are often left to the individual, who is expected to read through three, four, or more separate policies to determine what is proper.

The use of examples or case studies that deal with difficult rather than obvious issues is a valuable method of interpreting and explaining policy statements about normative or ideal conduct. Few doubt that manufacturing data or forging experimental results is wrong. It may be less clear, however, how preliminary results should be presented in grant applications, when “enough data” are needed to give confidence that a project will succeed but “enough work” remains to be done to justify the grant award.

Most normative rules provide important general principles but leave significant questions unanswered. This void has prompted some universities to take additional steps to foster responsible conduct in research, such as developing guidelines for the conduct of research.

Scope and Purpose of Institutional Guidelines for the Conduct of Research

By “guidelines for the conduct of research,” the panel means institutional policies that address practices such as those related to data management (including data collection, storage, retention, and accessibility), publication (including authorship policies), peer review and refereeing, and training and mentorship. Some institutions have guidelines that focus on a single topic, such as authorship, whereas others adopt a more comprehensive approach. The guidelines may be voluntary or compulsory, and they are administered through a variety of organizational units.

Several major research institutions, such as the National Institutes of Health (for its intramural research program), Harvard Medical School, Johns Hopkins University Medical School, and the University of Michigan Medical School, have formulated comprehensive guidelines for the conduct of research.10 Nevertheless, comprehensive guidelines for research conduct are not common. One study of 133 medical institutions indicated that 17 (13 percent) had such guidelines and that 25 (19 percent) were considering developing guidelines, while 91 (68 percent) were not (Nobel, 1990).

Guidelines for the conduct of research differ from institutional policies that are designed to address misconduct in science or conflict of interest or that, in response to regulatory requirements, govern research involving human subjects, hazardous materials, or recombinant DNA.11 Research conduct guidelines are intended to promote responsible conduct of research and, to the extent that questionable practices and misconduct in science are linked, to reduce the amount



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement