A survey of more than 3,000 scientists has revealed that a large fraction are acting in ways that could compromise the integrity of research, according to an article published today in Nature.
A third of participants in the survey acknowledged that they had engaged in actions such as overlooking others’ use of flawed data, failing to present data contradicting one’s own work, and circumventing minor requirements of human-subject research. While those actions do not rise to the level of fraud, fabrication, and plagiarism — the three cardinal sins of research — they nonetheless signal problems in the world of science, said Brian C. Martinson, a research investigator at the HealthPartners Research Foundation, a nonprofit center in Minneapolis.