00:01
We want to test a claim made by a magazine that 14 % of men use exercise to reduce stress.
00:07
So let's start with our null and alternative hypotheses.
00:14
The claim that has been made is that p, population proportion, is 14%.
00:19
The null hypothesis always has an equal sign.
00:23
So we'll put that here.
00:25
The alternative hypothesis will be that it's something different.
00:29
So we'll put p is not 14%.
00:31
We have no indication of directionality here.
00:35
We don't have anything saying it will be less than that or it will be more than that.
00:39
So we'll just do not equal to.
00:43
For our sample, we have a sample size of 100.
00:47
And out of that, 10 meet the criteria, which is a sample proportion, 10 out of 100, of 0 .1.
00:54
What is the test statistic? so this is based on assuming that the null hypothesis is true.
01:01
The central limit theorem covers normal approximations for sampling distributions.
01:07
If i took every sample of this size, took all of the sample proportions and plotted them out, i would have something approximately normal.
01:16
It follows a normal distribution.
01:17
Its mean is p.
01:19
Its standard deviation, root p, 1 minus p over n.
01:23
If you're wondering where those come from, look at the mean and standard deviation of the binomial, np and root np or minus p, respectively.
01:31
Divide those by n, you get these.
01:33
The same way we divided number of success states by n to get proportion.
01:40
So we're going to have two rejection regions here.
01:43
This is a two -tailed test.
01:45
We care if our statistic is significantly low or significantly high.
01:52
The test statistic is the standard score for our particular sample proportion.
01:58
It's z score.
01:59
Z tells you how many standard deviations away from a mean a value is in its distribution.
02:06
So here, the raw value is p hat.
02:08
The mean of this distribution is p, which we're pretending is 0 .14.
02:13
Its standard deviation, root p, 1 minus p over n...