The Complainant had alleged that 12 articles, of which the Defendant had been a co-author, were scientifically dishonest. Based on approaches from the co-authors of some of the 12 articles in which new information came to light, and on a request from the Defendant to have the case reopened, DCSD have now re-evaluated the case.
The unanimous decision is that the Defendant is still subject to criticism for four of the articles, however, two are no longer considered scientifically dishonest.
The new information that led to the reopening of the case made it clear that in two articles, in which the same subjects were used without the reader being made aware of this fact, the selection of subjects was not in fact concealed, as DCSD had ruled on 18 December 2013. Information was also forthcoming that led to the DCSD ruling that, in this specific case, it was not scientifically dishonest that some measurements in the same article were only performed on a subset of the overall study population without clarification and explanation to the reader.
Conclusions
The case concerns a charge of scientific dishonesty in 12 articles.
DCSD find the Defendant guilty of scientific dishonesty in four of the 12 articles. The main conclusions are as follows:
- That the Defendant, in articles 4 and 5, intentionally committed a serious breach of good scientific practice corresponding to undisclosed distorted interpretation of the Defendant’s own results, pursuant to the DCSD order section 2, no. 4, by failing to provide information that the biopsy material stemmed from the same subjects as in Article 1 and in another article authored by the Defendant. In these specific instances, knowledge was withheld from readers that results based on the same subjects were used to support the conclusions in articles 4 and 5. See also sections 7.16.1 and 7.16.2 in the ruling.
- That the Defendant, in articles 3 and 5, intentionally committed a serious breach of good scientific practice corresponding to unknown selective or surreptitious discarding of own undesired results, pursuant to the DCSD order section 2, no. 2, by failing to provide information that the biopsy material stemmed from previous studies with more subjects than the number cited in articles 3 and 5. As a result, knowledge was withheld from readers that a selection of subjects had taken place. See also sections 7.16.1 and 7.16.2 in the ruling.
- That the Defendant, in Article 12, by gross negligence, had committed a serious breach of good scientific practice corresponding to undisclosed construction of data, pursuant to the DCSD order, section 2, no. 1, by failing to provide information that eight subjects stemmed from a previous study (used in Article 8) and were therefore subject to a different research protocol than the one described in the methodology section of Article 12. See also section 7.16.4 in the ruling.
- That the Defendant, in Article 4, by gross negligence, committed a serious breach of good scientific practice corresponding to undisclosed construction of data, pursuant to the DCSD order, section 2, no. 1, by failing to respond to obvious image manipulation in the article. See also section 7.17.3 in the ruling.
Translation into English of the full ruling will follow.
How does DCSD investigate cases?
DCSD's rulings are of a complex nature. Here is a brief summary of some of the essential elements of the process:
1. DCSD’s mandate
DCSD’s remit is to investigate complaints of scientific dishonesty in order to improve the credibility of Danish research. The main concept is defined by the Act on research advisory system, etc.:
“Scientific dishonesty, i.e. falsification, fabrication, plagiarism and other serious violations of good scientific practice committed intentionally or due to gross negligence during the planning, implementation or reporting of research results”.
In cases where DCSD confirms scientific dishonesty, a ruling is issued to this effect. DCSD can also recommend that the work involved, typically a scientific article, be withdrawn by the publisher. DCSD is not empowered to issue more severe sanctions, i.e. it cannot force the parties to take certain actions.
DCSD does not rule on cases concerning scientific theories’ validity or truth or on cases related to the quality of the research associated with a scientific product. In other words: DCSD does not address the conclusions and results of research. Nor does DCSD determine whether the research is good or bad.
2. What constitutes a “serious breach of good scientific practice”?
The act refers to practice within science. Practice, which is not a written set of rules, changes over time, partly due to the fact that scientific research is in a constant state of development.
The academic members of the committee, comprising recognised researchers appointed on the recommendation of the Danish Council for Independent Research, are responsible, based on their knowledge of practice, for judging whether the particular case involves a serious breach of good scientific practice that, according to the law, corresponds to falsification, fabrication or plagiarism. It is therefore an academic assessment, similar to those conducted by other councils and boards, that determine whether a discipline’s standards have been met or violated.
3. Responsibility for intentional or gross negligence
It follows from the definition in the act that scientific dishonesty has been committed when there is a serious breach of good scientific practice (corresponding to falsification, fabrication, plagiarism, etc.) and if such a serious violation has been committed either intentionally (deliberately) or by gross negligence, i.e. in blatant disregard of the attentiveness or meticulousness normally expected in the specific situation. The dishonesty concept does not cover accidental or excusable mistakes, including those made with a limited (simple) degree of negligence.
Intentional or grossly negligent behaviour has been covered by the dishonesty concept since the first committee on scientific dishonesty was established in 1992. Similar concepts of responsibility are applied in other countries. It is also an integral part of the Danish legal tradition that a party is not only responsible for his/her intentional (conscious) behaviour but (at least) also for grossly negligent behaviour.
4. Joint responsibility
Authors of a scientific article do not have an objective (strict) responsibility for the contributions made by the other authors. However, it is part of good scientific practice that all authors of a scientific article have a certain degree of joint responsibility for its overall content, including responsibility for reading the final manuscript prior to submission to a journal.
DCSD is of the opinion that, generally speaking, the lead author(s) (often called the senior or final author(s)) of a scientific article has special responsibility for all of the article’s content, including reading the final manuscript carefully prior to submission to a journal.
Nevertheless, each case of dishonesty requires an evaluation of joint responsibility, and both a serious breach of good scientific practice and intentional or negligent actions by the person jointly responsible must be confirmed. DCSD bases its evaluations on the facts of each case, including the defendant’s qualifications and other relevant factors, so that the assessment relates to the defendant’s specific role and contribution to the scientific articles concerned.
5. Reopening a case
DCSD rulings are final in the sense that no other administrative authority is empowered to hear appeals about its rulings. DCSD does, however, have the power to reopen a case if significant new information is forthcoming that, had it had been available during the original proceedings, could have led to a different outcome.
After the initial ruling in this particular case, DCSD received new infor-mation from co-authors of some of the articles in question, and the Defendant also agreed to the case being reopened on that basis.
The ruling in the reopened case differs from the earlier ruling of 18 December 2013 most significantly in that DCSD now finds that the Defendant only committed scientific dishonesty in four of the six articles.
Copenhagen, 28 August 2014
Henrik Gunst Andersen
Chairperson of DCSD
For further information please contact:
Danish Agency for Science, Technology and Innovation
Head of Secretariat Charlotte Elverdam , tel. +45 72 31 82 20, email: chel@fi.dk