Tải bản đầy đủ - 0trang
VI. Ethics in the Conduct of Research
standpoint, it is more important to think about how we as scientists engage in those
debates and resolve those and other issues. How do we develop and test hypotheses, marshal1 data, draw and report inferences, engage in rational discourse, and
work with colleagues and students in an ethical manner?
In this section, I describe my perceptions of ethical and unethical behavior in
several specific situations encountered in agronomic research and in agricultural
research in general. In each case, I move from situations in which I believe there
is considerable agreement on ethical choices to ones in which ethical behavior is
harder to specify clearly. This is not an exhaustive treatment by any means but will
serve more as a partial list of research situations with ethical dimensions. In this
section, I used the terms scientist and researcher interchangeably.
1. Selecting Topics for Research
Scientists embark on specific research efforts because they are interested in the
subject matter, have expertise in the subject matter, wish to learn something, think
the subject matter is important, wish to perform a service, wish to gain personal
reward, or some combination of these. The decisions are conditioned by employment opportunities and the availability of resources to support research. I am not
convinced that these decisions, except those involving public service, have much
practical ethical content. Of course, each person must let hisher own ethical perceptions guide these decisions.
2. Designing Experiments
Creative scientists generate hypotheses. The more creative and knowledgeable
they are, the more likely they are to generate hypotheses that depart from the current paradigm. Although there is a definite resistance to new paradigms in the scientific community, researchers who bring about paradigm shifts are often rewarded. The greatest reward is the feeling of having gone where no one went before,
having been the first human in the history of the world to understand a phenomenon, no matter how minute and unimportant. Other more tangible rewards include
recognition, accolades, fame, and money.
Scientists may become emotionally involved with their hypotheses. They want
these products of their intuition and insight to be true. Unless scientists are vigilant, they will unconsciously do things in their research and analysis that will
“stack the deck” in favor of their hypotheses. Lay people placed in the role of experimenters do this in the extreme (Folwell, 1969).
It is important for scientists to design experiments that will disprove their hy-
PRACTICAL ETHICS IN AGRONOMIC RESEARCH
potheses if they are not true. Thus, scientists are put in the position of having to
scrutinize their own creations with cold objectivity and with a critical eye. They
must subject their creations to rigorous tests. Intellectual honesty and integrity are
of paramount importance in this situation. Only if a hypothesis is not proven false
by rigorous experimentation is it ethical for the scientist to promote it as a new
3. Collecting and Reporting Data
Both practical and ethical considerations dictate that scientists should carefully
and accurately record data. The levels of precision and accuracy required are functions of the size of the differences investigators need and wish to detect. It is generally accepted that fabrication and falsification of data constitute unethical behavior, except in special situations such as espionage. In reports, data should be
accurately represented so that readers are not mislead.
In many situations, it is hard to determine whether data have been fabricated or
falsified. The more a scientist knows about the subject matter of an experiment,
the easier to fabricate or falsify data without being detected. A knowledgeable person can shade the data so that they support a certain hypothesis without changing
them so much that the changes are apparent. Moving a few observations toward
or away from a treatment mean can alter the statistical analysis such that it supports one hypothesis over another. Such alterations of data are clearly dishonest
It is unethical for an investigator to provide only data that support a certain hypothesis if he/she collected other data. There are cases, however, in which extraneous forces ruin treatments and even entire experiments, rendering part or all of
the data useless. Whether or not experiments have been ruined or data are worthless or misleading is often a judgment call, requiring objectivity and integrity.
There are statistical techniques that accommodate missing data without changing variance estimates. These should not be used to deal with outliers, however,
unless the outliers are clearly artifacts. Intellectual honesty and integrity are very
important in this situation. It is best in these situations to err on the side of caution.
4. Analyzing Data
Most agronomic data are subject to statistical analysis. Statistical analyses are
conducted and reported so that those interested in an experiment can get a more
accurate perspective on the degree of variation, both controlled and extraneous,
encountered in the experimental material and conditions. With statistical information, others can decide whether they think an investigator’s conclusions and inferences are appropriate.
