00:01
A city is contemplating a ban on smoking.
00:03
They'll initiate it if more than 70 % of adults support it.
00:06
We want to see if that's the case based on our sample data.
00:09
So let's perform a hypothesis test.
00:12
We need the null and alternative hypotheses.
00:15
Now the null always gets some kind of equal sign.
00:17
The alternative doesn't.
00:19
So the check of more than 70 % is going to be the alternative.
00:23
So i'm going to put p, population proportion, more than 70%.
00:27
Null will be the opposite of that.
00:29
It's not more than 70%.
00:31
Our sample size is 150, looking at the proportion who support the ban, which is 108 people.
00:40
Giving me a sample proportion of 108 over 150, 0 .72.
00:46
To perform a hypothesis test, we start by assuming the null hypothesis is true.
00:51
So, we'll pretend p is 0 .7.
00:55
If i write every sample of size 150, take the sample of size 150, take the sample of a hypothesis, we start by assuming the null hypothesis is true.
00:58
And plot them out, i'd get something approximately normal.
01:03
P hat follows normal distribution, its mean is p, its deviation, root p, 1 minus p over n.
01:11
So this is the central limit theorem, and now i want to know where my sample would fall on this curve.
01:17
Maybe it falls here.
01:19
It's above 0 .7, but that's not unreasonable here.
01:22
It's very plausible how a sample came from a population with p equal to 0 .7.
01:28
But if p hat is way up here, i would say if the null hypothesis is true, this is really unlikely.
01:35
Therefore, i don't think it's true.
01:38
The cutoff point where we say, okay, that's too unlikely, is called alpha, the level of significance.
01:44
We've not been given one here, so we'll use the kind of default of 5%.
01:49
I'm going to use the critical value method.
01:51
So this is a right -tailed test.
01:55
I want the cut -off point for the top 5 % of this curve.
02:00
If it's just one tail, it gets all of alpha to itself.
02:04
It's called z sub 0 .05, and i know this is 1 .6 .5.
02:10
You might have a table of these, if not use the inverse normal function to get this from software or your calculator...