Probably scientists manipulate the data in their publications, when they were meaning, such a new study
Spectacular trap of the fraud or federation in the sciences as that of the sudkoreaner hwang woo-suk, the preliminary, human embryonic stem cells (korean cloning hero deconstructed online), and celebrated by the magazine science, or the physicist jan hendrik already come again and again (sluts, manipulate, false). Unknown and therefore very controversial, however, is the dark figure, how often frauds are actually in the scientists.
Daniele fanelli from the university of edinburgh has performed a first meta analysis of surveys among scientists in his study, which appeared in the open access magazine PLL ONE. Beauties or falsifications are therefore not so rare, particularly consistent is, little surprising, medical and pharmaceutical research. The known trap is probably only the tip of the iceberg.
In order to make the surveys made comparable in different countries among scientists of different specialists with different methods and questions, fanelli focused on the information of the scientists who encountered a certain type of fraud at least once or to know a colleague who has done this. There were not plagiarity trap, which probably also often in the digital age by light cut paste occur (but also easier to be revealed) or unethical authorships in the sciences that are not less frequently. Plagiates were therefore not included because they were not so damaging the scientific knowledge itself as the false or fabrication of data. However, fanelli makes it clear that scientific misconduct is difficult to define. For his meta analysis, he evaluated 18 surveys he considered sufficiently compatible.
On average, 1.97 percent of scientists admitted that they had changed or fabricated data at least once in order to improve the outcome of their study or to adapt the results to the requirements of the funder. One should ame that it will probably be more because of scientists objectivity and honesty as values are highly high – and this is required by them as well. 33.7 percent domed that they have used other fragged tricks, for example, to conceal information that contradicts their own results, or not to involve observations or data because the scientists had the feelings that can not be right. Fanelli points to surveys in which only a few percent of respondents said that they had changed data, but much more allowed them to improve the results have already selected or not use data.
When it comes to colleagues, the scientists are volatile (or suspicious?To). 14 percent say they got to know a colleague who has fabricated, changed or false data, with other fraghous cheating, do this even say 72 percent. What they are more likely to conceal themselves with themselves, they could rather overwhelm, especially as it is not clearly clear what a scientific misconduct is. So many scientists, says fanelli, does not matter if it is their own research that one "improvement" the data is a false or fraud. That could also be the reason why scientists look at themselves as honest as their colleagues (but that’s not only the case with scientists).
Striking is that obviously under fanelli’s surveys in the medical research areas, which is important to which clinical and pharmacological studies, is most manipulated. The konne is because here the coarse financial interests act, but it could, at least very theoretically, also because the physicians of the problem are rather aware and therefore rather admit it.