Friday, May 25, 2012

Washington Post warmist Juliet Eilperin: "Mann and his team used a statistical tool called “principal component analysis” to tease out which factors are most responsible for a given change over time..."

Book review: Michael Mann’s ‘The Hockey Stick and the Climate Wars’ - The Washington Post

Mann and his team used a statistical tool called “principal component analysis” to tease out which factors are most responsible for a given change over time and to compare data from all over the globe.

Two Canadians, mining consultant Stephen McIntyre and economist Ross McKitrick, challenged the hockey stick on a variety of fronts. First they went after the proxy data; later they employed the same statistical tool that Mann and his colleagues did, but in what Mann considers an inconsistent way. The result was a smoothing out of the pattern embodied in the hockey stick. “If there is a lesson” in these sorts of fights, Mann writes, “it is that scientific findings that rest on such technical complexities are prone to abuse by those with a potential ax to grind.”

1 comment:

Anonymous said...

Principal component analysis is only valid if the time series exhibits stationarity. Climate does not exhibit stationarity either as observed or in the output of climate models. Therefore, Mann's work is total bunk. Ditto, for all other climate reconstructions using PCA, which is nearly all of them.