00:01
Here we have some sample data.
00:02
We have a sample size n of a thousand people.
00:06
Out of these, 558 responded no to the question of if us nuclear weapons made them safer.
00:15
We want to use this to make a 90 % confidence interval for the population proportion.
00:21
Formula we need is the point estimate, p -hat, the sample proportion, plus and minus the margin of error, z -root, p -hat minus p -hat over n.
00:34
So we have n, what's p -hat, what's z? p -hat is the proportion of our sample who felt that way.
00:44
So just pass this.
00:47
They were asked if they felt it made them feel safer.
00:49
They said no.
00:50
We want the interval for those who felt it did not make them safer.
00:53
So we do have the information we need, 558 out of a thousand, 0 .58.
01:04
Z we get from a level of confidence.
01:08
Right now this is a binomial experiment.
01:11
A thousand independent trials, two outcomes, they don't feel safer or presumably they do, same probability p, for each person not feeling safer.
01:20
We don't know what p is yet, but we'll estimate it.
01:24
I take a normal approximation to this binomial and divide it all by n...