Statistical and Systematical Issues for the Neutrino Mass Ordering
Luca Stanco (right) gave a critical look at statistical analyses currently made to distinguish the normal and inverted mass hierarchy of neutrinos. He in particular criticized the use of the delta chisquared test statistic in the comparison of the two hypotheses, claiming that Wilks theorem does not apply, the two hypotheses are not nested, and the observable does not distribute as a chisquare.
He also discussed the material contained in the poster by Matteo Tenti, discussed in another article in this blog. This is summarized in arXiv:1606.09454v3.
A final point was made on the issue of whether the MH determination is a discovery or an exclusion business. For a discovery, sensitivity is the estimate to quote; one would then need a 5-sigma result. Jowever tat process for MH can also be seen as an exclusion one, and then a 95% CL quotation could be sufficient.The estimator should however be properly computed and many toy MCs, to properly evaluate its probability density function.
In conclusion, Stanco gave an optimistic view: with the next generation of experiments the mass hierarchy will soon be known.