Testing Hypothesis

Two- Sample Test

 Agenda

• Tests For Differences between

Means : Large Sample Sizes

• Tests For Differences between Means:

Small Sample Sizes

• Testing Differences between Means

with Dependent Samples

• Tests for Differences between

Proportions : Large Sample Sizes
 INDEPENDENT SAMPLE
Larger Sample Smaller Sample
Type of test Z – distribution t – distribution
Step 1 Define Ho & H1 Define Ho & H1

Step 2 Calculate Critical Values Calculate Critical Values

t (critical)
z (critical) Considering Degrees of Freedom
Step 3 Calculate ZH0 Calculate tH0

(𝑥ҧ1−𝑥ҧ2)−(𝜇1−𝜇2) (𝑥ҧ1−𝑥ҧ2 )−(𝜇1−𝜇2 )

z=
𝜎𝑥ഥ1 − 𝑥ഥ2
t=
σ 𝑛1 +𝑛1
p
1 2
Pooled Estimate
Pooled Estimate
Step 4 |Critical Values| > |Calculated Values|
(Interpret) Ho is not Rejected

|Calculated Values| > |Critical Values|

Ho is Rejected
 ഥ
𝒙 s n Problem 1
Semiautomatic line 198 32 150 Block Enterprises, a
manufacturer of chips for
Automatic line 206 29 200 computer, is in the process
of deciding whether to
replace its current
semiautomated assembly
line with a fully automated
assembly line. Block has
gathered some preliminary
test data about hourly chip
production, which is
summarized in the
following table (alongside),
and it would like to know
whether it should upgrade
its assembly line.

State (and test at α = 0.02)

appropriate hypotheses to
help Block decide.
 Solution to Problem 1
Standard Deviation of Sample 1 : 𝒔𝟏 = 𝟑𝟐
Mean of Sample 1 : 𝒙ഥ𝟏 = 𝟏𝟗𝟖
Size of Sample 1: 𝒏𝟏 = 150

Standard Deviation of Sample 2 : 𝒔𝟐 = 𝟐𝟗

Mean of Sample 2 : 𝒙ഥ𝟐 = 𝟐𝟎𝟔
Size of Sample 2: 𝒏𝟐 = 200

Sample 1 : Semi-Automatic Lines

Sample 2 : Automatic Lines
 Hypothesis
𝑯𝟎 : 𝝁𝟏 = 𝝁𝟐

There is no significant difference between the output from the

semi-automatic line and the output from the automatic line

𝑯𝟏 : 𝝁𝟏 < 𝝁𝟐
The output from the semi-automatic line is significantly lesser than
that from the automatic line

Alternatively,
𝑯𝟏 : 𝝁𝟐 > 𝝁𝟏
The output from the automatic line is significantly greater than that
from the semi- automatic line

Significance Level = 𝛼 = 0.02

 Z- Table
(1-Tailed)

𝜶 = 𝟎. 𝟎𝟐

Look for 0.4800

𝒁𝑳𝑳 = −𝟐. 𝟎𝟓
or
𝒁𝑪𝒓𝒊𝒕𝒊𝒄𝒂𝒍 = −𝟐. 𝟎𝟓
−𝟐. 𝟎𝟓
 Standardizing in terms of
Z- Statistics
(𝑥ҧ1 −𝑥ҧ2 )−(𝜇1 −𝜇2 )
z=
𝜎𝑥ഥ1 − 𝑥ഥ2

Standard Error of the Difference between Two Means

𝒔𝟏 𝟐 𝒔𝟐 𝟐
𝜎𝑥ҧ1− 𝑥ҧ2 = +
𝒏𝟏 𝒏𝟐
 Standard Error of the Difference
between Two Means
𝒔𝟏 𝟐 𝒔𝟐 𝟐
𝜎𝑥ҧ1− 𝑥ҧ2 = +
𝒏𝟏 𝒏𝟐

