They decided against detailed analysis of all the emails in the public domain. They examined just three instances of possible abuse of peer review, and just two cases when CRU researchers may have abused their roles as authors of IPCC reports. There were others. They have not studied hundreds of thousands more unpublished emails from the CRU. Surely openness would require their release.
All this makes it harder to accept Russell’s conclusion that the “rigour and
honesty” of the scientists concerned “are not in doubt”.
New Scientist magazine criticizes their
“failure to investigate whether emails were deleted to prevent their release
under freedom of information laws.”
What does it all mean? When engineers write a reliability report, they'll test some number such as 500 or 1000 parts at high temperature, high humidity, high operating voltage and other accelerated factors to see how quickly they fail. But often the test ends with no failures, which looks pretty good on paper. Nonetheless, a failure rate is calculated as if a failure was just about to occur if the test had gone a little longer. Therefore a short test with no failures is not as good as a long test with no failures.
The trick of the Muir Russell report is to only go so far as to find no wrongdoing. If the scientists' behaviour had problems, you need to investigate less. If they misbehaved in many areas simultaneously, you have to prune several more areas of investigation. In that way, the investigation was severly hampered, but the end result was achieved, it looked pretty good on paper, as far as it went. The only problem is nobody's buying it. Including Muir Russell, who felt it was necessary to stay as far away as possible.
A cynic could read, "the rigour and honesty [of the scientists caught up in Climategate] are not in doubt" in a different way than intended. Whether they are guilty or innocent is not in doubt, the investigation just failed to say which it is.
No comments:
Post a Comment