Chapter 9 9-1

Learning Objectives

Basic Business Statistics In this chapter, you learn:

12th Edition n The basic principles of hypothesis testing
n How to use hypothesis testing to test a mean or
proportion
Chapter 9
n The assumptions of each hypothesis-testing
procedure, how to evaluate them, and the
consequences if they are seriously violated
Fundamentals of Hypothesis
How to avoid the pitfalls involved in hypothesis testing
Testing: One-Sample Tests n

n The ethical issues involved in hypothesis testing

Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-1 Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-2

What is a Hypothesis? The Null Hypothesis, H0

n A hypothesis is a claim n States the claim or assertion to be tested
(assertion) about a
population parameter: Example: The average diameter of a
manufactured bolt is 30mm H( 0 : µ = 30 )
n population mean
n Is always about a population parameter,
Example: The mean monthly cell phone bill not about a sample statistic
in this city is µ = $42
n population proportion
Example: The proportion of adults in this H0 : µ = 30 H0 : X = 30
city with cell phones is π = 0.68
Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-3 Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-4

The Null Hypothesis, H0 The Alternative Hypothesis, H1

(continued)

n Begin with the assumption that the null n Is the opposite of the null hypothesis
hypothesis is true n e.g., The average diameter of a manufactured
bolt is not equal to 30mm ( H1: µ ≠ 30 )
n Similar to the notion of innocent until
proven guilty n Challenges the status quo
n Alternative never contains the “=”sign
n Refers to the status quo or historical value n May or may not be proven
n Null always contains “=“sign n Is generally the hypothesis that the
researcher is trying to prove
n May or may not be rejected

Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-5 Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-6
Chapter 9 9-2

The Hypothesis Testing Process The Hypothesis Testing Process

(continued)
n Suppose the sample mean age was X = 20.
n Claim: The population mean age is 50.
n H0: µ = 50, H1: µ ≠ 50 n This is significantly lower than the claimed mean
n Sample the population and find sample mean. population age of 50.

Population n If the null hypothesis were true, the probability of

getting such a different sample mean would be
very small, so you reject the null hypothesis.

n In other words, getting a sample mean of 20 is so

unlikely if the population mean was 50, you
Sample
conclude that the population mean must not be 50.

Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-7 Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-8

The Test Statistic and

The Hypothesis Testing Process Critical Values
(continued)
n If the sample mean is close to the stated
Sampling population mean, the null hypothesis is not
Distribution of X rejected.
n If the sample mean is far from the stated
population mean, the null hypothesis is rejected.
X
20 µ = 50 n How far is “far enough” to reject H0?
If H0 is true ... then you reject
If it is unlikely that you n The critical value of a test statistic creates a “line in
the null hypothesis
would get a sample
mean of this value ... ... When in fact this were
that µ = 50. the sand” for decision making -- it answers the
the population mean… question of how far is far enough.
Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-9 Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-10

The Test Statistic and Possible Errors in Hypothesis Test

Critical Values Decision Making
Sampling Distribution of the test statistic n Type I Error
n Reject a true null hypothesis

n Considered a serious type of error

Region of Region of
Rejection Rejection n The probability of a Type I Error is α
Region of
Non-Rejection n Called level of significance of the test
n Set by researcher in advance
n Type II Error
Critical Values
n Failure to reject a false null hypothesis

n The probability of a Type II Error is β

“Too Far Away” From Mean of Sampling Distribution

Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-11 Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-12
Chapter 9 9-3

Possible Errors in Hypothesis Test Possible Errors in Hypothesis Test

Decision Making Decision Making
(continued) (continued)

Possible Hypothesis Test Outcomes n The confidence coefficient (1-α) is the

probability of not rejecting H0 when it is true.
Actual Situation
n The confidence level of a hypothesis test is
Decision H0 True H0 False (1-α)*100%.

