Why do we need to consider ethical matters in the field of scientific research? -trend and guideline of response

Shigeaki YAMAZAKI (Professor, Aichi Syukutoku Daigaku)

1. Introduction

The Exploratory Committee of Scientific Misconduct was established in Osaka Univ. in the wake of revelation of misconduct in the Medical School. The Committee made the policy to respond to scientific misconduct in Osaka Univ. and discussed the procedure to deal with them. This is the first plan for scientific misconduct in Japan.

The Legislative Bureau House of Councilors is planning to move to validate the response to scientific misconduct. Without legislation, it is difficult for Cabinet-level ministry and agency to make guidelines.

I wrote a book "Kagakusya no Husei-koui(Misconducts by Scientists)". One reader, who committed misconduct and left the scientists community, give his correspondence that he felt that he had healed by my book. I feel that the research on scientific misconduct is a kind of treatment of illness of the scientists' community.

The reason why we study scientists' misconduct is shown in the words by Farthing:

Protecting the public from research misconduct is just another aspect of public health. It is no different, say to the Drinking Water Inspectorate which monitors our domestic water supply, or the Food Standards Agency which will set as an independent watchdog to ensure that the food we buy is safe to eat.

2. Regulation of Scientific Misconducts in the USA

Laws that relevant to research ethics:
1966: Animal Welfare Act
1974: National Research Act (1985: Health Research Extension Act

The Office Research Integrity (ORI) was established in 1992. Office Research Integrity is attached to Public Health Services (PHS), so ORI covers only research granted by PHS. If the research project is judged as a project that includes misconducts, the principal researcher cannot get a grant for 3 years and his or her name is disclosed to the public. In the USA, they can request information under the information disclosure law. The case of misconduct in research is not an exception. All information other than a whistle blower's name should be disclosed.

At the start-up time, ORI emphasized a judgement whether the case is guilty or not guilty. But they had to spend a lot of time for judgment. Now they are emphasizing education. I think research on integrity is very important. ORI shows model procedures for responding to allegations of scientific misconduct. Each university can follow it. This model procedure is recognized as a good model in many countries including the UK.

A researcher cannot get a grant from NIH without a response manual and a response organization to misconducts. There is a cultural difference about a response to misconduct between Japan and the US. In the US, if you find someone's misconduct, you should report it. In Japan, if you see someone's misconduct, you should keep it in secrecy.

• Organizational structure to respond to scientific misconduct
  Research Integrity Officer
  Vice President (in charge of research ethics)
  Committee of Inquiry
  Committee of Suit: Vice President (in charge of research ethics)
  Committee of Punishment
• Flow of Inquiry and Time Table:
  1) Accept of accusation
  2) Primary Inquiry
  3) Appeal and petition of objection
  4) Holding of Official Committee of Inquiry
  5) Official Inquiry
  6) End of Inquiry
  7) Recommendation to Committee of Punishment
  8) Public Announcement of the Result of Inquiry
  (1-3: the first month, 4-8:the second or the third month)

Fig 14 shows the cooperation among the organizations for inquiry of scientific misconduct. The cooperation prevents them from drawing a wrong conclusion.

3. Frequency of Occurring Scientific Misconduct

Research by ORI finds about 5 misconduct cases from 10,000 research projects. For example, in 1997, they have 32,000 PHS research project. 15 cases were judged as misconduct cases. I think these are the tip of an iceberg.

The article in Nature435 (9 June 2005) showed the new fact: 33% of 3247 researchers answered that they have committed at least one in the 10 major misconducts such as "fabrication of data" and "change of research design, method, and results" in the past 3 years. This means 33% of researchers confess that they have committed misconducts (Martinson BC et.al. Scientists behaving badly. Nature 435(9 June 2005))

4. History of Misconduct and Its Regulation

Establishment of Bayh-Dole Act in 1980:The objective of Bayh-Dole Act was an enhancement of industry-university cooperation. The law established that intellectual patent of research results by the federal government research fund is to belong to the individual researcher or his or her institute. Under this law, researchers and institutes can make a profit. They guessed that the frequency of research misconduct would increase and discussed how to respond to misconducts in the Congress (Goa etc.). Actually, under the Bayh-Dole Act, the number of patents universities received substantially increased from 1981 to 2001(especially for the top 100 universities). I think the essence of scientific research changed from intellectual interest to making profits by development of industry-university cooperation.

In 1980, people believed "scientific activities are self-governing activities by scientists. Self-regulation mechanism is working by referee system. Big responses to misconducts are not needed". They believed that government should not intervene in the problems within the scientific communities because scientists can solve their problems with self-regulation systems. However editorial misconducts that reviewers steal posts' ideas began to be pointed out after the International Congress on Peer Review in Biomedical Publication in 1989.

Discussion on research misconducts began in the Congress. In 1989, specialized agencies, OSI and OSIR, were established. Research misconducts are defined as FFP (Fabrication, Falsification, Plagiarism), however some problems are not included in FFP such as Gift authorship, duplicate publication, redundant publication, wasteful publication. Maybe the definition ought to be widened.

Federal government limited the definition of research misconduct to FFP in 2002. In 2005, the expression "認められた慣行からの重大な逸脱行為" was deleted, so the definition became more limited and more concrete. The definition of research misconduct became clear 20 years after Goa pointed out research misconducts.

On the other hand, ORI recommends the broader definition. I think the definition will become broader little by little. Scientists tend to limit a definition and a concept, but a funding organization tends to prefer the broad definition based on reality of the situation. The definition by the Federal government is the minimum definition.

5. What is a Cause of Misconduct

Think of the case of Kyushu Univ. assistant professor arrested for possession of ant-hypnotic agent. Was it just a personal scandal of an elite? No. I think events or misconducts are not caused only by personal circumstances but also by culture or the circumstances of a laboratory.