𝟑𝟐𝟐 +𝟐𝟗𝟐
➔𝜎𝑥ҧ1 − 𝑥ҧ2 =
𝟏𝟓𝟎 𝟐𝟎𝟎

➔𝜎𝑥ҧ1 − 𝑥ҧ2 = 3.3214 Chips per hour

 Standardizing in terms of
Z- Statistics
(𝑥ҧ1 −𝑥ҧ2 )−(𝜇1 −𝜇2 )
z=
𝜎𝑥ഥ1 − 𝑥ഥ2

(198−206)−(0)
➔
3.3214

➔- 2.41
𝑍𝐻0 = −2.41
−2.41 −𝟐. 𝟎𝟓
 Conclusion
| 𝑍𝐻0 | > |𝑍𝐶𝑟𝑖𝑡𝑖𝑐𝑎𝑙 |

➔ |- 2.41| > | -2.05|

𝑊𝑒 𝑅𝐸𝐽𝐸𝐶𝑇 𝐻𝑜 𝑖𝑛 𝑓𝑎𝑣𝑜𝑟 𝑜𝑓𝑯𝟏

The output from the automatic line is

significantly greater than that from the
semi- automatic line
 Problem 2
Notwithstanding the Equal Pay Act of 1976,
in 2015 it still appeared that men earned
more than women in similar jobs. A
random sample of 38 male machine-tool
operators found a mean hourly wage of Rs.
113.8, and the sample standard deviation
was Rs. 18.4. A random sample of 45
female machine-tool operators found their
mean wage to be Rs. 84.2, and the sample
standard deviation was Rs. 13.1. On the
basis of these samples, is it reasonable to
conclude (at α = 0.01) that the male
operators are earning over Rs. 20.0 more
per hour than the female operators?
 Solution to Problem 2
Standard Deviation of Sample 1 : 𝒔𝟏 = 𝟏𝟖. 𝟒
Mean of Sample 1 : 𝒙ഥ𝟏 = 𝟏𝟏𝟑. 𝟖
Size of Sample 1: 𝒏𝟏 = 38

Standard Deviation of Sample 2 : 𝒔𝟐 = 𝟏𝟑. 𝟏

Mean of Sample 2 : 𝒙ഥ𝟐 = 𝟖𝟒. 𝟐
Size of Sample 2: 𝒏𝟐 = 45

Sample 1 : Male
Sample 2 : Female
 Hypothesis
𝑯𝟎 : 𝝁𝟏 = 𝝁𝟐 + 20
𝑯𝟎 : 𝝁𝟏 − 𝝁𝟐 =20
Male operators are earning exactly Rs 20
more per hour than the female operators

𝑯𝟏 : 𝝁𝟏 − 𝝁𝟐 > 𝟐𝟎
Male operators are earning over Rs 20
more per hour than the female operators

Significance Level = 𝛼 = 0.01

 Z- Table
(1-Tailed)

𝜶 = 𝟎. 𝟎𝟏

Look for 0.49

𝒁𝑼𝑳 = 𝟐. 𝟑𝟑
or
𝒁𝑪𝒓𝒊𝒕𝒊𝒄𝒂𝒍 = 𝟐. 𝟑𝟑
2. 𝟑𝟑
 Standard Error of the Difference
between Two Means

𝒔𝟏 𝟐 𝒔𝟐 𝟐
𝜎𝑥ҧ1− 𝑥ҧ2 = +
𝒏𝟏 𝒏𝟐

𝟏𝟖.𝟒𝟐 𝟏𝟑.𝟏𝟐
➔𝜎𝑥ҧ1 − 𝑥ҧ2 = +
𝟑𝟖 𝟒𝟓

➔𝜎𝑥ҧ1 − 𝑥ҧ2 = 3.566

 Standardizing in terms of
Z- Statistics
(𝑥ҧ1 −𝑥ҧ2 )−(𝜇1 −𝜇2 )
z=
𝜎𝑥ഥ1 − 𝑥ഥ2

(113.8−84.2)−(20)
➔
3.566

➔ 2.7
𝑍𝐻0 = 2.7
2. 𝟑𝟑 < 2.7
 Conclusion
| 𝑍𝐻0 | > |𝑍𝐶𝑟𝑖𝑡𝑖𝑐𝑎𝑙 |

➔ |2.7| > |2.33|

𝑊𝑒 𝑅𝐸𝐽𝐸𝐶𝑇 𝐻𝑜 𝑖𝑛 𝑓𝑎𝑣𝑜𝑟 𝑜𝑓𝑯𝟏

Male operators are earning over Rs 20

more per hour than the female
operators
 Agenda

• Tests For Differences between Means :

