You are on page 1of 58

Chapter 9

Chapter 9: Tests of Hypotheses


Based on a Single Sample
Chapter Contents

9.1 Logic of Hypothesis Testing


9.2 Statistical Hypothesis Testing
9.3 Testing a Mean: Known Population Variance
9.4 Testing a Mean: Unknown Population Variance
9.5 Testing a Proportion
9.6 Power Curves and OC Curves (Optional)
9.7 Tests for One Variance (Optional)

9-1
Chapter 9
LO2 9.1 The Logic of Hypothesis Testing

A Motivating Example: Paper Manufacturing


• The Hammermill Company produces paper for laser
printers. Standard paper width is supposed to be 216mm,
or 8.5 inches. Suppose that the actual width is a random
variable that is normally distributed with a known standard
deviation of 0.023mm. This standard deviation reflects the
manufacturing technology currently in use and is known
from long experience with this type of equipment.
• A quality control inspector chooses 50 sheets at random
and measures them with a precise instrument, showing a
sample mean width of 216.007 mm. Does this sample show
that the quality of the products conforms to the industry
standard?
9-2
Chapter 9
LO2 The Paper Manufacturing Case

μ = 216mm
VS μ≠ 216mm
Population

sample size
Which hypotheses n=50
should be chosen? sample mean
X = 216.007mm Sample

9-3
Chapter 9
LO2 The Paper Manufacturing Case

How to make a decision?


• Proof by contradiction (反证法) in mathematics.

Assume μ=216, doubt if X is far away from


216.
• “Far away” needs to be interpreted with care.

9-4
Chapter 9
LO2 The Paper Manufacturing Case
Assumptions:

• Assume the thickness of the paper has a normal distribution.


• Moreover, we know the standard deviation is 0.023mm from
past experience.

9-5
Chapter 9
LO2 The Paper Manufacturing Case

A Sampling distributions of X
• The sample mean

E(X)  μ(  216)
• The sample standard deviation
SD(X)  σ n  0.023  0.00325
50
• The sample distribution
2
X~N(216,0.00325 )

9-6
Chapter 9
LO2 The Paper Manufacturing Case

Z-statistic
• Z-statistic
Xμ X  216 X  216
Z  
σ n 0.023 50 0.00325
• has a standard normal distribution.

0
• Checking whether X is far away from is equivalent to checking
whether Z is far away from 0.
9-7
Chapter 9
LO2 The Paper Manufacturing Case

Standard Normal Distribution

-2 -1 0 1 2

This is likely.
Z
9-8
Chapter 9
LO2 The Paper Manufacturing Case

Standard Normal Distribution

-2 -1 0 1 2

This is likely as
Z well.
9-9
9-9
Chapter 9
LO2 The Paper Manufacturing Case

Standard Normal Distribution

5%

Reject

-2 -1 0 1 2

This is unlikely.
Z
Xμ 216.007  216
Z|data  |data   2.154
σ n 0.023 50 9-10
Chapter 9
LO2 The Paper Manufacturing Case
A summary:

• Two competing hypotheses:


μ = 216 v.s. μ≠ 216
• Assume μ = 216 is true.
• From the information contained in X (or Z), we observe an
unlikely event occurs (with probability less than 5%).
• We reject the initial hypothetical claim: μ = 216

9-11
9.1 Hypotheses and Test Procedures
A Motivating Example

9-12
9.1 Hypotheses and Test Procedures
Example 9.5

9-13
9.1 Hypotheses and Test Procedures
Example 9.5

9-14
9.1 Hypotheses and Test Procedures
Example 9.5

9-15
Chapter 9
LO9-1 9.1 The Logic of Hypothesis Testing

LO9-1: List the steps in testing hypotheses.

9-16
Chapter 9
LO9-2 9.1 The Logic of Hypothesis Testing

LO9-2: Explain the difference between H0 and H1.

State the Hypothesis

• Hypotheses are a pair of mutually exclusive, collectively


exhaustive statements about some fact about a population.
• One statement or the other must be true, but they cannot both
be true.
• H0: Null Hypothesis 零假设
H1: Alternative Hypothesis 备择假设
• These two statements are hypotheses because the truth is
unknown.

9-17
Chapter 9
LO9-2 9.1 The Logic of Hypothesis Testing

State the Hypothesis


• Efforts will be made to reject the null hypothesis.
• If H0 is rejected, we tentatively conclude H1 to be the case.
• H0 is sometimes called the maintained hypothesis.
• H1 is called the action alternative because action may be required
if we reject H0 in favor of H1.

Can Hypotheses be Proved?


• We cannot accept a null hypothesis, we can only fail to reject it.

Role of Evidence
• The null hypothesis is assumed true and a contradiction is sought.

9-18
Chapter 9
LO9-3 9.1 The Logic of Hypothesis Testing

LO9-3: Define Type I error, Type II error, and power.


Types of Error
• Type I error 一类错误: Rejecting the null hypothesis when it is
true. This occurs with probability a (level of significance).
• Type II error二类错误 : Failure to reject the null hypothesis when it
is false. This occurs with probability b.

9-19
Chapter 9
LO9-3 9.1 The Logic of Hypothesis Testing

Probability of Type I and Type II Errors

• If we choose a = .05, we expect to commit a Type I error about 5


times in 100.
• b cannot be chosen in advance because it depends on a and the
sample size.
• A small b is desirable, other things being equal.

9-20
Chapter 9
LO9-3 9.1 The Logic of Hypothesis Testing

Power 功效 of a Test

• A low b risk means high power.


