Silflay Hraka

So it must be time for another heaping helping of heresy. Those who get emotional when their religion is questioned might want to skip to the next post.

Ans: It is precisely in a small specialized discipline that the likelihood of turning up sympathetic referees is highest. Within a small, focused discipline, there simply are fewer referees available. Also, there is always the possibility of the discipline becoming extinct or irrelevant. The referees have a vested interest in seeing that research is published, especially if there is a strong consensus. It has been my experience both in journals as well as with the awarding of grants that staying close to the consensus opinion is most likely to result in funding or publications because the reviewers like to see work that is similar to their own and work that reinforces their position.
…
We examined the list of references in the paleoclimate papers that we considered to find any evidence that these papers were using contemporary statistical tools, that they were citing the current statistics literature, and that they had basic knowledge of the statistics literature. We examined resumes of the most frequently published authors to understand where and with whom they obtained their statistical training. We examined the composition of the Probability and Statistics Committee2 of the American Meteorological Society searching for “mainstream statisticians.” We examined the scientific programs for the AMS’s Conferences on Probability and Statistics in the Atmospheric Sciences3 for “mainstream statisticians.” In every case, while there are a few examples of cooperation, they are the exception. The atmospheric science community, while heavily using statistical methods, is remarkably disconnected from the mainstream community of statisticians in a way, for example, that is not true of the medical and pharmaceutical communities.
…
Our work demonstrates that the methodology is incorrect. Because of the lack of proper statistical sampling and correct inferential methodology, we concluded that the statements regarding the decade of the 1990s probably being the hottest in a millennium and 1998 probably being the hottest year in a millennium are unwarranted. Indeed, I repeatedly testified that the instrumented temperature record from 1850 onwards indicated that there is a pattern of global warming. We have never disputed this.

via Climate Audit