Transcript ppt
Section 10.4.1
Inference as Decision
AP Statistics
March 4, 2010
CASA
1
When inference is used to make a
decision…
Either you reject H0 or you fail to reject H0.
You can reject H0 correctly
You can fail to reject H0 correctly
You reject H0 incorrectly
(Type
I error)
You can fail to reject H0 incorrectly
(Type
II error)
2
Fail to Reject
H0
Reject H0
H0 is true
Correct
Type I error
H0 is false
Type II error
Correct
3
Potato Chips Example
The salt content of the chips should have a mean of
2 mg with a standard deviation of .1 mg.
When deciding whether to accept or reject a batch
of potato chips, a company looks at the salt content
of 50 chips.
If the salt content is too far away from the mean, it
will reject the batch.
4
What range values are acceptable?
The company will check a 50 chip sample.
If our alpha is .05, the acceptable range is the same
as the 95% confidence interval:
z
*
n
*
z 1.96
0.1
2 1.96
50
2 .0277, 2 .0277
1.9723, 2.0277
5
Accept or reject?
We understand with
normal variation and
everything working
normally, we will get a
sodium value between
1.9723 mg and 2.0277 mg
95% of the time.
0.1
2 1.96
50
2 .0277, 2 .0277
1.9723, 2.0277
6
Accept or reject?
We understand with
normal variation and
everything working
normally, we will get a
sodium value between
1.9723 mg and 2.0277 mg
95% of the time.
This means the 5% of the
time you will reject a batch
of chips that are fine.
When we reject the batch
(and H0) incorrectly we
have committed a Type I
error.
7
Accept H0
Reject H0
Reject H0
95% Confidence Interval
1.9723
2.0277
8
9
Significance and Type I Error
The significance level α of any fixed level test is the
probability of a Type I error. That is, α is the
probability that the test will reject the null hypothesis
H0 when H0 is in fact true.
10
Probability of reject H0 correctly
The probability that we correctly reject H0 (that is, we say
there is a difference when the difference really exists) is
called the “Power”.
The probability of the Type II error is 1- “Power”
We increase the “Power” by either increasing
the
sample size
Alpha
Remember, when we increase alpha, we increase the
probability of the Type I error
11
12
Assignment
10.66-10.69 all
13