Scientific Misconduct in Environmental Science and
Toxicology
(JAMA. 1994;272:168-170)
Herbert N. Nigg, PhD, Gabriela Radulescu
Scientific misconduct easily occurs in environmental science
and toxicology; we encountered four such cases. Only one case was
discovered by editors; three were reported by other authors. All guilty
authors were eventually banned from future publication in the
Bulletin of Environmental Contamination and Toxicology. Cases
in the Bulletin indicate that scientific misconduct may occur
undetected across phyla, genera, and species; that distance from the
publishing source makes detection more difficult; that editors and
reviewers are not organized to take action against scientific
misconduct; that plagiarized authors are likely to report plagiarism;
and that there is only a small risk of censure from any source for
authors engaging in scientific misconduct.
(JAMA. 1994;272:168-170)
SCIENTIFIC MISCONDUCT is certainly not a new
phenomenon; it was recognized during peer review as early as
1731.[1] Unethical practices in scientific publishing range
from lack of acknowledgment of sources, duplicate publication, and
plagiarism to fabrication of data (New York Times. February
25, 1986).[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
Peer review is the first defense against
scientific misconduct, but numerous improper studies are published
despite peer review (New York Times. February 25, 1986, and
Time. April 25, 1993:59-60).[2,6,7,8,9] [10] [11]
A book on the
subject[12] has been published recently, and
at least one editorial asks if academic institutions are corrupt.[13] The
editors of the biomedical journals have led efforts to minimize
misconduct in the scientific literature.[1,2]
Editors and reviewers of peer-reviewed publications worldwide,
traditionally entrusted with judging only the quality of the scientific
work submitted for publication, are increasingly faced with having to
judge the integrity of the authors proffering it. The task of
gatekeeping to ensure the highest quality of science implies, aside
from scientific expertise, a probing, skeptical mind. Effective
policing, however, requires a rather different set of skills and
perhaps even a predisposition to consider all authors "suspect"
until a thorough check of the literature establishes the honest
authors. This is a difficult role to assume in an environment so
fundamentally based on trust and is, in part, why most editors are not
comfortable confronting a colleague with an accusation of scientific
misconduct. Editors are, in such cases, called on to establish the
"means, motive, and opportunity."
There is substantial agreement in discussions about scientific
misconduct that the major motive for misconduct is what
Huth[5] described as "authorship [as] a meal ticket."
As he noted 10 years ago, "academic appointment, advancement, tenure,
and the associated economic rewards may hinge upon the length, if not
the quality of one's bibliography."[5] Nothing has changed
in this respect during the past decade, and it is uniformly true across
all the disciplines of science. The means and opportunities for
scientific misconduct, however, vary from case to case.
Herein, we report four cases of scientific misconduct encountered
among manuscripts submitted for publication in the Bulletin of
Environmental Contamination and Toxicology to illustrate the ease
with which unethical scientific behavior can occur in the fields of
environmental contamination and toxicology.
The Bulletin was established in 1964 for the purpose of
publishing scientific studies that could be presented in eight
single-spaced pages. The Archives of Environmental Contamination
and Toxicology and Reviews of Environmental Contamination and
Toxicology (originally entitled Residue Reviews) are
companion journals. The Bulletin publishes articles about
analytic methods, aquatic toxicology-metals, aquatic
toxicology-organics, environmental distribution and human exposure,
general toxicology, metabolism and biochemistry, and wildlife
toxicology. Initially, manuscripts submitted to the Bulletin
underwent an editorial review. For the past 10 years, however,
manuscripts submitted to the Bulletin have been peer reviewed
by an associate editor in each subject area and a 120-member editorial
board. The editorial board members in each area provide reviews.
Non-editorial board scientists review as well. The rejection rate for
submitted manuscripts is about 50%. Manuscripts are rejected primarily
because of flawed science.
During the last 10 years, the Bulletin has occasionally
received manuscripts containing previously published data. These
manuscripts were rejected because the data had been previously
published, and the authors were warned not to participate in unethical
practices. Numerous instances of "salami publication," slicing one
study into several manuscripts, have been identified. These authors are
asked either to combine the manuscripts into one or to withdraw the
manuscript(s). However, since 1990, four cases of scientific misconduct
have been discovered among manuscripts submitted for publication in the
Bulletin. These are discussed herein in the order of their
discovery.
The first case was discovered by B. W. Wilson, PhD,
professor, University of California-Davis, Bulletin associate
editor in the area of wildlife toxicology. Dr Wilson noticed striking
similarities between an article published in the Bulletin in
1989 and a manuscript submitted for publication in 1991. Comparison of
the 1989 data tables with the 1991 data, which were presented as
several tables, revealed that the data in the two manuscripts were
identical. The senior author had changed his original data, usually in
the hundreds place, changed the title, dropped authors, added one
author, and submitted self-plagiarized data for peer review. The
original article was not referenced.