Typically, agronomists test hypotheses at P = probability of type one error =
0.05. Frustrated with the high degree of variation encountered, especially in field
environments, they may elect to conduct and report statistical tests at higher levels of R They may elect to report the actual level of P computed for each test.
Researchers may perform various transformations so that data conform more
closely to the assumption of normality associated with many standard analytic procedures. Such practices are not unethical, so long as they are accurately portrayed
to readers and other recipients of reports.
It is not a case, however, of “let the buyer beware.” It is ethically incumbent on
researchers to perform and present statistical analyses with as much objectivity,
accuracy, and intellectual rigor as they can reasonably muster. Again, the role of
experiments and statistical analyses is to disprove hypotheses that are not true, not
to support hypotheses.
Statistics can be used to mislead, either intentionally or unintentionally. For example, high correlation or multiple correlation coefficients result when there are
linear trends in dependent and independent variables, regardless of whether any
causal connections exist. As the number of parameters in a multiple regression
model approaches the number of observations in an analysis, the multiple correlation approaches 1 .O.
These spurious relationships may mislead both the researcher and hidher audience. It is obviously unethical to use statistical anomalies to mislead people deliberately. Whether it is unethical to use them unintentionally is in that gray area
5. Drawing and Reporting Inferences
Technically, inferences drawn from an agronomic experiment only apply to the
materials and conditions of that experiment. At the same time, it is usually impractical to test hypotheses with all possible materials and under all possible conditions. Compounding the problem is the necessity to provide research information to practitioners working within a wide range of materials and conditions.
Thus, there is great pressure on agronomists to extend their inference space beyond that included within their experiments.
It is not unethical to interpolate and/or extrapolate the results of experiments. It
is risky, however. It is important for the researcher to evaluate that risk and communicate it to audiences to the extent possible. It is unethical to interpolate or extrapolate with the intent to mislead. As in most research situations, there is a gray
area that requires good judgment, intellectual honesty, and integrity.
6. Establishing and Maintaining Credibility
When a researcher is competent, professional, ethical, and intellectually honest
in measuring, recording, analyzing, and presenting the results of experiments and
PRACTICAL ETHICS IN AGRONOMIC RESEARCH
inferences drawn from them, he/she establishes credibility and contributes to the
credibility of hisher organization. Without credibility, neither scientists nor their
organizations can function effectively.
Both scientists and practitioners have different expectations of public-sector and
private-sector scientists. Public-sector scientists are expected to be honest and
thorough in their portrayal of research results and inferences, even when they are
comparing competing products. A private-sector scientist might reasonably be expected to portray hisher company’s product in the best light and to remain silent
with respect to performance of competing products.
Private-sector scientists lose credibility for both themselves and their companies if they misrepresent their products and services or portray competing products dishonestly. Markets are good mechanisms for determining value and customers have long memories, so the punishment for lack of credibility can be severe
for private-sector scientists and their organizations.
There is a very large gray area associated with credibility. My experience suggests that, contrary to some expectations, public institutions and public-sector scientists do not have a corner on credibility compared to private-sector scientists.
There is considerable variation in each category. In both arenas, rigorous peer review and peer pressure provide strong incentives for intellectual honesty and ethical behavior.
B. ETHICSAND SCIENTIFIC
1. Preparing Proposals
In the quest for program and project support, scientists and science administrators spend much time preparing proposals. These may take the form of technical
proposals submitted to competitive grants programs in response to solicitations.
They may be solicited or unsolicited proposals to private firms seeking support in
the form of gifts, grants, or contracts. They may be political proposals, often prepared for lobbyists or legislative champions to use in supporting research-related
These categories of proposals are similar in requesting an investment in return
for something. Timeframes and deliverables may or may not be specified precisely, but some new information or technology, either prototype orproven, is expected
in return for the investment.
