In reply a letter to the editor by Tim Wysocki, Ph.D and R. Wayne Fuqua, Ph.D on "The Consequences of a Fraudulent Scientist on His Innocent Coinvestigators," Jama, 264(24) p.3145, December 26, 1990 (follows letter by Garfield and Welljams-Dorof)
Editor's Note. - Drs Wysocki and Fuqua draw attention to one of the worst consequences of scientific misconduct: damage to the reputations of the fraudulent worker's innocent coinvestigators. The editors believe that in the case under discussion, Breuning was the only per son who acted in any way dishonorably or offended against the traditions of good science. Indeed, it was due to the persistent and heroic efforts of one of Breuning's coworkers, Robert Sprague, that Breuning's misconduct first came to light. We also believe that no reader could reasonably interpret the article by Garfield and Welijams-Dorof as any reflection on the good names of Drs Wysocki and Fuqua or any other of Breuning's numerous coauthors.
The Consequences of a Fraudulent Scientist on His Innocent Coinvestigators
Eugene Garfield and Alfred Welljams-Dorof
We disagree with Wysocki and Fuqua's letter suggesting that our citation study of 2O cited publications by Dr Stephen E. Breuning in any way implied, intimated, or otherwise indicated that every one was fraudulent. We also disagree with their claim that the study in any way "impugned the scientific integrity" of the coauthors of these articles.
The publications that we studied were clearly and consistently referred to simply as cited publications by Breuning. No mention of any coauthors was made, nor were any valuations attached to these publications. The exceptions were three articles identified in Table 1 as having been formally retracted by their respective journals by the end of 1988.
The study purposefully did not draw these distinctions because the goal was to indicate the overall influence of Breuning's work-good, bad, or indifferent. A related purpose was to determine what overall effect public disclosure of Breuning's widespread fraud had on the citation of his work. Neither we nor JAMA'S editors and referees thought that our report unfairly characterized the publications discussed.
Of course, we agree with Wysocki and Fuqua that examples of scientific misconduct ought to be discussed with great sensitivity to coauthors who are innocent bystanders. Many of Breuning's colleagues, for example, were unaware that their names had been added to articles and grant applications. Also, as Wysocki and Fuqua point out, the NIMH Investigative Panel Report1 concluded that one of the articles2 in our study was carried out as reported and properly conducted. In the second example they cite,3 the panel confirmed that relevant tests and evaluations were performed, but concluded that "there are serious irregularities in the published reports" (Appendix 1, p 26) related to questions about whether medications were manipulated or placebos were used.
In fact, 15 other publications in our study were reviewed by the panel. It concluded that 11 of these reported studies were not carried out. Three others were deemed to be "scientifically unsound and misleading" in whole or in part because they relied so heavily on Breuning's discredited work. The panel could reach no conclusion on the validity of one article because Breuning's data could not be verified.In conclusion, to avoid possible misunderstandings about the method or purpose of our study, we wish to state explicitly that no valuation-good, bad, or indifferent-of Breuning's cited publications was intended, implied, or intimated. Thus, the study cannot fairly be interpreted as having impugned the scientific integrity of any coauthor listed.
Eugene Garfield, PhD
Institute for Scientific Information
1. back to textFinal Report: Investigation of Alleged Scientific Misconduct on Grants MH32206 and MH37449. Washington, DC: National Institute of Mental Health; April 1987.
2. back to text Wyzocki T, Fuqua RW, Davis VJ, Breuning SE. Effects of thioridazine (Mellaril) on titrating delayed matching to sample performance of mentally retarded adults. Am J Ment Defic. 1981;85:539-547.
3.back to text Davis VJ, Poling AD, Wysocki T, Breuning SE. Effects of phenytoin withdrawal on matching to sample and workshop performance of mentally retarded persons. J Nerv Ment Dis. 1981;169:718-725.