• Larger samples lead to increased power.
9-21
Chapter 9
LO3 The Power: Paper Manufacturing Case

 Examine Type II error


 The target alternative
hypothesis
H1: μ = 216.5
 Power=0.24

a/2 = 0.025 a/2 = 0.025

215.9936 216 216.0064


216.005

9-22
9-22
Chapter 9
LO9-3 9.1 The Logic of Hypothesis Testing

Relationship Between a and b

• Both a small a and a small b are desirable.


• For a given type of test and fixed sample size, there is a trade-off
between a and b.
• The larger critical value needed to reduce a risk makes it harder
to reject H0, thereby increasing b risk.
• Both a and b can be reduced simultaneously only by increasing
the sample size.

9-23
Chapter 9
LO9-3 9.1 The Logic of Hypothesis Testing
Consequences of Type I and Type II Errors

• The consequences of these two errors are quite different, and the
costs are borne by different parties.
• Example: Type I error is convicting an innocent defendant, so the
costs are borne by the defendant. Type II error is failing to convict a
guilty defendant, so the costs are borne by society if the guilty
person returns to the streets.
• Firms are increasingly wary of Type II error (failing to recall a product
as soon as sample evidence begins to indicate potential problems.)

9-24
9.2 Tests about a Population Mean
Example 9.5

9-25
9.2 Tests about a Population Mean
Example 9.5

9-26
9.2 Tests about a Population Mean
Example 9.5

9-27
Chapter 9
The Duality of Confidence Interval and
LO9
Hypothesis Testing

9-28
9-28
9.3 Tests Concerning a Population Proportion
Example 9.5

9-29
9.3 Tests Concerning a Population Proportion
Example 9.5

9-30
9.3 Tests Concerning a Population Proportion
Small Sample Case

9-31
Chapter 9
LO9-9
9.3 Testing a Proportion
Example: Return Policy
• Retailers such as GUESS, Staples are employing new technology
to crack down on “serial exchangers”—customers who abuse
their return and exchange policies.
• The historical return rate for merchandise at department stores is
13%. After implementing a new technology, there were 22 returns
in a sample of 250 purchases. Does this indicate that the true
return rate has fallen?

Steps in Testing a Proportion


• Step 1: State the hypotheses
For example, H0: p  .13
H1: p < .13

9-32
Chapter 9
LO9-9 9.3 Testing a Proportion

Steps in Testing a Proportion


• Step 2: Specify the decision rule
For a = .05
for a left-tail area,
reject H0 if z < -1.645,
otherwise do not
reject H0.

Figure 9.12 9-33


Chapter 9
LO9-9 9.3 Testing a Proportion
Steps in Testing a Proportion

• Step 3: Collect Sample Data and Calculate the Test Statistic


If H0 is true, then the test statistic should be near 0 because the
sample mean should be near μ0. The value of the test statistic is
given next.

• Step 4: Since the test statistic lies in the left-tail rejection region,
we reject the null hypothesis H0: p  .13.

Figure 9.12 9-34


Chapter 9
LO9-9 9.3 Testing a Proportion
Calculating the p-Value

9-35
An Alternative Way to Report: P-
value (P-值)
a‘/2 = 0.005 a‘/2 = 0.005
a/2 = 0.025 a/2 = 0.025

0.016 0.016

-2.57 -1.96 0 1.96 2.57


-2.154 2.154
P(Z  -2.36)  0.016
P-value
P(Z  2.36)  0.016
= 0.016+0.016 = 0.032
9-36
9-36
The Choice of The Significance
Levels: P-value (P-值)

a/2 = 0.025 a/2 = 0.025

1.96 0 1.96
2.154 P-value 0.032
P-value 0.049 1.97

9-37
9-37
9.4 P-Values
Example 9.5

9-38
9.4 P-Values
Example 9.5

9-39
9.4 P-Values
P-value for z tests

9-40
9.4 P-Values
P-value for z tests

9-41
9.5 Some Comments on Selecting a Testing
Procedure
Questions:

9-42
Chapter 9
LO3 9.5 Some Comments on Selecting a
Testing Procedure
Statistical Significance 统计意义下的显著性 versus
Practical Significance (Importance) 实际意义下的重要性
• The standard error of most sample estimators approaches 0 as
sample size increases.
• In this case, no matter how small, q – q0 will be significant if the
sample size is large enough.
• Therefore, expect significant effects even when an effect is too
slight to have any practical importance.

9-43
9-43
9.5 Some Comments on Selecting a Testing
Procedure
Statistical Significance 统计意义下的显著性 versus Practical Significance
(Importance) 实际意义下的重要性

9-44
9.5 Some Comments on Selecting a Testing Procedure

Likelihood Ratio Tests: Neyman-Pearson Theorem

9-45
9.5 Some Comments on Selecting a Testing Procedure

Example 9.20

9-46
9.5 Some Comments on Selecting a Testing Procedure

Example 9.20

9-47
9.5 Some Comments on Selecting a Testing Procedure
Example 9.21

9-48
9.5 Some Comments on Selecting a Testing Procedure

Power 功效

9-49
9.5 Some Comments on Selecting a Testing Procedure

Power 功效

9-50
9.5 Some Comments on Selecting a Testing Procedure

Power 功效

9-51
9.5 Some Comments on Selecting a Testing Procedure

Likelihood Ratio Tests: General Theory

9-52
9.5 Some Comments on Selecting a Testing Procedure

Likelihood Ratio Tests: General Theory

9-53
9.5 Some Comments on Selecting a Testing Procedure

Likelihood Ratio Tests: General Theory

9-54
9.5 Some Comments on Selecting a Testing Procedure

Example 9.24

9-55
9.5 Some Comments on Selecting a Testing Procedure

Example 9.24

9-56
9.5 Some Comments on Selecting a Testing Procedure

Example 9.25

9-57
9.5 Some Comments on Selecting a Testing Procedure

Example 9.25

9-58

You might also like