In research situations, outcomes are uncertain, almost by definition. There are
some things the investor can expect, however. Ethical behavior of proposal authors
in this situation includes promising only what can be delivered, clearly informing
the investor of the risks involved, and providing an objective estimate of the probability of achieving a goal.
If the agreed-on goal of a proposed research effort is to achieve some desired
practical outcome, both investor and applicant should be clearly informed of all
activities that will be required to achieve the goal, including those outside the proposed research effort. It is important that both researchers and investors start projects with reasonable expectations.
I find terminology associated with the recent Government Performance and Results Act useful in thinking about research organized around desired practical outcomes. The key terms are goals, activities, outputs, and outcomes. Outputs of research efforts often include publications. Sometimes prototype products and/or
processes are produced. Rarely is the desired outcome a direct result, that is, output, of a single research activity. Usually, other activities, including other kinds of
research, technology transfer, and commercialization efforts, are required.
It is unethical for a researcher to accept money from sponsors if the researcher
does not intend to help achieve the agreed-on goals. To do so is to accept money
under false pretenses. If a researcher does not believe the goal is appropriate or
ethical, he/she should not enter into an agreement specifying that goal. Often,
problems can be avoided by clearly specifying and agreeing on goals during contract negotiations.
Conversely, sponsors are obligated not to exploit researchers by using research
results for other than intended purposes. In my own experience, a manufacturer of
hay preservatives supported university research that supposedly showed a considerable benefit from using the product. The preliminary research results, which
were never subjected to peer review or published in scientific journals, were used
extensively in the firm’s advertising.
I pointed out to the firm that the design of the preliminary experiments was
flawed. My colleagues and I proved in other experiments that the product was ineffective. Our results were peer reviewed and published in a reliable journal. The
manufacturer continued to manufacture and market the product.
Unless the manufacturer really believed the product was cost-effective, it was
unethical to continue marketing it. I participated as an expert witness in litigation
initiated by dissatisfied users of the product. I did not enjoy or wish to be involved
in this litigation, but I felt morally obligated to make sure that research results were
It is appropriate to ask in this situation if it was ethical for researchers of three
universities to perform flawed experiments and allow the firm to use the results in
advertising. I believe there was no intent on the part of the researchers to mislead
the company or its customers. It was simply incompetence.
The researchers involved were not Ph.D. scientists and might not have been expected to understand all the nuances of experiment design. They were, however,
authorized to conduct independent research. Whether it is unethical to be incompetent or for a university not to have some safeguards against incompetence of its
researchers is one of the issues in the gray area.
PRACTICAL ETHICS IN AGRONOMIC RESEARCH
2. Proposal Budgets
Research is an open-ended process. There are almost always great uncertainties
about the future of any research effort. Researchers want to make sure there are
plenty of funds to support a research effort, be able to do the research thoroughly
and accurately, and deal with contingencies along the way. They want funds to provide continuity and stability so that they are not constantly expanding and contracting the project, which is wasteful, inconvenient, frustrating, and tends to fragment the effort. A researcher’s organization needs to recover indirect costs, one
way or the other.
Sponsors, on the other hand, need to keep costs to a minimum. They want to be
assured that, within reasonable limits, they are only paying for the research contracted for and not for other research that may be of interest to the researcher or
hisher organization. They definitely do not wish to pay for other activities or costs
of the researcher’s organization, such as publicity and fund-raising, unless they
have given their approval for such expenditures.
Budget numbers should be the result of careful, objective analyses of the costs
of achieving mutually agreed-on objectives within the agreed-on timeframe. As
much as is possible given the relative unpredictability of the future of a research
project, contingencies should be clearly identified and the methods to deal with
them should be explicit. It is unethical and almost always counterproductive for
either a researcher or a sponsor to try to mislead the other in budget negotiations.