Do Not No Error Type II Error n The power of a statistical test (1-β) is the
Reject H0 Probability 1 - α Probability β probability of rejecting H0 when it is false.
Reject H0 Type I Error No Error
Probability α Probability 1 - β

Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-13 Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-14

Type I & II Error Relationship Factors Affecting Type II Error

n All else equal,
§ Type I and Type II errors cannot happen at n β when the difference between
the same time hypothesized parameter and its true value
§ A Type I error can only occur if H0 is true
§ A Type II error can only occur if H0 is false n β when α
n β when σ
If Type I error probability (α) , then
n β when n
Type II error probability (β)

Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-15 Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-16

Level of Significance
and the Rejection Region Hypothesis Tests for the Mean
H0: µ = 30 Level of significance = α

H1: µ ≠ 30 Hypothesis
Tests for µ
α
/2 α
/2

σ Known σ Unknown
30 (Z test) (t test)
Critical values

Rejection Region

This is a two-tail test because there is a rejection region in both tails

Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-17 Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-18
Chapter 9 9-4

Z Test of Hypothesis for the Mean Critical Value

(σ Known) Approach to Testing
n Convert sample statistic X
( ) to a ZSTAT test statistic
n For a two-tail test for the mean, σ known:
Hypothesis
Tests for µ n Convert sample statisticX( ) to test statistic
(ZSTAT)
σσKnown
Known σσUnknown
Unknown n Determine the critical Z values for a specified
(Z test) (t test) level of significance α from a table or computer
The test statistic is:
n Decision Rule: If the test statistic falls in the
X −µ
Z STAT = rejection region, reject H0 ; otherwise do not
σ
reject H0
n
Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-19 Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-20

6 Steps in
Two-Tail Tests Hypothesis Testing
H0: µ = 30
n There are two 1. State the null hypothesis, H0 and the
cutoff values H1: µ ≠ 30
alternative hypothesis, H1
(critical values),
defining the 2. Choose the level of significance, α, and the
regions of α/2 α/2 sample size, n
rejection
30 X 3. Determine the appropriate test statistic and
Reject H0 Do not reject H0 Reject H0
sampling distribution
-Zα/2 0 +Zα/2 Z 4. Determine the critical values that divide the
Lower
rejection and nonrejection regions
Upper
critical critical
value value
Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-21 Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-22

6 Steps in
Hypothesis Testing Hypothesis Testing Example
(continued)
Test the claim that the true mean diameter
5. Collect data and compute the value of the test of a manufactured bolt is 30mm.
statistic (Assume σ = 0.8)
6. Make the statistical decision and state the
1. State the appropriate null and alternative
managerial conclusion. If the test statistic falls hypotheses
into the nonrejection region, do not reject the n H0: µ = 30 H1: µ ≠ 30 (This is a two-tail test)
null hypothesis H0. If the test statistic falls into 2. Specify the desired level of significance and the
the rejection region, reject the null hypothesis. sample size
Express the managerial conclusion in the n Suppose that α = 0.05 and n = 100 are chosen
context of the problem for this test

Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-23 Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-24
Chapter 9 9-5

Hypothesis Testing Example Hypothesis Testing Example

(continued) (continued)

3. Determine the appropriate technique n 6. Is the test statistic in the rejection region?
n σ is assumed known so this is a Z test.

4. Determine the critical values α/2 = 0.025 α/2 = 0.025

n For α = 0.05 the critical Z values are ±1.96

5. Collect the data and compute the test statistic

Reject H0 if Reject H0 Do not reject H0 Reject H0
n Suppose the sample results are
ZSTAT < -1.96 or -Zα/2 = -1.96 0 +Zα/2 = +1.96
n = 100, X = 29.84 (σ = 0.8 is assumed known) ZSTAT > 1.96;
So the test statistic is: otherwise do
not reject H0 Here, ZSTAT = -2.0 < -1.96, so the
X − µ 29.84 − 30 − .16
ZSTAT = = = = −2.0 test statistic is in the rejection
σ 0.8 0.08
n 100 region
Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-25 Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-26

