Research misconduct definitions adopted by U.S. research institutions.

In 2000, the U.S. federal government adopted a uniform definition of research misconduct as fabrication, falsification, or plagiarism (FFP), which became effective in 2001. Institutions must apply this definition of misconduct to federally-funded research to receive funding. While institutions are free to adopt definitions of misconduct that go beyond the federal standard, it is not known how many do. We analyzed misconduct policies from 183 U.S. research institutions and coded them according to thirteen different types of behavior mentioned in the misconduct definition. We also obtained data on the institution's total research funding and public vs. private status, and the year it adopted the definition. We found that more than half (59%) of the institutions in our sample had misconduct policies that went beyond the federal standard. Other than FFP, the most common behaviors included in definitions were "other serious deviations" (45.4%), "significant or material violations of regulations" (23.0%), "misuse of confidential information" (15.8%), "misconduct related to misconduct" (14.8%), "unethical authorship other than plagiarism" (14.2%), "other deception involving data manipulation" (13.1%), and "misappropriation of property/theft" (10.4%). Significantly more definitions adopted in 2001 or later went beyond the federal standard than those adopted before 2001 (73.2% vs. 26.8%), and significantly more definitions adopted by institutions in the lower quartile of total research funding went beyond the federal standard than those adopted by institutions in the upper quartiles. Public vs. private status was not significantly associated with going beyond the federal standard.