Betrokken en zorgvuldig - Kennis maakt ons beter

14. Scientific misconduct and how to prevent it

14. Scientific misconduct and how to prevent it

14.1 Introduction

Research integrity requires honesty in presenting goals and intentions, in reporting methods and procedures and in conveying interpretations. Research must be reliable and its communication fair and full. Objectivity requires facts capable of proof, and transparency in the handling of data. Researchers should be independent and impartial and communication with other researchers and with the public should be open and honest. All researchers have a duty of care for the humans, animals, the environment or the objects that they study. They must show fairness in providing references and giving credit for the work of others and must show responsibility for future generations in their supervision of young scientists and scholars. European Code of Conduct for Research Integrity, 2011 .

The basic premise for scientific research is indeed honesty. Or, as Schuyt puts it most simply: 'Do not lie, do not steal.' *1   Yet, unfortunately, scientific misconduct does occur. Scientific misconduct has large negative consequences, firstly for science itself. Results have to be trustworthy for science to have the possibility to advance our knowledge. Similarly, scientific misconduct has negative consequences for its users, patients and society. They have to be able to trust the methods followed and the results obtained. Especially in biomedical research, misconduct may result in damage to patients' health and to healthcare in general. Also, fellow researchers need to be certain of the absence of misconduct, as misconduct would render impossible both cooperation and progress. Likewise, institutions must be sure that the research conducted in their name is trustworthy. They must be able to present an image that is worthy of support and grants. Finally, for individual researchers there are enormous risks connected to committing fraud. Even though in the short run their embellished output and increased funding may have advantages, the long-term consequences can be disastrous, as was recently shown by several cases of misconduct in, for example, the Netherlands. *2  

At the same time, independence in research is under growing pressure as a result of external factors: the increasing need to publish and the researchers' dependence on external funding. Individual researchers may also be driven by internal factors such as a desire to be the first and to produce positive results. This may make them overlook the need to be careful and to respect the process and the time needed to conduct good studies. Therefore, in this chapter we more extensively explain the general principles of what scientific misconduct entails and what can be done to prevent it.

Research misconduct may appear in many guises. The major forms of misconduct are: *3  

  1. Fabrication - making up results and recording them as though they were real.
  2. Falsification - manipulating research processes or changing or omitting data.
  3. Plagiarism - appropriating other people's material without giving proper credit.

Other forms of misconduct include:

  • Failure to meet clear ethical and legal requirements, such as misrepresentation of interests, breach of confidentiality, lack of informed consent and abuse of research subjects or materials.
  • Improper dealing with infringements, such as attempts to cover up misconduct and reprisals on whistle-blowers.

More specifically, Dutch Universities (VSNU, 2012) have stated in their Code of Conduct for Scientific Practice that scientific misconduct extends at least to the following:

  1. Falsifying data
  2. Entering fictional data
  3. Secretly omitting unfavourable results
  4. Deliberately misusing statistical methods to achieve conclusions other than those justified by the data
  5. Deliberately interpreting results and conclusions falsely
  6. Plagiarizing results or other authors' publications
  7. Pretending to be an author or co-author
  8. Deliberately ignoring or not recognizing the contribution of other authors
  9. Failing to exercise due care when conducting research.

Dishonesty or a lack of integrity can pertain to the research itself, to the dissemination of scientific findings by publication (including reports, applications and articles), to reviewing, and to applications for funding and jobs. Scientific misconduct can relate to all stages of research.

*1:Kees Schuit, Hypothese (NWO), 2013, pp. 08-11  
*2: Such as the cases of Prof. D. Stapel (University of Tilburg) and Prof. D. Poldermans (Erasmus University). See also Simonsohn U. Science 2012; 337:21; Simonsohn U. The data detective. Interview by Ed Yong. Nature. 2012;487(7405):18-9. 
*3: European Code of Conduct for Research Integrity, 2011 (www.esf.org).

14.2 Fraud in different stages of research

Planning of the research:

  1. Presuming the research question is scientifically driven while it is actually based on commercial or political interests, namely the interests of those granting the study.
  2. Using other people's ideas in research proposals and applications for funding without permission and/or citation.
  3. Making up data or pilot data for research proposals and applications for funding.
  4. Consciously omitting relevant knowledge from research proposals and applications for funding.

Data collection:

  1. Fabrication of data.
  2. Adding fictitious data.
  3. Wholly or partly failing to observe the inclusion and exclusion criteria in the protocol.
  4. Infringing the privacy of persons (patients as well as healthy subjects) taking part in the research.

Analysis:

  1. Massaging data to produce better results; adding and discarding data.
  2. Selective and unreported omission of unwanted data (e.g. outliers) and results.
  3. Improper use of statistical techniques to produce more desirable conclusions.
  4. Distorted interpretation of data or distorted conclusions.

Reporting:

  1. Incorrect or distorted representation of other people's findings (misquotation).
  2. Intentionally and unjustifiably presenting oneself as author/co-author (guest or honorary authorship) and leaving out qualified researchers (ghost authorship).
  3. Intentional and unjustifiable referencing.
  4. Failing to acknowledge other people's original observations (under-citation).
  5. Exaggerated self-citation to inflate one's own citation index (even citing one's non-existent papers).
  6. Plagiarism; self-plagiarism and slothing (copying of sentences without referencing)
  7. Untrue mention of other information (e.g. CCMO/MEC or DEC permission; granting body; untrue acknowledgement).

