If 𝑥1 and 𝑥2 are the means of independent random

samples of size 𝑛1 and 𝑛2 from populations with known

variances 𝜎12 and 𝜎22 , respectively, a (1-α)100% confidence
interval for 𝜇1 − 𝜇2 is given by

𝜎12 𝜎22 𝜎12 𝜎22

𝑥1 − 𝑥2 − 𝑧𝛼/2 + < 𝜇1 − 𝜇2 < 𝑥1 − 𝑥2 + 𝑧𝛼/2 +
𝑛1 𝑛2 𝑛1 𝑛2

Where 𝑧𝛼/2 is the z value leaving an area of α/2 to the right.

1
A standardized chemistry test was given to 50
girls and 75 boys. The girls made an average
grade of 76 with a standard deviation of 6,
while the boys made an average grade of 82
with a standard deviation of 8. Find a 96%
confidence interval for the difference 𝜇1 − 𝜇2 ,
where 𝜇1 is the mean score of all boys and 𝜇2
is the mean score of all girls who might take
this test.
2
𝑥1 − 𝑥2 =
82 – 76 = 6
𝜎1 = 8
𝜎2 = 6
@96% confidence interval 𝑧𝛼/2 = 𝑧0.02 = 2.05.

3
 𝜎12 𝜎22 𝜎12 𝜎22
𝑥1 − 𝑥2 − 𝑧𝛼 + < 𝜇1 − 𝜇2 < 𝑥1 − 𝑥2 + 𝑧𝛼/2 +
2 𝑛1 𝑛2 𝑛1 𝑛2

82 62 82 62
82 − 76 − 2.05 + < 𝜇1 − 𝜇2 < 82 − 76 + 2.05 +
75 50 75 50

3.429 < 𝜇1 − 𝜇2 < 8.571

4
 If 𝑥1 and 𝑥2 are the means of small independent random
samples of size 𝑛1 and 𝑛2 , respectively, from approximate
normal populations with unknown but equal variances, a (1-
α)100% confidence interval for 𝜇1 − 𝜇2 is given by

1 1 1 1
𝑥1 − 𝑥2 − 𝑡𝛼/2 𝑠𝑝 + < 𝜇1 − 𝜇2 < 𝑥1 − 𝑥2 + 𝑡𝛼/2 𝑠𝑝 +
𝑛1 𝑛2 𝑛1 𝑛2

Where 𝑠𝑝 is the pooled estimate of the population standard

deviation and 𝑡𝛼/2 is the t value with 𝑣 = 𝑛1 + 𝑛2 − 2 degrees of
freedom, leaving an area of α/2 to the right.
5
Pooled estimate, 𝑠𝑝2 , of the common variance, 𝜎 2 , is

2 2
2
𝑛1 − 1 𝑠1 + 𝑛2 − 1 𝑠2
𝑠𝑝 =
𝑛1 + 𝑛2 − 2

6
 A course in mathematics is taught to 12 students by the
conventional classroom procedure. A second group of
10 students was given the same course by means of
programmed materials. At the end of the semester, the
same examination was given to each group. The 12
students meeting in the classroom made an average
grade of 85 with a standard deviation of 4, while the 10
students using programmed materials made an average
grade of 81 with a standard deviation of 5. Find a 90%
confidence interval for the difference between the
population means, assuming the populations are
approximately normally distributed with equal variances.
7
Conventional classroom procedure:
𝑛1 = 12
𝑥1 = 85
𝑠1 = 4
Programmed materials
𝑛2 = 10
𝑥2 = 81
𝑠2 = 5
Find a 90% confidence interval.
8
 2 2
2
𝑛1 − 1 𝑠1 + 𝑛2 − 1 𝑠2
𝑠𝑝 =
𝑛1 + 𝑛2 − 2
12−1 42 +(10−1)52
𝑠𝑝2 =
12+10−2

𝑠𝑝2 = 20.05

𝑠𝑝 = 20.05
𝑠𝑝 = 4.478
9
@90% confidence interval:
𝑣 = 𝑛1 + 𝑛2 − 2
𝑣 = 12 + 10 − 2
𝑣 = 20