The two authors of the article were sent three successive letters. The
first asked for an explanation. The second insisted on an explanation.
The third banned them from ever again publishing in the Bulletin.
Ironic coincidences occurred during this incident. An additional
manuscript by the senior author was undergoing review when the
plagiarism was discovered. The original article and the plagiarized
manuscript concerned the characteristics of rat blood after pesticide
treatment. The article in review concerned fish blood characteristics after
pesticide treatment. All three articles contained
identical data as well as nearly identical data. After receiving the
third letter, the senior author replied asking that his coauthor be
absolved of any responsibility because he (1) had been out of the
country, (2) had no knowledge of what was done, and (3) had many
graduate students who would be hurt by the banning. The coauthor then
submitted an article with seven coauthors before a reply from the
Bulletin editors could be sent. It appears that the coauthor
is the first author's supervisor. The article with seven coauthors was
rejected. During the investigation, an additional article on mollusks
with one of the plagiarizing authors' names as coauthor was accepted
into the Bulletin peer-review process. This article was also
rejected because of previous scientific misconduct.
A reply was never received from the coauthor (perhaps supervisor). This
is a flagrant case of remanufacturing data, even across phyla and
species, and submitting for publication. The authors' institution was
in India.
The second case was brought to our attention by authors who had
published an article in the Bulletin in 1986. A similar study
was published in the Bulletin in 1993. Approximately 80% of
the 1986 article was copied verbatim into the 1993 article. The 1993
authors were contacted by the Bulletin editors. They explained
that they had difficulty with written English and, consequently, copied
large sections from the 1986 article into their 1993 manuscript. In
fact, the 1993 article did contain a few different references, and the
data were presented differently and may have been original. However,
the "Materials and Methods" section was copied verbatim from the
1986 article, except for a few words. No scientist exactly reproduces
another's methods. This led us to suspect that the 1993 article was
not only a copy of the 1986 article but that the data were possibly
manufactured as well. The original 1986 article was not referenced in
the 1993 article, but other articles by the 1986 authors were cited.
The 1993 authors were banned from again publishing in the Bulletin.
These authors' institution was also in India.
The third case was brought to our attention by the original
author, who had published an article in Spanish in 1990 in a Mexican
journal. The Bulletin published an article in 1993 that turned
out to be an English translation of the 1990 Mexican article and also
contained falsified data. The first author of the fraudulent article
was confronted by the plagiarized author and signed a confession that
was forwarded to the Bulletin. Very strong letters were mailed
to the first author, the coauthor, the president of the university, and
the department head.
No replies were received from the authors. The department head and
president gave essentially the same reply. They objected to the
strength of the letters. They believed substantial responsibility lay
with the journal and the review process. They related that the first
author "was being separated" from the university. Both the
department head and the university president ignored a request for
their written policy for cases of scientific misconduct. The
implication that "separation from the university" was disciplinary
was untrue. The first author retired as scheduled. The authors'
institution was in Mexico.
The fourth case was brought to our attention by a former graduate
student, now a PhD. A professor in the student's former department and
not her adviser published five figures from her thesis in the
Bulletin in 1993. These were published without permission and
without acknowledgment of the source. A strong letter was sent to the
author banning him from further publication in the Bulletin.
The author who stole the figures replied with a letter that stated that
he and the former student are friends and that it was merely an
oversight not to acknowledge the source of the figures. Queries were
made to the president of the university, the head of the funding
agency, and the department head. The department head answered with an
apology and the fact that no written policy for scientific misconduct
existed in his organization. Regardless, this author was banned from
publishing again in the Bulletin. The institution of the
author who stole the figures is in India.
These four cases of scientific misconduct have several elements in
common. Plagiarized articles were not referenced. Each first author
admitted their plagiarism and misconduct in writing. Each case was from
a non-US university. In three cases, the authors were arrogant in their
excuses for the event. In one case, the author begged to be solely
responsible. Two cases appeared to be part of a long-standing pattern
of scientific misconduct by the first author.
Our cases indicate that (1) scientific misconduct may occur undetected
across phyla, genera, and species; (2) results are difficult to
evaluate, let alone duplicate, in cases in which the data are only
available halfway across the globe; (3) editors and reviewers in the
field have not yet communicated, organized, and taken concerted action
on scientific misconduct; (4) plagiarized authors are likely to report
scientific misconduct; and (5) the risk of discipline to the fraudulent
authors from their organization, funding agencies, or any other source
is small.