Large Sample Sizes

• Tests For Differences

between Means: Small
Sample Sizes

• Testing Differences between Means

with Dependent Samples

• Tests for Differences between

Proportions : Large Sample Sizes
 INDEPENDENT SAMPLE
Larger Sample Smaller Sample
Type of test Z – distribution t – distribution
Step 1 Define Ho & H1 Define Ho & H1

Step 2 Calculate Critical Values Calculate Critical Values

t (critical)
z (critical) Considering Degrees of Freedom
Step 3 Calculate ZH0 Calculate tH0

(𝑥ҧ1−𝑥ҧ2)−(𝜇1−𝜇2) (𝑥ҧ1−𝑥ҧ2 )−(𝜇1−𝜇2 )

z=
𝜎𝑥ഥ1 − 𝑥ഥ2
t=
σ 𝑛1 +𝑛1
p
1 2
Pooled Estimate
Pooled Estimate
Step 4 |Critical Values| > |Calculated Values|
(Interpret) Ho is not Rejected

|Calculated Values| > |Critical Values|

Ho is Rejected
 Problem 3
A large stock-brokerage firm wants to
determine how successful its new account
executives have been at recruiting clients.
After completing their training, new
account execs spend several weeks calling
prospective clients, trying to get the
prospects to open accounts with the firm.
The following data give the number of
new accounts opened in their first 2
weeks by 10 randomly chosen female
account execs and by randomly chosen
male account execs. At α = 0.05, does it
appear that the women are more effective
at generating new accounts than the men
are?
Number of New Accounts
Female account execs 12 11 14 13 13 14 13 12 14 12
Male account execs 13 10 11 12 13 12 10 12
 Female Account Executives
Number of New Accounts 𝑥𝑖 − 𝑥ҧ (𝑥𝑖 − 𝑥)ҧ 2
12 - 0.8 0.64
11 -1.8 3.24
14 1.2 1.44
13 0.2 0.04
13 0.2 0.04
14 1.2 1.44
13 0.2 0.04
12 -0.8 0.64
14 1.2 1.44
12 -0.8 0.64

Mean: 𝑥1 = 12.8 σ(𝑥𝑖 − 𝑥ҧ )2 = 9.6

 Male Account Executives
Number of New Accounts 𝑥𝑖 − 𝑥ҧ (𝑥𝑖 − 𝑥)ҧ 2
13 1.375 1.891
10 -1.625 2.641
11 -0.625 0.391
12 0.375 0.141
13 1.375 1.891
12 0.375 0.141
10 -1.625 2.641
12 0.375 0.141

Mean: 𝑥1 = 11.625 σ(𝑥𝑖 − 𝑥ҧ )2 = 9.875

 Variance

2
σ(𝑥𝑖 − 𝑥ҧ )
𝑠2 =
𝑛−1
 Solution to Problem 3

Variance of Sample 1 :𝑺𝟐𝟏 = 𝟏. 𝟎𝟔𝟔𝟕

Mean of Sample 1 : 𝒙ഥ𝟏 = 𝟏𝟐. 𝟖
Size of Sample 1: 𝒏𝟏 = 10

Variance of Sample 2 :𝑺𝟐𝟐 = 𝟏. 𝟒𝟏𝟎𝟕

Mean of Sample 2 : 𝒙ഥ𝟐 = 𝟏𝟏. 𝟔𝟐𝟓
Size of Sample 2: 𝒏𝟐 = 8

Sample 1 : Female
Sample 2 : Male
 Hypothesis
𝑯𝟎 : 𝝁𝟏 = 𝝁𝟐
The women are equally effective as the
men in generating new accounts

𝑯𝟏 : 𝝁𝟏 > 𝝁𝟐
The women are significantly more
effective than the men in generating new
accounts

Significance Level = 𝛼 = 0.05

 t- Table
(1-Tailed)

