Every so often, the fallibility of scientific institutions is pushed into the spotlight. Recently, that spotlight has been focused on the retraction of a Science paper because one of its authors allegedly faked groundbreaking evidence that someone's political opinions could be changed by a short conversation with an affected person—specifically, in this case, that a conversation with a gay canvasser could persuade people to be more in favour of gay marriage.
This case has brought public scrutiny to issues of research ethics and transparency. In scientific communities, potential solutions to these problems have been debated and trialled for a considerable amount of time. The latest issue of Science includes three articles that lay out options for journals, universities, and individual researchers who hope to improve transparency and accountability in research.
You can't repeat what you can't see
Reproducibility is one of the cornerstones of science. If one lab tries something and finds that it works, but five other labs try it and come up with nothing, that’s an indication that something was strange about the first group’s results—perhaps they faked the data, or perhaps, more innocuously, they just did something differently.
Read 12 remaining paragraphs | Comments
No comments:
Post a Comment