Monday, January 16, 2006

Trying to prevent fraud

We're all well aware that the practice of peer review in scientific journals does not appear to be working too well. David Dobbs, in yesterday's NY Times, suggests adopting a Wikipedia approach where papers are posted on the web and reviewed not by anonymous, assigned reviewers but by anyone who can access the web site.

Yes, this approach has also gotten some bad press recently, but, overall, Wikipedia's articles are quite good; Nature magazine has found Wikipedia's science entries almost as accurate as Brittanica's. If access to the 'article review' web site is tightly controlled, it's unlikely that the results can be worse than the peer review process.

No comments: