Also in this month’s issue of Science & Engineering Ethics is a report entitled Mentoring and Research Misconduct: An Analysis of Research Mentoring in Closed ORI Cases by David Wright, Sandra Titus (who in June published data on the underreporting of misconduct), and Jered Cornelison.
In this study, they focus on trainee misconduct investigated by ORI from 1990 to 2004 (45 of the 158 closed cases). All but 3 of the cases involved fabrication and/or falsification. In reviewing the records, the authors discovered that 73% of the mentors had not looked at the raw data generated by their trainees, and 62% did not set standards for collecting and maintaining data; of greatest concern is the finding of
a troublingly high incidence of missing data or of no lab books at all (even in the laboratories of renowned scientists).
Further, 24 of the trainees reported stress as a contributing factor to their misconduct, with 15 of these indicating that “they fabricated or falsified data because they felt internal pressure to perform well.”
The authors conclude with a thoughtful observation and suggestion:
Surprisingly, in doing our literature review we found that there appear to be no agreed upon standards or best practices in the research community recommending that mentors or lab directors review trainee raw data at regular intervals, whereas there should be.
Institutions, perhaps through peer review of mentoring practices, should assure that research standards are implemented and enforced.
Indeed, with the Hellinga case again serving as a poster child for this and other reform to provide better institutional oversight of mentor practices. The Chronicle offers some additional food for thought as well.
Medical Writing, Editing & Grantsmanship 
drugmonkey said
Further, 24 of the trainees reported stress as a contributing factor to their misconduct, with 15 of these indicating that “they fabricated or falsified data because they felt internal pressure to perform well.”
This is not “inadequate” mentoring. This is mentors setting an environment in which falsification is encouraged. Ever seen labs like this? I have. Guess what? The PI is more culpable than is the trainee. Yet the trainee is always left holding the bag…
In defense of reviewers said
“Internal pressure to perform well” is a bad thing?
David said
“Internal pressure to perform well” should be fully replaced with “Internal pressure to do good science.” If you are doing good science, you might start off getting a lot of negative results, but they will be meaningful and should suggest other experiments or switching projects. Good science means good controls (negative and positive). Evidently the Hellinga case had problems with controls.