of the Executive Order on Scientific Dishonesty, but that objectively viewed, it was a deficient and misleading account of the cost groups used in the study. However, this was not considered a serious breach of good scientific practice, but rather a question of the quality of the research, on which the DCSD has no authority to take a position.
• DCSD Annual Report 2010 (case no. 5/2010 and case no. 6/2010)
• DCSD Annual Report 2009 (case no. 10)