In this case, Mr.Y, who is the present head of the institute, was a mentor of the arrested professor. The first author for 4 of 10 papers which Mr.Y published when he was in Nagoya Univ. was the arrested professor. This means the arrested researcher was at Mr.Y's right hand. And also, Mr.M, whom I mentioned in my book about research misconducts, also belonged to the same laboratory.

I think that the personality of a researcher is not the only risk factor of misconducts but the circumstance of the laboratory is also an important risk factor. Continued existence of the institute is not certain because of introduction of fixed term and transformation into independent administrative institutions of Kyushu University. We can imagine the level of the researchers' stress was high in the competitive circumstance.

Schon events in the Bell institute happened in bad circumstances after the slide of stock price of the parent company, Lucent Technologies. When the stock price of Lucent Technologies slid, the numbers of papers from the Bell institute suddenly decreased. A circumstance is a serious risk factor of misconducts.

What is a cause of misconduct? A person or a system? I think both. In order to solve this kind of misconduct problem, we should emphasize a cause analysis and improvement of research circumstances in order to solve this kind of problem.

The definition of Research misconduct:
FFP(Fabrication, Falsification, Plagiarism)

Other derogation (a wide definition of research misconduct):
Gift authorship, duplicate publication, redundant publication, wasteful publication, unfair use of funds, sexual harassment

[Back Ground]
Performance based:patents, major journals, impact factors
Conflict of interest
Increase of competitive research funds
Need of a star researcher
Fixed terms
Increase of collaborative research project (large scale project)
↓result
[Circumstance]
Research elbowroom depleted, Competitive relationship between researchers,
Publish results without enough verification
↓
[Back Ground 2] In Japan, self-regulation and legal control for scientific misconduct are insufficient.
↓
[Conclusion] Restructuring research circumstances is necessary

6. Issues in the Future

Q&A:

1. Staff training

The objective of our research project is to develop a program of research ethics. What kind of training is necessary for the staff in a research integrity research center? Would you please tell me the program of Osaka University?

A:In Osaka Univ., we developed a procedure, but it was a problem as to who put it in operation. The conclusion of the discussion was that the existing administrative officials should study it and should put it in operation. It is true that we have a problem on human resources. For example, American people are familiar with observation and supervision. In Japan, we have had an ombudsman system. So I think that it may be a good idea that we apply the ombudsman system to the research ethics area

2. Peer review and Referee system

I would like to ask about peer review and referee system. You said that in the 1980's they had an idea that scientific activities, which are autonomic activities, have a self- regulation system, a referee system, so they don't have to have particular countermeasures against misconducts. If the referee system is as valuable as to be continued in 21st century, how we should improve the system.

I don't have a decisive idea. Now, in most cases, 2 referee judge in single blind system. However the chief editor of BMJ proposed an open reviewer system around 2000. He thought that misconduct by reviewers will be decreased by introducing an open reviewer system. In this proposal, names of the reviewer should be disclosed, but the correspondence is managed by the editorial desk.

In 19th century, they had such kind of system. A reviewer personally judges a paper and report to the academic society if the paper is worth recommending.

3. Systematic Exploitation of Labor Forces by Professor

Q: Some professors write a lot of papers using his or her students as a labor force. In Osaka university, has this kind of problem been discussed?

A: It is true that there are some on-campus politics, but I think it is important to make one's opinion public.

Q: I heard the that a new monitoring system of misconducts will be established for engineering management in Japan.

A: Monitoring from the top may be effective, but I think local input and case studies are important.

4. Double Blind Referee System

Q: In the UK, a double blind referee system is adopted by some journals. Today, introducing an open referee system is being discussed; on the other hand, the double blind system is also being discussed.

A: I think a double blind system is one possibility to prevent reviewer's misconducts. However, the double blind system requires much administration, so it is difficult to use the double blind system in major journals.

5. Achievement Oriented Culture

Q: I think an achievement oriented culture in research community is one of the backgrounds of misconducts. Could you give me some advice on how to get rid of such background for misconducts?

6. Employment type of staffs on an organization of research integrity

Q. Do you think the staffs on an organization of research integrity should be full-time, or interlocking?

A. Staffs of research integrity center should consist of both researchers and officers.

7. Actual Situation of Research Community

Q: I felt that it would be difficult to prevent misconduct in the present research communities. Think of this case as an example. A researcher did the same experiment three times. The researcher got good data in the 2nd and the 3rd experiments. And the researcher used the 2nd and the 3rd data because he or she recognized that the first one was an exercise. Is this a fabrication?

Now the researchers have to do research in a competitive circumstance without being shown what the definition of fabrication is. They think that they cannot win the competition if they try to follow the ethical norm completely and that they are lucky if their misconducts are not discovered. How can we establish a good system in which they can and would like to follow ethical norms?

A: As an example, in the Osaka Univ. Case, they did not record the experiment data in a notebook. They wrote a paper soon after they obtained good data and did not record them in a notebook, so other people could not do a check experiment. Even a layman recognizes this as a problem. I think that the universities should teach their students how to take notes, at least.

Q: In active laboratories, they are doing research in such ways. Now the researchers do not learn that they should record the data in a notebook. They do not learn rules about recording data.

A: On the ORI, it is explicitly stated how many years they should keep experiment notebooks and the property of the data.

Q: Japanese researchers go to overseas institutes to study for 2 or 3 years very often. However 2 years are too short to do good experiments, so it is highly possible that they push themselves. So it is often said that you must treat the data of Japanese with skepticism. This is the real situation of current researchers. We should establish education systems of research ethics reflecting such actual situations, or the education system is just imposing rules.