Licence to publish will restore trust in science

It may perhaps be possible to design an accreditation system that could be used by all scientists - but by this argument, do we not need one that covers all researchers in all fields, not just the sciences? Experts in general - not just in the sciences - seem to be less respected than they once were. Can we imagine a set of key research skills that ALL researchers need to show they have, that would work for economists, sociologists and historians, as well as scientists?

As a PhD is supposed to be proof of the ability to carry out and publish research, perhaps what's needed is to make PhD programmes more rigours, and PhD examinations more demanding.

Having trained in the UK and worked in the US as a scientist and medical writer, I think there is something to be said for the American system. I was lucky, in that my primary supervisor for my D.Phil (Oxford) is American, and he made sure I received what he thought was complete training in all aspects of research conduct. Others in the same Department got all the way to their Viva (final doctoral "interrogation") to find that they could only graduate with an MSc. Now that I teach in the US system, I see that their longer training, with 2 years of classes before they start full-time intensive laboratory research produces more consistently educated and rounded scientists. Perhaps Ph.D programs have changed in the UK since I graduated (1999), but I was certainly expected to have all the skills listed on this article when I began my first post-doc in the US.