Articles on Law, Science, and Engineering

Scientific Misconduct: Part 5 - Making Scientific Self-Regulation Work (cont'd)

What Should A Researcher's Code of Responsibility Address? [index]

A threshold issue is the minimum training required to undertake research. The inferior quality of much research done by individuals who lack rigorous doctoral or postdoctoral research training illustrates that special training in research should be required of all individuals who receive public dollars to support their research. This training should be standardized with criteria for minimum levels of performance in constructing and carrying out research protocols, with less emphasis on the popularity or luck of the research topic draw. The old saw rejecting training, techniques or experience gained from experiments with "negative results" should be modified to account for competence levels achieved during training. Breakthroughs based on good experiments should be rewarded for the competence in the design and conduct of the research rather than the apparent value of the results. Random information acquired while conducting research, however useful, cannot form the basis for a standard because research excellence requires much more than anecdotal experiences.

Licensure is another important issue. Most importantly, it would afford researchers an opportunity for self-control that is not now available in the profession. Licensure would facilitate communication of research results and the review of applications for research grants. It might also help solve the problem of self-designated researchers whose incompetent and outrageous claims are the mainstay of scientific yellow journalism. The black eye these individuals give to the profession need not be shared by licensed researchers.

Strict standards for recording and storing data would be particularly beneficial to the profession. In addition to eliminating many honest mistakes that can happen using informal methods, such standards would provide protection from false accusations of misconduct. For example, standards can be set so that data records are kept in a manner that minimizes the possibility of subsequent alterations by the author. In the case of data readings directly from burets, meters, etc., it has long been the accepted practice by scientists and inventors to enter raw data in ink on a dated, numbered page of a permanently bound laboratory book. The better practice is to have each page acknowledged by another person's initials.

The integrity of data generated in charts by machines can be safeguarded by similar indexing. Data in electronic form can be protected in magnetic media by using a security format. Originals of all data records can be maintained in secure locations, with copies, but not the originals, available to the author. Falsification of data is substantially less likely using such practices because the likelihood of getting caught where misconduct is witnessed by others. Reliable records kept in this manner have the added advantage that they are convincing support in contests about invention dates for patents.

Establishing standards for routine research techniques would also benefit the profession. Good laboratory technique is essential to good research. If licensure of the author of a research publication or proposal included a representation that, unless noted otherwise, standard research techniques were employed, the reader would be better placed to evaluate the results reported.

It is also important to establish professional standards for administrative conduct at institutions receiving federal support for research. At most such institutions, department chairs, deans, and other administrators are hired on the basis of their records as researchers rather than their administrative experience. In such cases the administrator-researcher is usually taking part in proposing, conducting or reporting research. Basically, administrative conduct that uses resources to promote more or better research should be encouraged while conduct that wastes resources should be discouraged. Thus, decisions about matters such as employment and space allocation should be made on the basis of efficient use of tax research dollars rather than tenure, seniority, and other irrelevant considerations. Further, at institutions where faculty participates in these administrative decisions, those with conflicting interests should decline to take part.

Professional standards for researchers must also address the relationships amongst those who participate in the research. The general standard must be that relationships remain strictly within professional bounds. At issue is not solely the moral question of using power to exploit employees or favors to manipulate the boss, but the integrity of research data obtained under conditions where power or bribes purchase results. At a minimum, a serious deviation from professional research standards occurs if an unprofessional relationship is not noted when the data is published or used in a proposal to conduct additional research.

Any set of professional standards must address how misconduct is handled when it occurs. On the individual level, when a researcher knows of misconduct by another, there is an obligation to the profession to file a factual report (under oath) to appropriate authorities in the professional governing organization. As an example, many states now require physicians to report incompetent or impaired colleagues to the state board of medical examiners.

On the organizational level, when it is determined that misconduct occurred, there is an obligation to administer a fair and appropriate punishment. Support within the profession for these two standards are vital because the ultimate test of professional standards lies in the belief by the profession that the standards are necessary and just. If the profession fails to believe in its own standards, the public will lose confidence in the profession. The clearest signal the profession can give that it does believe in its standards is its willingness to report misconduct and police itself when misconduct occurs.

Next - Conclusions
Previous - The Failure of Physician Self-Regulation