Hypothesis Testing Example p-Value Approach to Testing

(continued)
6 (continued). Reach a decision and interpret the result
n p-value: Probability of obtaining a test
statistic equal to or more extreme than the
α = 0.05/2 α = 0.05/2
observed sample value given H0 is true
n The p-value is also called the observed level of
Reject H0 Do not reject H0 Reject H0
significance
-Zα/2 = -1.96 0 +Zα/2= +1.96
-2.0 n H0 can be rejected if the p-value is less than α
Since ZSTAT = -2.0 < -1.96, reject the null hypothesis
and conclude there is sufficient evidence that the mean
diameter of a manufactured bolt is not equal to 30
Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-27 Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-28

p-Value Approach to Testing: The 5 Step p-value approach to

Interpreting the p-value Hypothesis Testing

n Compare the p-value with α 1. State the null hypothesis, H0 and the alternative hypothesis,
H1
n If p-value < α , reject H0 2. Choose the level of significance, α, and the sample size, n
n If p-value ≥ α , do not reject H0 3. Determine the appropriate test statistic and sampling
distribution

Collect data and compute the value of the test statistic and
n Remember 4.
the p-value
n If the p-value is low then H0 must go 5. Make the statistical decision and state the managerial
conclusion. If the p-value is < α then reject H0, otherwise
do not reject H0. State the managerial conclusion in the
context of the problem
Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-29 Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-30
Chapter 9 9-6

p-value Hypothesis Testing Example p-value Hypothesis Testing Example

(continued)
Test the claim that the true mean
3. Determine the appropriate technique
diameter of a manufactured bolt is 30mm. n σ is assumed known so this is a Z test.
(Assume σ = 0.8) 4. Collect the data, compute the test statistic and the
1. State the appropriate null and alternative p-value
hypotheses n Suppose the sample results are
n H0: µ = 30 H1: µ ≠ 30 (This is a two-tail test) n = 100, X = 29.84 (σ = 0.8 is assumed known)
2. Specify the desired level of significance and the So the test statistic is:
sample size
n Suppose that α = 0.05 and n = 100 are chosen
X − µ 29.84 − 30 − .16
ZSTAT = = = = −2.0
σ 0.8 0.08
for this test
n 100

Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-31 Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-32

p-Value Hypothesis Testing Example:

Calculating the p-value p-value Hypothesis Testing Example
(continued)
4. (continued) Calculate the p-value. n 5. Is the p-value < α?
n How likely is it to get a ZSTAT of -2 (or something farther from the
mean (0), in either direction) if H0 is true? n Since p-value = 0.0456 < α = 0.05 Reject H0
n 5. (continued) State the managerial conclusion
in the context of the situation.
P(Z < -2.0) = 0.0228 P(Z > 2.0) = 0.0228
n There is sufficient evidence to conclude the average diameter of
a manufactured bolt is not equal to 30mm.

0 Z

-2.0 2.0
p-value = 0.0228 + 0.0228 = 0.0456

Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-33 Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-34

Connection Between Two-Tail Tests

Do You Ever Truly Know σ?
and Confidence Intervals
n For X = 29.84, σ = 0.8 and n = 100, the 95% n Probably not!
confidence interval is:
n In virtually all real world business situations, σ is not
0.8 0.8 known.
29.84 - (1.96) to 29.84 + (1.96)
100 100
n If there is a situation where σ is known then µ is also
29.6832 ≤ µ ≤ 29.9968 known (since to calculate σ you need to know µ.)

n Since this interval does not contain the hypothesized n If you truly know µ there would be no need to gather a
mean (30), we reject the null hypothesis at α = 0.05 sample to estimate it.

Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-35 Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-36
Chapter 9 9-7

Hypothesis Testing: t Test of Hypothesis for the Mean

σ Unknown (σ Unknown)
n Convert sample statistic ( X ) to a tSTAT test statistic
n If the population standard deviation is unknown, you
instead use the sample standard deviation S. Hypothesis
Tests for µ
n Because of this change, you use the t distribution instead
of the Z distribution to test the null hypothesis about the
mean. σσKnown
Known σσUnknown
Unknown
(Z test) (t test)
n When using the t distribution you must assume the
population you are sampling from follows a normal The test statistic is:
distribution.
X −µ
All other steps, concepts, and conclusions are the same.
t STAT =
n S
n
Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-37 Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-38