𝜶 = 𝟎. 𝟎𝟓

Look for
Df = 18 - 2 = 16

𝒕𝑼𝑳 = 𝟏. 𝟕𝟒𝟔
or
𝒕𝑪𝒓𝒊𝒕𝒊𝒄𝒂𝒍 = 𝟏. 𝟕𝟒𝟔
𝟏. 𝟕𝟒𝟔
 Standardizing in terms of
t- Statistics
(𝑥ҧ1 −𝑥ҧ2 )−(𝜇1 −𝜇2 )
t= 1 1
σ 𝑛 +𝑛
p
1 2

Standard Error (Pooled Estimate)

(𝑛1 −1) 𝑠1 2 + (𝑛2 −1) 𝑠2 2

𝜎𝑝 =
𝑛1 +𝑛2 −2
 Standard Error (Pooled Estimate)
(𝑛1 −1) 𝑠1 2 + (𝑛2 −1) 𝑠2 2
𝜎𝑝 =
𝑛1 +𝑛2 −2

(10−1) 1.06670 +(8−1) 1.41072

➔
10+8−2

➔1.1032 Accounts
 Standardizing in terms of
t- Statistics
(𝑥ҧ1 −𝑥ҧ2 )−(𝜇1 −𝜇2 )
t= 1 1
σ 𝑛 +𝑛
p
1 2

(12.8−11.625)−(0)
➔
1 1
1.1032 +
10 8

➔2.40
𝑡𝐻0 = 2.40
𝟏. 𝟕𝟒𝟔 2.40
 Conclusion
| 𝑡𝐻0 | > |𝑡𝐶𝑟𝑖𝑡𝑖𝑐𝑎𝑙 |

➔ |2.40| > |𝟏. 𝟕𝟒𝟔|

𝑊𝑒 𝑅𝐸𝐽𝐸𝐶𝑇 𝐻𝑜 𝑖𝑛 𝑓𝑎𝑣𝑜𝑟 𝑜𝑓𝑯𝟏

The women are significantly more

effective than the men in generating
new accounts
 Problem 4
Because refunds are paid more
quickly on tax returns that are filed
electronically, the Commissioner of
the Internal Revenue Service was
wondering whether refunds due on
returns filed by mail were smaller
than those due on returns filed
electronically. Looking only at
returns claiming refunds, a sample
of 17 filed by mail had an average
refund of $563, and a standard
deviation of $378. The average
refund on a sample of 13
electronically filed returns was $958,
and the sample standard deviation
was $619. At significance level of
0.01, do these data support the
commissioner’s speculation?
 Solution to Problem 4
Standard Deviation of Sample 1 : 𝒔𝟏 = 𝟑𝟕𝟖
Mean of Sample 1 : 𝒙ഥ𝟏 = 𝟓𝟔𝟑
Size of Sample 1: 𝒏𝟏 = 17

Standard Deviation of Sample 2 : 𝒔𝟐 = 𝟔𝟏𝟗

Mean of Sample 2 : 𝒙ഥ𝟐 = 𝟗𝟓𝟖
Size of Sample 2: 𝒏𝟐 = 13

Sample 1 : Mail
Sample 2 : Electronic
 Hypothesis
𝑯𝟎 : 𝝁𝟏 = 𝝁𝟐
Refunds due on returns filed by mail were similar
to those due on return filed electronically

𝑯𝟏 : 𝝁𝟏 < 𝝁𝟐
Refunds due on returns filed by mail were smaller
than those due on return filed electronically

Significance Level = 𝛼 = 0.01

 Solution to Problem 4

t- Table
(1-Tailed)

𝜶 = 𝟎. 𝟎𝟏

Look for
Df = 30 - 2 = 28

𝒕𝑳𝑳 = −𝟐. 𝟒𝟔𝟕

or
𝒕𝑪𝒓𝒊𝒕𝒊𝒄𝒂𝒍 = −𝟐. 𝟒𝟔𝟕
- 2.467
 Standard Error (Pooled Estimate)
(𝑛1 −1) 𝑠1 2 + (𝑛2 −1) 𝑠2 2
𝜎𝑝 =
𝑛1 +𝑛2 −2

(17−1) 3782 +(13−1) 6192

➔
17+13−2

➔$ 495.84
 Standardizing in terms of
t- Statistics
(𝑥ҧ1 −𝑥ҧ2 )−(𝜇1 −𝜇2 )
t= 1 1
σ 𝑛 +𝑛
p
1 2

