More than half of scientists in the Netherlands regularly violate scientific norms by omitting unwanted research results, obfuscating methodological problems with research, or selectively citing available literature. In fact, more than 8 percent have fabricated or falsified research findings in the past three years. This is clear from National Scientific Integrity Survey, a project led by Lex Bouter, Professor of Methodology and Integrity at Vrije Universiteit and Amsterdam UMC. The results were published on Tuesday in the man Article – commodity. This is on prepress server. This means that the publication has not yet undergone peer review.
The project aims to improve the quality and reliability of scientific research in the Netherlands, says Bauter. “Next year we will be talking to universities about how to use the search results.”
The study found links between breaking scientific rules and a number of external factors. If the researchers think they have a strict Peer review [beoordeling door vakgenoten] They committed fewer frauds. Scientists who felt compelled to perform due to publication pressures or the need to secure new funding engaged in questionable research practices. Men and people early in their careers also admitted to making mistakes relatively often.
Read also: How do you measure the quality of scientists?
“It probably won’t be great, but it’s the first time this project has been built this way,” Potter says. “Note: These are associations. Our investigation did not prove cause and effect. So it is not said that if you turn on the confirmation button as people experience it, this behavior decreases immediately. The effect of these phenomena is complex.”
The study of inappropriate behavior
Potter and his team sent out surveys among employees of 22 Dutch universities and university medical centers. In it, questions were raised about scientifically incorrect behavior and the factors that can influence it. Only eight institutions actively cooperated by providing email addresses and promoting participation, while the researchers collected the addresses themselves for 14. In total, more than 63,778 questionnaires were sent. Potter: “The response of the eight participating organizations actively was more than 21 percent. We cannot comment on the other fourteen because we do not know how accurate our email address pool is.”
Strictly speaking, Bauter asserts that the research findings say nothing about the participants and not about Dutch science as a whole, but he is confident in the quality and representation of his data. “The answer is comparable to similar studies abroad, and in absolute numbers I’m not aware of any greater international research on the subject.”
Participants were allowed to answer the most sensitive questions using . method random response Potter says it’s called, which is a solid guarantee of anonymity. This technology is also used in investigations of doping in sports and social security violations. This results in a two to three times higher percentage of people admitting to breaking the rules. So our 8 percent is much closer to reality than when we look at this in a conventional way.”
But isn’t it conceivable that many scientists who didn’t complete the survey didn’t because they had to admit something – and that the numbers are higher in real life? Potter: “It could be, but we don’t know. Either way, it wouldn’t be much less, because I can’t imagine respondents admitting mistakes they weren’t guilty of.”
On the morning of July 9, 2021, a version of this article also appeared on NRC