Example: Two-Tail Test Example Solution:

(σ Unknown) Two-Tail t Test
The average cost of a hotel
room in New York is said to H0: µ = 168 α/2=.025 α/2=.025
be $168 per night. To H1: µ ≠ 168
determine if this is true, a
n α = 0.05 Reject H0 Do not reject H0 Reject H0
t 24,0.025
random sample of 25 hotels -t 24,0.025 0
n n = 25, df = 25-1=24 -2.0639 2.0639
is taken and resulted in an X 1.46
of $172.50 and an S of n σ is unknown, so
H0: ______ t STAT =
X−µ
=
172.50 − 168
= 1.46
$15.40. Test the appropriate use a t statistic S 15.40
H1: ______ n 25
hypotheses at α = 0.05. n Critical Value:

(Assume the population distribution is normal) ±t24,0.025 = ± 2.0639 Do not reject H0: insufficient evidence that true
mean cost is different from $168

Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-39 Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-40

Example Two-Tail t Test Using A p- Example Two-Tail t Test Using A

value from Excel p-value from Minitab
1
n Since this is a t-test we cannot calculate the p-value
without some calculation aid.
2
n The Excel output below does this:
t Test for the Hypothesis of the Mean

Data
Null Hypothesis µ= $ 168.00 3
p-value > α
Level of Significance 0.05 So do not reject H0
Sample Size 25
Sample Mean $ 172.50
Sample Standard Deviation $ 15.40

Intermediate Calculations One-Sample T

Standard Error of the Mean $ 3.08 =B8/SQRT(B6)
Degrees of Freedom 24 =B6-1
4
t test statistic 1.46 =(B7-B4)/B11 Test of mu = 168 vs not = 168
Two-Tail Test
p-value > α Lower Critical Value -2.0639 =-TINV(B5,B12) N Mean StDev SE Mean 95% CI T P
Upper Critical Value 2.0639 =TINV(B5,B12)
So do not reject H0 p-value 0.157 =TDIST(ABS(B13),B12,2)
25 172.50 15.40 3.08 (166.14, 178.86) 1.46 0.157
Do Not Reject Null Hypothesis =IF(B18<B5, "Reject null hypothesis",
"Do not reject null hypothesis")

Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-41 Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-42
Chapter 9 9-8

Connection of Two-Tail Tests to

Confidence Intervals One-Tail Tests

n For X = 172.5, S = 15.40 and n = 25, the 95% n In many cases, the alternative hypothesis
confidence interval for µ is: focuses on a particular direction

172.5 - (2.0639) 15.4/ 25 to 172.5 + (2.0639) 15.4/ 25 This is a lower-tail test since the
H 0: µ ≥ 3
alternative hypothesis is focused on
H 1: µ < 3 the lower tail below the mean of 3
166.14 ≤ µ ≤ 178.86
H 0: µ ≤ 3 This is an upper-tail test since the
n Since this interval contains the hypothesized mean (168), alternative hypothesis is focused on
H 1: µ > 3
we do not reject the null hypothesis at α = 0.05 the upper tail above the mean of 3

Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-43 Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-44

Lower-Tail Tests Upper-Tail Tests

H 0: µ ≥ 3
H 0: µ ≤ 3
n There is only one H 1: µ < 3 n There is only one
H 1: µ > 3
critical value, since critical value, since
the rejection area is the rejection area is
in only one tail
α
in only one tail α

Reject H0 Do not reject H0

Z or t Do not reject H0 Reject H0
-Zα or -tα 0 Z or t
0 Zα or tα
µ X _
X µ

Critical value
Critical value

Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-45 Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-46

Example: Upper-Tail t Test

for Mean (σ unknown) Example: Find Rejection Region
(continued)
A phone industry manager thinks that nSuppose that α = 0.10 is chosen for this test and
customer monthly cell phone bills have n = 25.
increased, and now average over $52 per Find the rejection region: Reject H0
month. The company wishes to test this
claim. (Assume a normal population) α = 0.10

Form hypothesis test:

