Global Fit Kills (or at least Wounds) the Standard Model

Abstract

A few days ago I showed how to combine the Higgs confidence level plots by adding in inverse square. At the time I did not understand why this worked (I am a bit slow at statistics.) Since then I have looked again at the work on electroweak precision tests and the global fit where you can find the same calculation being done. The inverse square of the 95% confidence level limits is just one-quarter of the Δχ² estimator. For independent variables these can be directly added to give an overall Δχ² which can then be mapped back to an overall confidence level limit. This is exactly what I was doing in my combinations. So now I know that these combo plots are essentially correct, neglecting any correlations which should be zero.