𝑡𝛼/2 = 𝑡0.10/2 = 𝑡0.05

From Table A.5 𝑡0.05 for 𝑣 = 20 = 1.725

10
 1 1 1 1
𝑥1 − 𝑥2 − 𝑡𝛼/2 𝑠𝑝 + < 𝜇1 − 𝜇2 < 𝑥1 − 𝑥2 + 𝑡𝛼/2 𝑠𝑝 +
𝑛1 𝑛2 𝑛1 𝑛2

1 1 1 1
85 − 81 − 1.725 4.478 + < 𝜇1 − 𝜇2 < 85 − 81 + 1.725(4.478) +
12 10 12 10

0.693 < 𝜇1 − 𝜇2 < 7.307

The fact that both confidence limits are positive indicates that the classroom method of
learning this particular course is superior to the method using programmed materials.
11
If 𝑥1 and 𝑠12 , and 𝑥2 and 𝑠22 , are the means and variances of
small independent samples of size 𝑛1 and 𝑛2 , respectively,
from approximate normal distributions with unknown and
unequal variances, an approximate (1-α)100% confidence
interval for 𝜇1 − 𝜇2 is given by

𝑠12 𝑠22 𝑠12 𝑠22

𝑥1 − 𝑥2 − 𝑡𝛼/2 + < 𝜇1 − 𝜇2 < 𝑥1 − 𝑥2 + 𝑡𝛼/2 +
𝑛1 𝑛2 𝑛1 𝑛2

12
 Where 𝑡𝛼/2 is the 𝑡 value with

2 2 2
𝑠1 𝑠2
+
𝑛1 𝑛2
𝑣= 2 2 2 2 (round-off)
𝑠1 𝑠2
𝑛1 𝑛2
+
𝑛1 −1 𝑛2 −1

Degrees of freedom, leaving an area of α/2 to the right.

13
 Records for the past 15 years have shown the
average rainfall in a certain region of the country for
the month of May to be 4.93 centimeters, with a
standard deviation of 1.14 centimeters. A second
region of the country has had an average rainfall in
May of 2.64 centimeters with a standard deviation of
0.66 centimeters during the past 10 years. Find a 95%
confidence interval for the difference of the true
average rainfalls in these two regions, assuming that
the observations come from normal populations with
different variances.
14
𝑛1 = 15
𝑥1 = 4.93 centimeters
𝑠1 = 1.14 centimeters

𝑛2 = 10
𝑥2 = 2.64 centimeters
𝑠2 = 0.66 centimeters

Find a 95% confidence interval.

15
@95% confidence interval: 𝑡𝛼/2 = 𝑡0.05/2 = 𝑡0.025

2 2 2
𝑠1 𝑠2 2
+ 1.142 0.662
𝑛1 𝑛2 +
= 22.693 ≈ 23
15 10
𝑣= 2 2 2
𝑣= 2 2
𝑠2
1 𝑠2 1.14 2 0.66 2
𝑛1 𝑛2 15 10
+ +
𝑛1 −1 𝑛2 −1 15−1 10−1

From Table A.5 𝑡0.025 for 𝑣 = 23 = 2.069

16
 𝑠12 𝑠22 𝑠12 𝑠22
𝑥1 − 𝑥2 − 𝑡𝛼/2 + < 𝜇1 − 𝜇2 < 𝑥1 − 𝑥2 + 𝑡𝛼/2 +
𝑛1 𝑛2 𝑛1 𝑛2

1.142 0.662 1.142 0.662

4.93 − 2.64 − 2.069 + < 𝜇1 − 𝜇2 < 4.93 − 2.64 + 2.069 +
15 10 15 10

1.543 < 𝜇1 − 𝜇2 < 3.037

17
 If 𝑑 and 𝑠𝑑 are the mean and standard deviation of the
differences of 𝑛 random pairs of measurements, a (1-α)100%
confidence interval for 𝜇𝐷 = 𝜇1 − 𝜇2 is

𝑠𝑑 𝑠𝑑
𝑑 − 𝑡𝛼/2 < 𝜇𝐷 < 𝑑 + 𝑡𝛼/2
𝑛 𝑛

Where 𝑡𝛼/2 is the 𝑡 value with 𝑣 = 𝑛 − 1 degrees of freedom,

leaving an area of α/2 to the right.