In addition to banning authors from publication, the heads of the
departments, the funding agencies, and the presidents of the
universities were asked for their written policies on scientific
misconduct. Funding agencies did not respond. One department head
admitted that no written policy existed. Although replies were received
from the other parties, the question of scientific misconduct policy
was ignored. There is no policy within these organizations to deal with
scientific misconduct, in agreement with the 1990 study of
Nobel.[14]
All authors of these manuscripts and articles were placed on a
"List of Shame" after review of each case by all Bulletin
editors. This internal list is distributed to our editors and editorial
board with instructions that these authors are not to publish and are
not to be referenced in articles published in the Bulletin.
This policy applies by editorial agreement to all cases regardless of
source and regardless of editorial opinions on disadvantaged authors
and other extenuating circumstances.
How widely should the details of scientific misconduct be distributed?
It was suggested by the Bulletin editors, for example, that
the list be distributed to all authors submitting manuscripts and also
to editors of journals with similar subject matter. The legal staff of
Springer-Verlag, New York, NY, vetoed these suggestions. According to
the legal staff, the actions that the Bulletin editors took
were as much as could be done to correct these situations.
Scientific misconduct demands strong action by the employer and
by journal editors. From our experience and the experience of
others,[13] disciplinary policies are not in place, and
organizations would rather keep these cases from public scrutiny. There
is also no systematic way to retract scientific
literature,[14] [15] although the number of citations of
invalid studies declines from the number expected once the scientific
misconduct is publicized.[14]
The American Journal of Public Health has a written policy to
deal with scientific integrity.[16] Most journals, including
the Bulletin, do not have policies concerning scientific
misconduct clearly published in each issue. [15] We can hardly
criticize others for lack of action when, in general, most journal
editors have taken no action themselves. We agree with
Nobel[14] that the scientific misconduct issue should be
addressed openly, intensively, and head on. We also agree with
Friedman[17] that scientific journals have an obligation to
develop written policies and to respond to allegations of scientific
misconduct and unethical publishing practices.
From the Citrus Research and Education Center, University of
Florida, Lake Alfred (Dr Nigg), and Journals Division, Springer-Verlag,
New York, NY (Ms Radulescu).
Presented at the Second International Congress on Peer Review in
Biomedical Publication, Chicago, Ill, September 11, 1993.
Florida Agricultural Experiment Station Journal Series No.
R-03734.
Reprint requests to Citrus Research and Education Center, University of
Florida, 700 Experiment Station Rd, Lake Alfred, FL 33850 (Dr Nigg).
References
1. Kronick DA. Peer review in 18th-century
scientific journalism. JAMA. 1990;263:1321-1322.
2. Rennie D, Gunsalus CK. Scientific misconduct: new
definition, procedures, and office---perhaps a new leaf.
JAMA. 1993;269:915-917.
3. Lock S. Repetitive publication: a waste that must stop.
BMJ. 1984;288:42-43.
4. Lock S. Notice of inadvertent repetitive publication.
BMJ. 1984;288:43.
5. Huth EJ. Responsibilities of coauthorship. Ann
Intern Med. 1983;99:266-267.
6. Sojka RE, Moon HW. Ethical issues of concern in the
United States Department of Agriculture-Agricultural Research Service.
In: Kral DM, ed. Research Ethics, Manuscript Review and Journal
Quality. Madison, Wis: American Society of Agronomy; 1992:chap 4
(pp 25-33).
7. Sojka RE, Mayland HF, Gbur EE. American Society of
Agronomy member experiences and perceptions of the peer
reviewing-editing process. In: Kral DM, ed. Research Ethics,
Manuscript Review and Journal Quality. Madison Wis: American
Society of Agronomy; 1992:chap 4:35-53.
8. Williamson RCN, Farndon JR, Murie JA, Johnson CD.
Editors' announcements. Br J Surg. 1992;79:3.
9. Bard AJ. Misconduct in scientific publishing. Chem
Eng News. May 31, 1993:44, 54-55.
10. Zurer P. Chemist fights scientific misconduct charge.
Chem Eng News. August 16, 1993:5.
11. Macdonald JR. What's wrong with these publications?
or, some pathologies of scientific authorship. Physics Today.
August 1993:51-52.
12. LaFollette MC. Stealing Into Print: Fraud,
Plagiarism, and Misconduct in Scientific Publishing. Berkeley:
University of California Press; 1992:294.
13. Are academic institutions corrupt? Lancet.
1993;342:315-316. Editorial.
14. Nobel JJ. Comparison of research quality guidelines in
academic and nonacademic environments. JAMA.
1990;263:1435-1437.
15. Pfeifer MP, Snodgrass GL. The continued use of
retracted, invalid scientific literature. JAMA.
1990;263:1420-1423.
16. Susser M, Yankauer A. Prior, duplicate, repetitive,
fragmented, and redundant publication and editorial decisions. Am
J Public Health. 1993;83:792-793.
17. Friedman PJ. Correcting the literature following
fraudulent publication. JAMA.
1990;263:1416-1419.
Table of Contents