(563−958)−(0)
➔
1 1
495.84 +
17 13

➔- 2.162
𝑡𝐻0 = −2.162
 - 2.162

- 2.467
 Conclusion
| 𝑡𝐻0 | < |𝑡𝐶𝑟𝑖𝑡𝑖𝑐𝑎𝑙 |

➔ |-2.162| < |-2.467|

𝑊𝑒 𝐷𝑜𝑛𝑜𝑡 𝑅𝐸𝐽𝐸𝐶𝑇 𝐻𝑜
Refunds due on returns filed by
mail were similar to those due on
return filed electronically
 Agenda

• Tests For Differences between Means :

Large Sample Sizes

• Tests For Differences between Means:

Small Sample Sizes

• Testing Differences
between Means with
Dependent Samples
• Tests for Differences between
Proportions : Large Sample Sizes
 DEPENDENT SAMPLE
Smaller Sample

Type of test t – distribution

Step 1 Define Ho & H1
Step 2 Calculate Critical Values
t (critical)
Considering Degrees of Freedom

Step 3 Calculate tH0

x̅ −µ
t= σ x

Where x̅ = 𝑎𝑣𝑒𝑟𝑎𝑔𝑒 𝑑𝑖𝑓𝑓𝑒𝑟𝑒𝑛𝑐𝑒 𝑏𝑒𝑡𝑤𝑒𝑒𝑛 𝑝𝑎𝑖𝑟𝑠

Pooled Standard Deviation calculation
Step 4 |Critical Values| > |Calculated Values|
(Interpret) Ho is not Rejected

|Calculated Values| > |Critical Values|

Ho is Rejected
Problem 5
Aquarius Health Club has been
advertising a rigorous program for
body conditioning. The club claims
that after 1 month in the program,
the average participant should be
able to do eight more push-ups in
2 minutes than he or she could do
at the start. Does the random
sample of 10 program participants
given below support the club’s
claim? Use the 0.025 Level of
Significance.

1 2 3 4 5 6 7 8 9 10
Before 38 11 34 25 17 38 12 27 32 29
After 45 24 41 39 30 44 30 39 40 41
Solution to Problem 5
Participant Before After

1 38 45

2 11 24

3 34 41

4 25 39

5 17 30

6 38 44

7 12 30

8 27 39

9 32 40

10 29 41
Standard Deviation
2
σ 𝑥 − 𝑛𝑥ҧ 2
S=
𝑛−1

Note:
𝒙 = (𝒙𝑨 −𝒙𝑩 )
Participant Before After Change (𝒙𝑨 −𝒙𝑩 )𝟐
𝒙𝑩 𝒙𝑨 𝒙𝑨 − 𝒙𝑩

1 38 45 7 49
2 11 24 13 169
3 34 41 7 49
4 25 39 14 196
5 17 30 13 169
6 38 44 6 36
7 12 30 18 324
8 27 39 12 144
9 32 40 8 64
10 29 41 12 144
Sum 110 1344 ෍ 𝑥2
Average (𝑥)ҧ 11
 Standard Deviation

2
σ 𝑥 − 𝑛𝑥ҧ 2
S=
𝑛−1

1344 −10×(11)2
=
10−1

= 3.8586
 Standard Error (Characterizing 1-Sample HT)
𝑠
𝜎𝑥ҧ =
𝑛
3.8586
➔
10

➔ 1.2202 push-ups
 Hypothesis
𝑯𝟎 : 𝝁𝑨 = 𝝁𝑩 +8
No Change in Body Conditioning

𝑯𝟏 : 𝝁𝑨 − 𝝁𝑩 > 𝟖
The data supports the club’s claim

Significance Level = 𝛼 = 0.025

 t- Table
(1-Tailed)

𝜶 = 𝟎. 𝟎𝟐𝟓

Look for
Df = 10 - 1 = 9

𝒕𝑼𝑳 = 𝟐. 𝟐𝟔𝟐
or
𝒕𝑪𝒓𝒊𝒕𝒊𝒄𝒂𝒍 = 𝟐. 𝟐𝟔𝟐
2.262
 Standardizing in terms of
t- Statistics
x̅ −µ Note:
t=
𝜎𝑥ഥ 𝑋ത = 𝒙𝑨 − 𝒙𝑩
µ = 𝝁𝑨 − 𝝁𝑩
11− 8
➔
1.2202

