Confidence interval question

[senmeis]

Hi,

on the last page of this document https://faculty.uccs.edu/mwickert/wp-content/uploads/sites/58/2026/03/N5630_5.pdf the confidence interval diagram is to read. Somehow I feel difficult to understand it.

1. Why isn’t it related to n/N instead of n?

2. Do y+ and y- have absolute meaning besides scaled factors?

 

[berkeman]

1. Why isn’t it related to n/N instead of n?

But that is what they show, no?

 

[senmeis]

Let’s take a look at the shown example. With the confidence level = 0.95 the scaled error probability is only a function of n. Is it not strange?