unusual connection between Blasey and @MichaelEMann.
Kraemer &Blasey 2006, "Centring in regression analyses: a strategy to prevent errors in statistical inference" observed that "Using non‐centred data in regression analysis..often leads to inconsistent and misleading results".
Kraemer &Blasey 2006, "Centring in regression analyses: a strategy to prevent errors in statistical inference" observed that "Using non‐centred data in regression analysis..often leads to inconsistent and misleading results".
2/ use of non-centered data in principal components is much worse, as observed in McIntyre and McKitrick 2005 climateaudit.info/pdf/mcintyre.m… where we observed that Mann's centering on a subsegment profoundly biased results towards a hockey stick. In response, Mann argued that centering.
3/ was merely an arbitrary convention and advocated alt-statistics - an ideology which satisfied the highly ideological climate community. While Blasey did not expressly turn her attention to the perils of non-centering in principal components, it is a logical consequence of her
4/ concern over the perils of non-centering in the more forgiving operation of linear regression. Hopefully, Blasey will have an opportunity to address Mannian principal components when her present tribulations are over.