➔ 2.459
𝑡𝐻0 = 2.459
2.262 2.459
 Conclusion
| 𝑡𝐻0 | > |𝑡𝐶𝑟𝑖𝑡𝑖𝑐𝑎𝑙 |

➔ |2.459| > |𝟐. 𝟐𝟔𝟐|

𝑊𝑒 𝑅𝐸𝐽𝐸𝐶𝑇 𝐻𝑜 𝑖𝑛 𝑓𝑎𝑣𝑜𝑟 𝑜𝑓𝑯𝟏

The data supports the club’s claim

 Problem 6

Jeff Richardson, the receiving clerk for

a chemical-products distributor, is
faced with the continuing problem of
broken glassware, including test-tubes,
petri dishes, and flasks. Jeff has
determined some additional shipping
precautions that can be undertaken to
prevent breakage, and he has asked
the Purchasing Director to inform the
suppliers of the new measures. Data
for 8 suppliers are given below in Supplier 1 2 3 4 5 6 7 8
terms of average number of broken
items per shipment. Before 16 12 18 7 14 19 6 17
Do the data indicate, at 𝜶 = 0.05, that
the new measures have lowered the After 14 13 12 6 9 15 8 15
average number of broken items?
Supplier Before After Change 𝒙𝑨 − 𝒙𝑩 𝟐
𝒙𝑩 𝒙𝑨 𝒙𝑨 − 𝒙𝑩

1 16 14 -2 4

2 12 13 1 1

3 18 12 -6 36

4 7 6 -1 1

5 14 9 -5 25

6 19 15 -4 16

7 6 8 2 4

8 17 15 -2 4

Sum -17 91 ෍ 𝑥2

Average (𝑥)ҧ -2.125

 Standard Deviation

2
σ 𝑥 − 𝑛𝑥ҧ 2
S=
𝑛−1

91 − 8×(−2.125)2
=
8−1

= 2.7999 Items
Note:
𝒙 = (𝒙𝑨 −𝒙𝑩 )
 Standard Error
𝑠
𝜎𝑥ҧ =
𝑛
2.7999
➔
8

➔ 0.9899 items
 Hypothesis
𝑯𝟎 : 𝝁𝑨 = 𝝁𝑩
The new measures have not significantly lowered
the average number of broken items

𝑯𝟏 : 𝝁𝑨 < 𝝁𝑩
The new measures have lowered the average
number of broken items

Significance Level = 𝛼 = 0.05

 Solution to Problem 6
t- Table
(1-Tailed)

𝜶 = 𝟎. 𝟎𝟓

Look for
Df = 8 - 1 = 7

𝒕𝑳𝑳 = −𝟏. 𝟖𝟗𝟓

or
𝒕𝑪𝒓𝒊𝒕𝒊𝒄𝒂𝒍 = −𝟏. 𝟖𝟗𝟓
- 1.895
 Standardizing in terms of
t- Statistics
x̅ −µ Note:
t=
𝜎𝑥ഥ 𝑋ത = 𝒙𝑨 − 𝒙𝑩
µ = 𝝁𝑨 − 𝝁𝑩
−2.125−0
➔
0.9899

➔-2.147
𝑡𝐻0 = −2.147
- 2.147 - 1.895
 Conclusion
| 𝑡𝐻0 | > |𝑡𝐶𝑟𝑖𝑡𝑖𝑐𝑎𝑙 |

➔ |-2.147| > |-1.895|

𝑊𝑒 𝑅𝐸𝐽𝐸𝐶𝑇 𝐻𝑜 𝑖𝑛 𝑓𝑎𝑣𝑜𝑟 𝑜𝑓𝑯𝟏

The new measures have lowered the

average number of broken items
 Agenda

• Tests For Differences between Means :

Large Sample Sizes

• Tests For Differences between Means:

Small Sample Sizes

• Testing Differences between Means

with Dependent Samples

• Tests for Differences

between Proportions :
Large Sample Sizes
 INDEPENDENT SAMPLE
Proportion

Type of test Z – distribution

Step 1 Define Ho & H1

Step 2 Calculate Critical Values

Z (critical)
Step 3 Calculate ZH0
(𝑝ҧ1 −𝑝ҧ2 )−(𝑝1 −𝑝2 )
z=
𝜎𝑝
ഥ 1 −𝑝
ഥ2
Pooled Proportion of Success

Step 4 |Critical Values| > |Calculated Values|

(Interpret) Ho is not Rejected

|Calculated Values| > |Critical Values|

Ho is Rejected
 Problem 7
MacroSwift has recently released a
new word-processing product, and
they are interested in determining
whether people in the 30-39 age
group rate the program any
differently than members of the 40-
49 age group. MacroSwift randomly
sampled 175 people in the 30-39 age
group who purchased the product
and found 87 people who rated the
program as excellent, with 52 people
who would purchase an upgrade.
They also sampled 220 people in the
40-49 age group and found 94 people
who gave an excellent rating, with 37
people who plan to purchase an
upgrade.

Is there any significant difference in

the proportions of people in the two
age groups who rate the program as
excellent at the α = 0.05 level?

Is the same result true for

proportions of people who plan to
purchase an upgrade?
 Solution to Problem 7
(Excellent)
𝟖𝟕
Sample proportion of successes: 𝒑 ഥ𝟏 = = 𝟎. 𝟒𝟗𝟕𝟏
𝟏𝟕𝟓
ഥ𝟏 = 𝟏 − 𝒑
Sample proportion of failures: 𝒒 ഥ𝟏 = 𝟎. 𝟓𝟎𝟐𝟗
Size of Sample 1: 𝒏𝟏 = 175

𝟗𝟒
Sample proportion of successes: 𝒑 ഥ𝟐 = = 𝟎. 𝟒𝟐𝟕𝟑
𝟐𝟐𝟎
ഥ𝟐 = 𝟏 − 𝒑
Sample proportion of failures: 𝒒 ഥ𝟐 = 𝟎. 𝟓𝟕𝟐𝟕
Size of Sample 2: 𝒏𝟐 = 220

Sample 1 : 30 - 39 age group

Sample 2 : 40 – 49 age group
 Hypothesis
𝑯 𝟎 : 𝒑𝟏 = 𝒑𝟐

There is no significant difference in the proportions

of people in the two age groups who rate the
program as excellent

𝑯 𝟏 : 𝒑𝟏 ≠ 𝒑𝟐
There is a significant difference in the proportions of
people in the two age groups who rate the program
as excellent

Significance Level = 𝛼 = 0.05

 Z- Table
(2-Tailed)

𝜶 = 𝟎. 𝟎𝟓

Look for 0.475

𝒁𝑳𝑳 = −𝟏. 𝟗𝟔
𝒁𝑼𝑳 = 𝟏. 𝟗𝟔
or
𝒁𝑪𝒓𝒊𝒕𝒊𝒄𝒂𝒍 = ±𝟏. 𝟗𝟔
−𝟏. 𝟗𝟔 𝟏. 𝟗𝟔
 Weighted Average Concept

Best estimate of the overall proportion of

successes in the population if the two proportions
are hypothesized to be equal

𝑛1 𝑝1 + 𝑛2 𝑝2
𝑝ҧ =
𝑛1 + 𝑛2
Overall proportion of successes in
two population
𝑛1 𝑝1 + 𝑛2 𝑝2
ഥ=
𝒑
𝑛1 + 𝑛2

175 0.4971 + 220 (0.4273)

=
175+220

= 0.4582

ഥ = 1 − 𝑝ҧ = 0.5418
𝒒
 Estimated Standard Error
1 1
𝝈 ഥ𝟏 − 𝒑
𝒑 ഥ𝟐 = 𝑝ҧ 𝑞ത ( + )
𝑛1 𝑛2

1 1
➔ 0.4582 × 0.5418 × (
175
+
220
)

➔ 0.0505
 Standardizing in terms of
Z- Statistics
(𝑝ҧ1 −𝑝ҧ2 )−(𝑝1 −𝑝2 )
z=
𝜎𝑝
ഥ 1 −𝑝
ഥ2

