LikeLike

]]>I am 73 y.o. and exercise vigorously. N. N. Taleb is excercise for the intellect.

LikeLike

]]>LikeLike

]]>LikeLike

]]>LikeLike

]]>LikeLike

]]>This is an error that should have been noticed. This is a first year stats student type of mistake, how did it go past all those eyes to get published?

LikeLike

]]>Thanks, will update with a comment on bounds.

LikeLike

]]>It’s not clear you can come up with much of a “best estimate” from data with y=0. Maybe all that can be given in general (without using some prior information on p) are reasonable bounds. I like the Agresti-Coull procedure which gives a reasonable 95% interval. I’ve used this in a consulting project (in that case, the data were y=75 out of n=75) and put it in Regression and Other Stories. ]]>

LikeLike

]]>