If a budget represents a researcher’s best effort to estimate the costs of proposed
research and the sponsor is unable to invest enough to cover the costs, the project
may be scaled back, its scope diminished, or its proposed duration lengthened. Either the researcher or the sponsor may decide that the project is not financially feasible. As in most such situations with considerable gray area, intellectual honesty
and integrity are of paramount importance.
3. Indirect Cost Recovery
As a community of agricultural research scientists and administrators, we have
failed to come to grips with the ethical dimensions of indirect cost recovery.
Given the internal and external conflicts of interest involved and the potentially high
cost of adhering strictly to rules and protocols, we have handled this issue loosely.
Research incurs indirect costs, such as the costs of utilities, depreciation of facilities and equipment, administration, and various support services. These costs,
although indirect, are real and must be recovered from some source.
Most institutions, organizations, agencies, and private firms have indirect cost
policies that guide decisions about what or how much indirect cost they will agree
to pay when they sponsor research and how much they will try to recover from
outside sponsors. Government-audited rates for many universities are between 40
and 70% of modified total direct costs of projects. Private firms often have indirect cost rates exceeding 100%.
For obvious reasons, universities do not ask donors and benefactors to pay additional sums to cover indirect costs incurred when gifts are used to support research. Universities have considerable discretion in their use of gifts. Donors may
specify broad areas they wish their gifts to support but ordinarily would not specify in detail how the gift should be spent. In many cases, an institution can use gift
income to pay for items usually classified as indirect costs.
In the case of grant or contract research, however, activities, objectives, and associated direct costs are specified in proposals, agreements, and contracts. Money
provided to cover direct costs cannot be used to pay indirect costs. Universities
deem it appropriate in those situations to insist on full indirect cost recovery from
Intermediate indirect cost rates may be negotiated in individual cases, depending on how large the grants are, whether or not the university stands to gain royalties, the prospects for future gifts or grants, in-kind contributions by sponsors,
etc. Institutions may establish relatively low “internal” indirect cost recovery rates
for other public or quasi-public institutions, organizations, or agencies.
A university researcher may encourage a sponsor to provide research support in
the form of a gift rather than a grant or contract. In that situation, no indirect costs
are recovered and the entire gift may go directly into the researcher’s program to
cover the direct costs of research. There may be a tacit agreement between the researcher and the sponsor specifying the research to be done, the deliverables, and
timeframes. Because the university does not know the tacit contract exists, it does
not require indirect cost recovery.
When these tacit contracts are funded with gifts, the indirect costs incurred by
the research are shifted to other parts of the university’s budget. In effect, the money comes out of other research and education programs, without the approval of
those managing, conducting, and sponsoring those programs. Increasingly, universities are imposing internal surcharges on gift accounts to help offset indirect
costs incurred when gifts are used for research.
Knowingly entering into such tacit contracts for the purpose of avoiding indirect costs is clearly unethical behavior on the part of both researchers and sponsors. It is ethically incumbent on researchers, sponsors, and others directly involved in negotiations on gifts, grants, and contracts to be open and honest about
what is expected of each participant and what restrictions are being imposed on
use of sponsors funds.
The ethical gray areas include negotiated rates and situations in which sponsors
make zero or inappropriately low indirect cost recovery a condition of their sponsorship. The most prominent example is the legislative mandate that the federal
government will pay only 17% of direct costs to cover indirect costs associated
with USDA competitive grants.
The result of this policy is that other sponsors, particularly state governments,
PRACTICAL ETHICS IN AGRONOMIC RESEARCH
are required to share the unrecovered indirect costs of USDA competitively funded projects. Other federal research-sponsoring agencies pay full audited indirect
cost rates. It is appropriate to ask if it is ethical for a government agency or institution to accept or provide grants with arbitrary limits on indirect cost recovery. It
would be better if sponsors would simply look at the bottom line and decide if they
are receiving the best value for their investment.