The Consequences of a Fraudulent Scientist on His innocent Coinvestigators
Tim Wysocki and R. Wayne FuquaTo the Editor. - In their recent citation analysis of articles authored or coauthored by Dr Stephen E. Breuning, Garfield and Welljams-Dorof1 have misled readers by implying that all of these studies are fraudulent. This misrepresentation is evident in several sections of their article.
Their quotations from the National Institute of Mental Health (NIMH) Investigative Panel Report2 that "none of the described studies of psychopharma cologic treatment had been carried out," "only a few of the experimental subjects . . . were ever studied," and "the complex designs and rigorous methods reported were not employed" erroneously intimate that these statements apply to all of the 20 studies under consideration in their article. In fact, these quotes were drawn from a secondary source3 that extracted these statements from sections of the NIMH report dealing with three psychopharmacologic treatment studies and with the panel's specific conclusions about Breuning's individual culpability. Other studies that were carefully scrutinized by the NIMH panel were found to have been conducted as described and to be free of evidence of fraud. Garfield and Welijams-Dorof irresponsibly failed to discuss these important exceptions to the panel's overall findings in the Breuning case.
We are familiar with two such incidents. With respect to the first of these studies,4 the NIMH panel "confirmed through its site visit to Coldwater and through interviews that this work was carried out as reported" (p 23) and the panel "confirmed that this study was carried out and found no information to suggest that it was conducted improperly"(Appendix 1, p 24). A recent published report5 confirmed the findings of this study in a systematic replication. With respect to the second of these articles,6 the panel identified contradictions regarding placebo administration, but nevertheless concluded that "the test and workshop performance evaluations had been carried out" (p 24) and "panel members confirmed that matching-to-sample tests were carried out. A site visit to the Evergreen School adjacent to Coldwater confirmed that staff there had recorded workshop performance" (Appendix 1, p 25).
The failure of Garfield and Welljams Dorof to acknowledge these exceptions to the NIMH panel's determination of widespread fraud in Breuning's work, their citation of a secondary source rather than the original NIMH report, their inclusion of all 20 of the publications together in Table 1, and their imprecise portrayal of the NIMH panel findings combine to imply that every article that Breuning appears on as an author or coauthor is fraudulent. In doing so, they have impugned the scientific integrity of several of Breuning's former colleagues who are without complicity in his scientific fraud. While we applaud the authors' efforts to provide a systematic analysis of the sequelae of scientific misconduct, this enterprise must occur with responsible, meticulous attention to documented facts and with appropriate sensitivity to the serious implications of these episodes for those honest scientists who are innocent bystanders. Neglect of these concerns will only corn-pound the tragic consequences of scientific fraud.
1. back to text Garfield E, Welljams-Dorof A. The impact of fraudulent research on the scientific literature: the Stephen E. Breuning case. JAMA. 1990;263:1424-1426. Tim Wysocki, PhD
Ohio State University
R. Wayne Fuqua, PhD
Western Michigan University
2. back to textFinal Report: Investigation of Alleged Scientific Misconduct on Grants MH32206 and MH37449. Washington, DC: National Institute of Mental Health; April 1987.
3. back to text Holden C. NIMH finds a case of 'serious misconduct.' Science. 1987;235:1566-1567.
4. back to text Wysocki T, Fuqua RW, Davis VJ, Breuning SE. Effects of thioridazine (Mellaril) on titrating delayed matching to sample performance of mentally retarded adults. Am J Ment Defic. 1981;85:539-547.
5. back to text Carpenter M, Cowart CA, McCallum RS, Bell SM. Effects of antipsychotic medication on discrimination learning for institutionalized adults who have mental retardation. Behav Residential Treatment. 1990;5:1O5-120.
6. back to text Davis VJ, Poling AD, Wysocki T, Breuning SE. Effects of phenytoin withdrawal on matching to sample and workshop performance of mentally retarded persons. J Nerv Ment Dis. 1981; 169:718-725.