Jeffrey K Aronson

BMJ doi: https://doi.org/10.1136/bmj.q2188 (Published 04 October 2024) Cite this as: BMJ 2024;387:q2188

BMJ doi: https://doi.org/10.1136/bmj.q2250 (Published 11 October 2024) Cite this as: BMJ 2024;387:q2250

Although the concept of academic integrity has been long with us, and violations of it not uncommon, formal research into the topic began only in the 1990s, particularly by a pioneer in the field Donald McCabe of Rutgers University, whose main interest lay in the US collegiate system. In recent years concern about academic integrity, or rather the lack of it, has increased markedly, particularly in the area of research integrity. Having failed to find a satisfactory definition of academic integrity, I propose the following: uncorrupted moral virtue in relation to truth, uprightness, honesty, and sincerity in the pursuit of research, education, and scholarship. Several codes of conduct outline what is expected of academic researchers, of which the European Charter for Researchers is an example. A companion publication, the European Code of Conduct for Research Integrity, which deals in greater detail with the principles outlined in the European Charter for Researchers, and which itself refers to many other codes of practice, also outlines a range of possible violations of academic integrity, i.e. matters of academic dishonesty or misdemeanours. These include fabrication, falsification, and plagiarism, misuse of statistics, concealing the use of AI, withholding information, and selective or inaccurate citation. Other matters of concern include predatory journals, paper mills, and the large increase in the numbers of papers being retracted after publication.

To understand how academic science can be violated we must first understand the norms that determine how it should be successfully prosecuted. The work of the sociologist Robert K Merton is informative. In the 1930s and 1940s he described, in three related essays, the ways in which academic norms had developed and what they were. In the first essay, he described what he supposed had been the influence of Puritanism in the 17th century, while acknowledging that the Protestant religious ethic was only one factor in the development of rational and empirical science. In the second essay he focused on factors that might induce hostility against science, leading to the enunciation of some academic norms: scientific autonomy (preventing contradictory influences from other spheres), integrity (the appreciation of which enhances the likelihood that scientific results will receive public acceptance in the absence of complete understanding of how they were achieved), the use of comprehensible language in describing outcomes and their interpretation, and the restriction of academic skepticism to scientific thought. In the third essay he outlined other norms, based on the cultural values and mores that govern scientific activities: universalism, communality (which he called communism), disinterestedness, and organised skepticism; to these we can add appropriate curiosity and restraint of excess enthusiasm. In addition to these, there are norms relating to the methods that are used to obtain evidence and to the methods used to analyse the knowledge thus obtained. Together these constitute a scientific ethic and imply a scientific conscience.

Academic norms include: preservation of academic autonomy; integrity in academic practices; a priori formulation of hypotheses and publication of protocols; the use of appropriate methods by which evidence is obtained; the use of appropriate methods for interpreting the accumulated knowledge so obtained; the need to take heed of the possible consequences, both intended and unintended, of research; the use of comprehensible language in describing outcomes and their interpretation; universalism; communality; disinterestedness; organised skepticism, but limited to one's own sphere of academic interest; appropriate curiosity; avoidance of inappropriate enthusiasm; and respect for the work of others. Violation of all of these is possible, and different types of violations are known by different terms: “research misconduct” refers to three major violations, fabrication, falsification, and plagiarism; I prefer to call these collectively by the legal term “felonies,” emphasising their gravity and importance. Other violations are known as “questionable research practices,” which I prefer to call by the stronger term “misdemeanours,” again emphasising their importance, although they are less grave than felonies. Interest in detecting, reporting, and discussing these violations has increased markedly since 1990, but I have found only two systematic reviews, one of which, disappointingly, but perhaps not unexpectedly, shows that there is little useful evidence on the subject, and a second, which shows that the likely prevalences of both types, felonies and misdemeanours, are higher than one would want, although probably estimated at lower than they truly are.

Fraudulent academic activities that I have referred to as “felonies” comprise fabrication (defined by the US Office of Research Integrity as “making up data or results and recording or reporting them”), falsification (“manipulating research materials, equipment, or processes, or changing or omitting data or results such that the research is not accurately represented in the research record”), and plagiarism (“the appropriation of another person's ideas, processes, results, or words without giving appropriate credit”). Many cases of fabrication and falsification reported in the 1970s increasingly drew attention to this, and none was perhaps more frequently referred to and discussed than the Darsee Affair, which involved extensive fabrication and falsification of research data by John R Darsee, working in Eugene Braunwald's laboratory in Boston. Several of Darsee's papers were subsequently retracted, but they continue to be cited as if they had not.

Of the various types of academic misdemeanours and felonies, perhaps plagiarism is the most difficult and complex to study and write about. Even defining it is difficult. The US Office of Research Integrity (ORI) defines plagiarism as “the appropriation of another person's ideas, processes, results, or words without giving appropriate credit,” but how do you define when it has occurred? What are the quantitative and qualitative thresholds? What are the different types of plagiarism? Why does plagiarism occur? How does one detect it? What is self-plagiarism and should we even call it that? There is an enormous literature on the subject and it needs to be tamed; an annotated bibliography would help, if anyone would create one. Perhaps they already have?

Estimates of the frequencies of the different types of academic felonies (research misconduct), fabrication, falsification, and plagiarism, are in single figures, at between 2% and 5% of all publications. Misdemeanours are more common. However, these figures are probably underestimates of the true frequencies, and even if the numbers seem relatively low, felonies are likely to have marked effects on research results. There is, for example, some evidence that they may contribute to the problem of poor reproducibility of research results. Furthermore, the quality of the data included in meta-analyses of published studies can certainly affect the conclusions of the analyses, as the example of ivermectin in the treatment of covid-19 shows.

BMJ doi: https://doi.org/10.1136/bmj.q2817 (Published 13 December 2024) Cite this as: BMJ 2024;387:q2817