Monday, December 07, 2009

Painting The Mice
I had a chance to discuss "climategate" or "stolen-email-gate" as Barbara Boxer prefers to call it with a colleague today at lunch. He didn't think it was too big a deal, and expressed a belief that the behaviors described in the emails were not exceptional, nor evidence of any misconduct. I challenged him on that point, and reminded him of the case of William Summerlin, in the 1970's. Summerlin was a researcher at the Sloan-Kettering Institute who claimed to have solved the allograft rejection problem, which, if true, would have likely won him a Nobel Prize in Medicine. Unfortunately, others were not able to replicate his results. When his boss asked him to show his data, Summerlin apparently painted black patches on white mice. He was caught by a whistleblower. Since that time, "painting the mice" has been a term to describe deliberately fudging scientific data.
Scientific misconduct is not common, and generally takes the form of plagiarism. Actually manipulating data to produce a fraudulent result is probably much less common. Because even the suspicion that something like that has occurred would be disastrous to a scientist's career and reputation, no reputable scientist would ever refuse to show his original data and the methods used to refine the data. No real scientist would destroy his original data, especially after questions arose. The destruction of the original data calls the entire output of the CRU into question. There have always been questions about the alarmism over this issue. The discrediting of the "hockey stick", for example. The extreme claims of Al Gore, and some of the proposals of our current government and the extreme environmentalists can now be seen to have no sound scientific basis, and will inevitably end up as the Piltdown man of this century.