Submission for publication:

  1. Unreported multiple submissions or publications.
  2. Duplicate and redundant publications ('salami publication').
  3. Unreported conflict of interest.
  4. Unwarranted waiting for or omitting of reporting to serve commercial or political interests.

14.3 Fraud in judging others

Colleagues:
Colleagues are favourably or unfavourably judged on something other than scientific arguments.

Reviewing of articles and applications for grants:

  1. Reviewing of manuscripts or grant applications in which one has too much interest (e.g. manuscripts written by friends or enemies, or grant applications from researchers whose work is too close to one's own field of research).
  2. Stealing of original ideas by reviewers or editors.

14.4 Preventing fraud


Although there will always be persons who commit serious scientific misconduct for their own reasons, measures can be taken and a culture can be created in which misconduct is less likely to occur. Such measures can be divided into activities on an individual, a departmental and an institutional level. Here, we limit ourselves to the individual and the departmental or research programme level.

The basic premises of research integrity are transparency and independence. Of these, transparency is most likely to be influenced by the measures taken. At the same time, a culture in which the importance of research integrity is explicit and evident is also important. Measures can thus be promoted to stimulate a culture of decent scientific thinking and behaviour.

I. Individual researchers

An individual researcher should be aware of the high demands involved in science: carefulness, precision, readiness to be accountable. The individual researcher is ultimately responsible for the behaviour related to doing research, be it the design, data collection, data storage or result reporting. *1 The researcher should always keep in mind that everything he/she does should be amenable to an open discussion. If not - that is, if one believes that something is better not seen or heard - one should not proceed, but should start thinking about what is at stake and perhaps discuss the issue with others.

  1. To be transparent, it is crucial that steps taken and decisions made are well documented. By doing this, the process of argumentation can be properly reconstructed afterwards. Moreover, data, arguments and decision should be documented in such a manner that they can be found and understood by others. It must be remembered that data and lab notebooks are the institute's property. Thus, they should be properly stored in a safe place. Also, the archiving of data should be transparent; that is, not only the data itself but also the ways of coding should be well documented (see Chapters 5, 6 and 7).
  2. Research is usually not an isolated activity, and even if it is, it is important to organize collaboration and feedback. In this way, unclear actions and irregularities can been discovered more easily. Organize regular feedback. Discuss dilemmas with peers and supervisors in time. Realize that the ultimate aim is not competition but the progression of science and knowledge.


II. The level of department or research programme

At the level of the department or research programme, it is possible to create a culture in which science flourishes in a transparent manner. The heads of departments and leaders of research lines are responsible for organizing the working environment in such manner that adequate scientific thinking and behaviour is stimulated. Two issues are crucial: a safe environment and continuous attention to careful handling of the scientific endeavour.

  1. A safe environment is helpful for scientists to be able to discuss dilemmas and possible errors. Young people especially should feel free to ask questions and express their concerns. This is possible only when the culture is such that doubts and temporary lesser performance are allowed and the pressure on people is not too heavy. Therefore, leaders of research lines or programmes are responsible for fruitful and honest scientific exchange within their research groups. Indeed, most important is the fair play of all collaborators involved.
  2. The best safeguards against fraud are cooperation between researchers, mutual evaluation of research, and a policy on publication that includes a thorough, independent peer review.
  3. Programme leaders can encourage cooperation among researchers by organizing a project team to do the research and internal reporting. Thorough discussions help prevent fraud.
  4. Although there may be a clear division of labour within a project team, it is important for the team to work as a unit when deciding how the data are to be collected, assessed and interpreted, and how the results are to be reported. Regular in-group reporting/discussions, with the responsible leader as organizer, are essential. Regular checks of one another and mutual feedback reduce the risk of fraud. Isolated individuals are more prone to scientific misconduct due to the lack of social control. This also applies to reporting: proper supervision and feedback to ensure that the rules on citing other authors are observed prevent plagiarism. In other words, good mentorship also reduces the likelihood of misconduct.
  5. Many research projects (especially those of larger research groups and consortia) have a steering committee in addition to the project team. Again, the risk of fraud is reduced if the progress of the project is discussed and the results are presented to the steering committee, and possibly outsiders, on a regular basis. The work done by staff members of research steering committees contributes substantially to the quality and integrity of the research as a whole, even if it does not immediately produce publication points for individual members of staff.
  6. Regular audits of individual studies can be organized within the research group. See for example the audit procedures described in the Quality Handbook of EMGO+.
  7. The leaders of research lines or research programmes should introduce all PhD students to the rules of good scientific conduct. The mentors should discuss these issues at the start of a contract. Courses that deal with aspects of integrity can be taken by the PhD students elsewhere. Obviously, good scientific conduct and the occurrence and dangers of misconduct can also be discussed within research teams.


On scientific misconduct, see also http://ori.hhs.gov/TheLab/TheLab.shtml

*1: Schuyt et al., Zorgvuldig en integer omgaan met wetenschappelijke gegevens. KNAW, Amsterdam, 2012


print