H0: µ ≤ 52 the average is not over $52 per month Do not reject H0 Reject H0
0 1.318
H1: µ > 52 the average is greater than $52 per month
(i.e., sufficient evidence exists to support the
manager’s claim) Reject H0 if tSTAT > 1.318

Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-47 Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-48
Chapter 9 9-9

Example: Test Statistic Example: Decision

(continued) (continued)

Obtain sample and compute the test statistic Reach a decision and interpret the result:
Reject H0

Suppose a sample is taken with the following

results: n = 25, X = 53.1, and S = 10 α = 0.10

n Then the test statistic is:

Do not reject H0 Reject H0
1.318
X−µ 53.1 − 52 0
t STAT = = = 0.55 tSTAT = 0.55
S 10 Do not reject H0 since tSTAT = 0.55 ≤ 1.318
n 25 There is insufficient evidence that the
mean bill is over $52.
Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-49 Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-50

Example: Utilizing The p-value for Excel Spreadsheet Calculating The p-

The Test value for The Upper Tail t Test
n Calculate the p-value and compare to α (p-value below t Test for the Hypothesis of the Mean
calculated using Excel spreadsheet on next page)
Data
p-value = .2937 Null Hypothesis µ= 52.00
Level of Significance 0.1
Sample Size 25
Reject H0 Sample Mean 53.10
α = .10 Sample Standard Deviation 10.00

Intermediate Calculations
0 Standard Error of the Mean 2.00 =B8/SQRT(B6)
Do not reject Reject Degrees of Freedom 24 =B6-1
H0 1.318 H0 t test statistic 0.55 =(B7-B4)/B11
tSTAT = .55
Upper Tail Test
Upper Critical Value 1.318 =TINV(2*B5,B12)
Do not reject H0 since p-value = .2937 > α = .10 p-value
Do Not Reject Null Hypothesis
0.2937 =TDIST(ABS(B13),B12,1)
=IF(B18<B5, "Reject null hypothesis",
"Do not reject null hypothesis")

Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-51 Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-52

Using Minitab to calculate The p-value for

The Upper Tail t Test Hypothesis Tests for Proportions
1

n Involves categorical variables

2
n Two possible outcomes
3 n Possesses characteristic of interest
n Does not possess characteristic of interest
p-value > α
So do not reject H0
n Fraction or proportion of the population in the
One-Sample T category of interest is denoted by π
4 Test of mu = 52 vs > 52

95% Lower
N Mean StDev SE Mean Bound T P
25 53.10 10.00 2.00 49.68 0.55 0.294
Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-53 Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-54
Chapter 9 9-10

Proportions Hypothesis Tests for Proportions

(continued)
n Sample proportion in the category of interest is
denoted by p n The sampling
distribution of p is Hypothesis
X number in category of interest in sample approximately Tests for p
p =n = normal, so the test
n sample size
statistic is a ZSTAT
value: X≥5 X<5
When both X and n – X are at least 5, p can be and or
n
p−π n–X≥5 n–X<5
approximated by a normal distribution with ZSTAT =
mean and standard deviation π (1 − π )
Not discussed
n
µp = π π (1− π ) n
σp = in this chapter
n
Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-55 Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-56

Z Test for Proportion in Terms of

Number in Category of Interest Example: Z Test for Proportion

n An equivalent form A marketing company

to the last slide, Hypothesis claims that it receives
but in terms of the Tests for X responses from 8% of
number in the those surveyed. To test
category of this claim, a random
interest, X: X≥5 X<5
and or sample of 500 were
Check:
n-X ≥ 5 n-X < 5 surveyed with 25
X = 25
ü
responses. Test at the α
X − nπ
ZSTAT = Not discussed
= 0.05 significance level. n-X = 475
nπ (1 − π ) in this chapter

Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-57 Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-58

