@IanSudbery @NSSDeviations Yeah, we adjust the local p-value to a global p-value with the "LEE" and it is the _global_ p-value that must be greater than 5-sigma to call discovery. The literature isn't great here, but there is "Trial factors for the look el
RT @HEPfeickert: Wrapping up Friday with (statistical) #physics and #coffee. https://t.co/bH7VoE2uJx https://t.co/Jrjcsshq4b
RT @HEPfeickert: Wrapping up Friday with (statistical) #physics and #coffee. https://t.co/bH7VoE2uJx https://t.co/Jrjcsshq4b
Wrapping up Friday with (statistical) #physics and #coffee. https://t.co/bH7VoE2uJx https://t.co/Jrjcsshq4b
@DrAndreDavid @CMSvoices @KyleCranmer @aartheijboer much needed de-mystification & mythbusting. From dice-rolling to https://t.co/I14Xo3zn8r
@RobertGaristo Not quite, because the trials factor is proportional to z₀. See Eq 11 in https://t.co/ND3pbrRxdn Pinging @KyleCranmer