Scientists are asked to provide peer reviews of proposals submitted to granting
agencies and manuscripts submitted for publication in scientific journals. Program
managers and journal editors are responsible for selecting reviewers with expertise and experience in the subject matter addressed by a specific proposal orjournal article. Reviewers are usually asked to decline the invitation to review a proposal or project if they think they do not have the required expertise, experience,
time, or other resources required to generate a thorough, useful, and discriminating review.
It is unethical for a scientist knowingly to undertake or complete a review if
he/she lacks the necessary qualifications. It is also unethical, in my opinion, to provide a cursory, superficial review. In the gray area are reviews for which the reviewer is only partially qualified and reviews that are unnecessarily critical, laced
with sarcasm and/or personal innuendos, or simply incompetent. Not surprisingly, the latter characteristics are often found in the same review.
5. Authorship and Shared Recognition
It is clearly unethical for a scientist to claim to author something he/she did not
author. This is plagiarism, which joins falsification and fabrication of data as the
unforgivable sins of scientists. Authorship in this context requires an author to
make a meaningful contribution to the research being reported.
Gratuitous authorship, that is, including people in a list of authors for the sole
purpose of rewarding them for past favors or attaching more credibility to research
than it otherwise would have, is unethical. Intentionally refusing or failing to include or at least acknowledge an author who made a meaningful contribution is
likewise unethical. Needless to say, the gray area is in the interpretation of meaningful in any specitic situation.
Perhaps this is best illustrated with an example. One of my colleagues is convinced that another colleague deliberately authored false statements in a peerreviewed article, intending to mislead the readers into placing more confidence in
some data than is justified. He asked me if he should write a letter to the editor of
the journal exposing this misconduct. I suggested that we hold a closed, internal
hearing in which the accuser and the accused could make statements to a panel of
their peers. That group could decide what steps to take next.
I am concerned that such a serious accusation, publicized internationally before
it can be thoroughly investigated, can devastate a person’s career, whether or not
it is true. To further complicate matters, the accuser does not wish to confront the
This situation raises a number of the questions that are typical of the issues encountered in whistle-blowing. Is it ethical for this person who has become aware
of serious misconduct not to reveal the misconduct? Is it ethical for the accuser to
make accusations to others but refuse to confront the accused in an appropriate forum, thus giving the accused a chance to defend himself?
Given a long history of personality conflict between these two individuals, are
the accusations the result of an objective analysis by the accuser? Is it ethical for
me, having been provided somewhat convincing evidence that some wrongdoing
is involved, to let this matter drop without taking some kind of action to explore the
situation further, even without the cooperation of the accuser? Is it ethical of me to
reveal things that were revealed to me in confidence, knowing that the accuser does
not wish this information revealed? The answer to all of these questions is no.
I must weigh the potential good and bad consequences of the alternative courses
of action. Some pertinent questions include the following: Could this apparent
wrongdoing simply be a mistake that could be corrected by the accused writing a
letter of explanation to the journal? What harm is done to this person and his students if he is guilty of wrongdoing, is not confronted, and is allowed to continue
the wrongdoing? What effect would a serious, open confrontation on this issue
have on morale and the image of our institution? Will having wrong information
in the literature cause problems for other scientists or practitioners? How much administrative and faculty time can we afford to spend on this particular incident?
A potential conflict of interest arises when a scientist is in a position to use
hisher position and the influence associated with it for personal gain beyond the
contractual compensation associated with the position or to benefit disproportionately within hisher organization or system. Whether or not an actual conflict exists depends on several factors.
It is unethical for a scientist to let outside relationships prevent himher from
fulfilling contractual obligations to hisher employer. It is unethical for a scientist
to attempt to manipulate competitive processes unfairly in hisher own favor or in
favor of groups to which the faculty member is attached or obligated in some way.
