Deception And Fraud In The Publication Of Scientific Research: Are There Solutions?

A number of events in the U.S. and abroad have refocused the scientific community on historical issues of whether, and how, integrity of our technical literature can be assured. Solutions to this problem are neither simple nor certain. Professional societies have addressed scientific misconduct, and effective responses by the research community will require cooperation of scientific publications. While the incidence of scientific fraud is difficult to estimate with precision and certainly varies with discipline, identified and publicized recent cases beg attention from editorial boards. Several egregious cases are described. The peer review system serves the function of examination and critique by scientists in relevant disciplines to assess submitted papers prior to publication. There is even a developing literature and several specific journals dedicated to the subject of fraud, professional integrity and ways to monitor or correct existing conditions. Underlying the field of professional and scientific publication is a fundamental assumption that data are real and that research actually occurred. Typically, the process is “blind” in both directions, although some journals permit “author-directed” reviews. A reviewer’s responsibilities include ensuring that text properly reflects data, that tables and figures are necessary/appropriate, and that conclusions fairly and reasonably reflect results and the body of information. Thus, existing peer review systems probably cannot detect anything but the most obvious fraud. In addition to imposing or perpetuating stringent review protocols, journals also can amend author guidelines to speak explicitly about publishing requirements. Cases of properly documented fraud warrant immediate public announcement, followed by official withdrawal or retraction. Reflection on these issues led editors of one journal to institute changes in editorial policies and develop a code of ethics for authors, reviewers, and editors. Prevention of dishonest research is already difficult, and we should ensure that this remains the case. Editors should formally commit reviewers/authors to ethical conduct in technical publications prior to publication and review