Scientific Misconduct in Management Research
| Nicolai Foss |
Fraudulent behavior in research is the ultimate academic gossip. It is hardly surprising that our post on Thomas Basbøll’s claim that management theory heavyweight Karl Weick has engaged in plagiarism (here) was one of O&M’s most popular posts in 2010. One of my own papers was once directly copied. All that was changed was the front page. In one of those strange coincidences, the journal editor asked my co-author to review the paper. The plagiarist was a consultant, not an academic, so it is possible that the case had no consequences for him.
How prevalent is scientific misconduct in management research? And how strongly should we care? After all, what gets published in the management journals does not have the same direct impact as what gets published in the medicine journals, or what the UN’s Intergovernmental Climate Panel utters. While management research may not cure cancer, it likely has considerable impact on resource allocation, and therefore on what is available for curing cancer. Moreover, there are strong externalities: A reputation for “bad science” in one field or discipline may easily spill over to other fields and disciplines. Hence, misconduct should be regarded with as severely in management research as in other fields and disciplines.
With respect to the incidence of fraudulent research behavior, rather little is known. While fraud in, particularly, medicine tends to draw major headlines in the press, I cannot recall anything similar in the case of management research. It seems unlikely that management researchers should be significantly more honest than researchers in medicine, so our lack of knowledge in this seems troublesome. In ”Management Science on the Credibility Bubble: Cardinal Sins and Various Misdemeanors,” recently published in the Academy of Management Learning and Education, Arthur G. Bedeian, Shannon G. Taylor, and Alan N. Miller present evidence that research misconduct is quite a prevalent phenomenon. Briefly, they collected data from faculty in 104 PhD-granting management departments in the US. Questions identified “eleven different types of questionable research conduct, including data fabrication, data falsification, plagiarism, inappropriately accepting or assigning authorship credit, and publishing the same data or results in two or more publications.”
Some of Bedeian et al.’s examples of “questionable research conduct” seem somewhat open to interpretation and questioning (e.g., “developing ‘ins’ with journal editors” — in fact, the initiative for such “ins” often emerge from the editor side; “published the same data or results in two or more publications” — presumably, there is nothing necessarily wrong with publishing “the same data … in two or more publications”), and the procedure of asking faculty to indicate their “knowledge of faculty engaging in” research misconduct is questionable, as different faculty may relate to the same episode of research misconduct (they acknowledge this problem). Still, the numbers are quite striking. More than 70% reports knowledge of cases of not giving due credit to originators of ideas (i.e., plagiarism). Even more report knowledge of data manipulation, although only (?) 27% report knowledge of outright data fabrication.