PRACTICAL ETHICS IN AGRONOMIC RESEARCH
It is not unethical for a scientist to receive shares of royalties or other compensation as a result of contractual relationships between the investigator’s organization and other organizations. It is not unethical for a scientist to own, operate, or
otherwise participate in a business, so long as the relationship does not prevent the
scientist from meeting contractual obligations to his organization and moral obligations to its constituents.
Employers may ask employees to reveal their relationships with outside organizations that compensate them more than a specified amount. This allows the employer to investigate and decide if a conflict of interest exists.
Conflicts of interest are often subtle and virtually undetectable from the outside.
Individuals know best when they have a conflict of interest. Ethical behavior includes revealing potential conflicts of interest and behaving in an even-handed and
objective manner even when potential conflicts of interest exist. A person who has
the level of intellectual honesty and integrity required to be a good scientist should
be able to think and act objectively even when hisiher own interests are at stake.
I have always been frustrated when conflict of interest rules require certain people on a proposal review panel to leave the room. The rules often preclude people
from the same institution or even the same state as an applicant to refrain from participating in a review of the applicant’s proposal. In my experience, this usually
has the effect of barring from the discussion the people who know the most about
the proposal and the applicant. It also suggests that other panel members are so
naive that they do not know when they are being manipulated.
A problem is created in a proposal review process when more detailed information is available about some projects than others. This is given as another reason for barring those with a potential conflict from the review. An “ignorance is
bliss” argument usually raises a flag with me, but I can understand the concerns of
those who take this position.
Increasingly, public institutions and agencies are expanding and improving systems to identify and protect intellectual property generated in their research operations. This represents a significant cultural as well as administrative change in
many public institutions, particularly universities.
A few decades ago, the most common attitude in universities and especially in
colleges of agriculture was that intellectual property generated with public funds
belonged to the public and should be freely available to any member of the public
who wished to see or use it. Now, universities see their intellectual property as being owned by the public, but they feel a responsibility to manage it in the best interests of the public.
Transferring intellectual property to the public domain, thus making it freely
available to all members of the public, is not always in the public’s best interest.
Sometimes, in order to assure that useful new information and technology will be
commercialized and thus benefit the public, it is necessary to give some individual or group proprietary access to it.
Often, intellectual property generated by public institutions and agencies takes
the form of prototype products, services, or information. More investment, research and development, and marketing is often necessary before such prototype
technology can be successfully commercialized. By licensing discoveries and inventions exclusively or semiexclusively to interested and capable firms or individuals, public institutions provide incentives for further investment in commercializing these developments.
Public institutions use patents, copyrights, and trade secrets to protect intellectual property from being appropriated by unauthorized individuals. The management, including sale and licensing, of intellectual property involves several ethical issues.
1. Patent and Copyright Infringement
It is illegal and unethical to commercialize patented material or duplicate, disseminate, or commercialize copyrighted intellectual property without the permission of the owner of the patent or copyright. This unethical behavior is similar to
plagiarism. The law has never been specific on whether it is illegal to reproduce a
patented item in order to conduct research on that item. Differences of opinion on
whether patents or copyrights have been infringed upon are usually settled through
litigation. Often the technical issues are quite complex. As more cases are litigated, attorneys and courts gain experience, and clearer precedents are set.
2. Ethics and Trade Secrets
Until relatively recently, trade secrets were not protected, but laws now have
words to the effect that if a secret is “sufficiently secret,” it is protected, much like
a patent. Sufficiently secret means that the information or material has not been
broadly disseminated. Also, those with whom it has been shared should know that
they were not granted commercialization rights nor are they free to pass the information or materials to others. Trade secrets are not protected from the possibility that someone else may rediscover or reinvent the secret technology.
Items of germplasm and other genetic material may be patented but more frequently are treated as trade secrets. Increasingly, institutions insist that people requesting and receiving germplasm from scientists sign material transfer agreements. These are usually papers acknowledging that the recipient has not been
granted rights to commercialize the material. Recipients are not allowed to pass
the material to others without permission of the owner of the trade secret.