0.4971−0.4273 − 0
➔
0.0505

➔ 1.38
𝑍𝐻0 = 1.38
−𝟏. 𝟗𝟔 𝟏. 𝟑𝟖 𝟏. 𝟗𝟔
 Conclusion
| 𝑍𝐻0 | < |𝑍𝐶𝑟𝑖𝑡𝑖𝑐𝑎𝑙 |

➔ |1.38| < | ±𝟏. 𝟗𝟔|

𝑊𝑒 𝐷𝑜𝑛𝑜𝑡 𝑅𝐸𝐽𝐸𝐶𝑇 𝐻𝑜
The proportions of people in the two
age groups who rate the program as
excellent are not significantly different
 Solution to Problem 7
(Upgrade)
𝟓𝟐
Sample proportion of successes: 𝒑 ഥ𝟏 = = 𝟎. 𝟐𝟗𝟕𝟏
𝟏𝟕𝟓
ഥ𝟏 = 𝟏 − 𝒑
Sample proportion of failures: 𝒒 ഥ𝟏 = 𝟎. 𝟕𝟎𝟐𝟗
Size of Sample 1: 𝒏𝟏 = 175

𝟑𝟕
ഥ𝟐 =
Sample proportion of successes: 𝒑 = 𝟎. 𝟏𝟔𝟖𝟏
𝟐𝟐𝟎
ഥ𝟐 = 𝟏 − 𝒑
Sample proportion of failures: 𝒒 ഥ𝟐 = 𝟎. 𝟖𝟑𝟏𝟗
Size of Sample 2: 𝒏𝟐 = 220

Sample 1 : 30 - 39 age group

Sample 2 : 40 – 49 age group
 Hypothesis
𝑯𝟎 : 𝒑𝟏 = 𝒑𝟐

There is no significant difference in the proportions of

people in the two age groups who wish to upgrade

𝑯 𝟏 : 𝒑𝟏 ≠ 𝒑𝟐
There is a significant difference in the proportions of
people in the two age groups who wish to upgrade
Significance Level = 𝛼 = 0.05
 Z- Table
(2-Tailed)

𝜶 = 𝟎. 𝟎𝟓

Look for 0.475

𝒁𝑳𝑳 = −𝟏. 𝟗𝟔
𝒁𝑼𝑳 = 𝟏. 𝟗𝟔
or
𝒁𝑪𝒓𝒊𝒕𝒊𝒄𝒂𝒍 = ±𝟏. 𝟗𝟔
−𝟏. 𝟗𝟔 𝟏. 𝟗𝟔
 Weighted Average Concept

Best estimate of the overall proportion of

successes in the population if the two proportions
are hypothesized to be equal

𝑛1 𝑝1 + 𝑛2 𝑝2
𝑝ҧ =
𝑛1 + 𝑛2
Overall proportion of successes in
two population
𝑛1 𝑝1 + 𝑛2 𝑝2
ഥ=
𝒑
𝑛1 + 𝑛2

175 0.2971 + 220 (0.1681)

=
175+220

= 0.2252

ഥ = 1 − 𝑝ҧ = 0.7748
𝒒
 Estimated Standard Error
1 1
𝝈 ഥ𝟏 − 𝒑
𝒑 ഥ𝟐 = 𝑝ҧ 𝑞ത ( + )
𝑛1 𝑛2

1 1
➔ 0.225 × 0.7748 × (
175
+
220
)

➔ 0.0423
 Standardizing in terms of
Z- Statistics
(𝑝ҧ1 −𝑝ҧ2 )−(𝑝1 −𝑝2 )
z=
𝜎𝑝
ഥ 1 −𝑝
ഥ2

0.2971−0.1681 − 0
➔
0.0423

➔ 3.049
𝑍𝐻0 = 3.049
−𝟏. 𝟗𝟔 𝟏. 𝟗𝟔 3. 𝟎𝟒𝟗
 Conclusion
| 𝑍𝐻0 | > |𝑍𝐶𝑟𝑖𝑡𝑖𝑐𝑎𝑙 |

➔ |3.049| > | ±𝟏. 𝟗𝟔|

𝑊𝑒 𝑅𝐸𝐽𝐸𝐶𝑇 𝐻𝑜
There is a significant difference in
the proportions of people in the two
age groups who wish to upgrade