Z Test for Proportion: Solution p-Value Solution

(continued)
Test Statistic: Calculate the p-value and compare to α
H0: π = 0.08
p −π .05 − .08 (For a two-tail test the p-value is always two-tail)
H1: π ≠ 0.08 ZSTAT = = = −2.47
π (1 − π ) .08(1 − .08) Reject H0
Do not reject H0
Reject H0
α = 0.05 p-value = 0.0136:
n 500 α/2 = .025 α/2 = .025
n = 500, p = 0.05
P(Z ≤ −2.47) + P(Z ≥ 2.47)
Critical Values: ± 1.96 Decision: 0.0068 0.0068
Reject H0 at α = 0.05 = 2(0.0068) = 0.0136
Reject Reject
-1.96 0 1.96
Conclusion:
.025 .025
There is sufficient Z = -2.47 Z = 2.47
-1.96 0 1.96 z evidence to reject the
company’s claim of 8% Reject H0 since p-value = 0.0136 < α = 0.05
-2.47
response rate.
Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-59 Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-60
Chapter 9 9-11

Potential Pitfalls and

Ethical Considerations Chapter Summary

Use randomly collected data to reduce selection biases

n
n Addressed hypothesis testing
n Do not use human subjects without informed consent
n Choose the level of significance, α, and the type of test
methodology
(one-tail or two-tail) before data collection
n Performed Z Test for the mean (σ known)
n Do not employ “data snooping” to choose between one-
tail and two-tail test, or to determine the level of n Discussed critical value and p–value
significance
approaches to hypothesis testing
n Do not practice “data cleansing” to hide observations
that do not support a stated hypothesis n Performed one-tail and two-tail tests
n Report all pertinent findings including both statistical
significance and practical importance
Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-61 Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-62

Chapter Summary
(continued)

n Performed t test for the mean (σ

Basic Business Statistics
unknown) 12th Edition

n Performed Z test for the proportion

Online Topic
n Discussed pitfalls and ethical issues
Power of A Hypothesis Test

Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-63 Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-64

The Power of a Test Type II Error

n The power of the test is the probability of
correctly rejecting a false H0 n Suppose we do not reject H0: µ ≥ 52 when in fact
the true mean is µ = 50
Suppose we correctly reject H0: µ ≥ 52
when in fact the true mean is µ = 50 This is the true
This is the range of X where
H0 is not rejected
distribution of X if µ = 50
Power Prob. of type II
= 1-β error = β
α

50 52 50 52
Reject Do not reject Reject Do not reject
H0: µ ≥ 52 H0 : µ ≥ 52 H0: µ ≥ 52 H0 : µ ≥ 52

Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-65 Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-66
Chapter 9 9-12

Type II Error Calculating β

(continued)
n Suppose n = 64 , σ = 6 , and α = .05
n Suppose we do not reject H0: µ ≥ 52 when in fact
the true mean is µ = 50 σ 6
cutoff = Xα = µ − Zα = 52 − 1.645 = 50.766
(for H0 : µ ≥ 52) n 64
Here, β = P( X ≥ cutoff ) if µ = 50
So β = P( X ≥ 50.766 ) if µ = 50

α β α

50 52 50 50.766 52
Reject Do not reject Reject Do not reject
H0: µ ≥ 52 H0 : µ ≥ 52 H0: µ ≥ 52 H0 : µ ≥ 52

Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-67 Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-68

Calculating β and
Power of the test Power of the Test
(continued)

n Suppose n = 64 , σ = 6 , and α = 0.05 n Conclusions regarding the power of the test:

⎛ ⎞
⎜ 50.766 − 50 ⎟
P( X ≥ 50.766 | µ = 50) = P⎜ Z ≥ ⎟ = P(Z ≥ 1.02) = 1.0 − 0.8461 = 0.1539 n A one-tail test is more powerful than a two-tail test
⎜ 6 ⎟
⎝ 64 ⎠ n An increase in the level of significance (α) results in
an increase in power
Power Probability of n An increase in the sample size results in an
=1-β type II error: increase in power
= 0.8461
β = 0.1539

The probability of 50 50.766 52

correctly rejecting a
Reject Do not reject
false null hypothesis is H0: µ ≥ 52 H0 : µ ≥ 52
0.8641
Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-69 Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-70

Online Topic Summary

n Examined how to calculate and interpret

the power of a hypothesis test.

Copyright ©2012 Pearson Education, Inc. publishing as Prentice Hall Chap 9-71