18
The variance, 𝑠𝑑2 , of the sample differences is

𝑛 𝑑𝑖2 − 𝑑𝑖 2
𝑠𝑑2 =
𝑛(𝑛 − 1)

19
Twenty college freshmen were divided into 10 pairs,
each member of the pair having approximately the
same IQ. One of each pair was selected at random
and assigned to a mathematics section using
programmed materials only. The other member of
each pair was assigned to a section in which the
professor lectured. At the end of the semester each
group was given the same examination and the
following results were recorded.

20
 Pair Programmed Materials Lectures
1 76 81
2 60 52
3 85 87
4 58 70
5 91 86
6 75 77
7 82 90
8 64 63
9 79 85
10 88 83

Find a 98% confidence interval for the true difference in

the two learning procedures. 21
@98% confidence interval: 𝑡𝛼/2 = 𝑡0.02/2 = 𝑡0.01
𝑣 = 𝑛 − 1 = 10 − 1 = 9

From Table A.5 𝑡0.01 for 𝑣 = 9 = 2.821

22
 Programmed
Pair Lectures 𝑑 𝒅𝟐
Materials
1 76 81 -5 25
2 60 52 8 64
3 85 87 -2 4
4 58 70 -12 144
5 91 86 5 25
6 75 77 -2 4
7 82 90 -8 64
8 64 63 1 1
9 79 85 -6 36
10 88 83 5 25
𝑑𝑖 = -16 𝑑𝑖2 = 392
𝑑 = -1.6 23
 2 2
2 𝑛 𝑑 𝑖 − 𝑑 𝑖
𝑠𝑑 =
𝑛(𝑛 − 1)
2
2 10 392 − −16
𝑠𝑑 =
10(10 − 1)

𝑠𝑑 = 40.711
2

𝑠𝑑 = 6.381

24
 𝑠𝑑 𝑠𝑑
𝑑 − 𝑡𝛼/2 < 𝜇𝐷 < 𝑑 + 𝑡𝛼/2
𝑛 𝑛
(6.381) (6.381)
−1.6 − 2.821 < 𝜇𝐷 < −1.6 + (2.821)
10 10

−7.292 < 𝜇𝐷 < 4.092

Since this interval allows for the possibility of 𝜇𝐷 being equal to zero, we are
unable to state that one method of instruction is better than the other, even
though this particular sample of differences shows the lecture procedure to be
superior.
25
If 𝑝 is the proportion of successes in a random sample of size 𝑛,
and 𝑞 = 1 − 𝑝, an approximate (1-α)100% confidence interval
for the binomial parameter 𝑝 is given by

𝑝𝑞 𝑝𝑞
𝑝 − 𝑧𝛼/2 < 𝑝 < 𝑝 + 𝑧𝛼/2
𝑛 𝑛
Where 𝑧𝛼/2 is the z value leaving an area of α/2 to the right.

26
In a random sample of 500 people eating lunch at a
hospital cafeteria on various Fridays, it was found that
x = 160 preferred seafood. Find a 95% confidence
interval for the actual proportion of people who eat
seafood on Fridays at this cafeteria.

27
@95% confidence interval 𝑧𝛼/2 = 𝑧0.025 . = 1.96.

𝑥 160
𝑝= = = 0.32
𝑛 500

𝑞 = 1 − 𝑝 = 1 − 0.32 = 0.68

28
 𝑝𝑞 𝑝𝑞
𝑝 − 𝑧𝛼/2 < 𝑝 < 𝑝 + 𝑧𝛼/2
𝑛 𝑛
0.32 0.68 (0.32)(0.68)
0.32 − 1.96 < 𝑝 < 0.32 + (1.96)
500 500

0.279 < 𝑝 < 0.361

29
If 𝑝 is used as an estimate of 𝑝, we can be
(1-𝛼)100% confident that the error will not
exceed a specified amount e when the
sample size is

2
𝑧𝛼/2 𝑝𝑞
𝑛=
𝑒2
30
 The previous theorem is somewhat
misleading in that we must use 𝑝 to
determine the sample size 𝑛, but 𝑝 is
computed from the sample.

31
 If a crude estimate of 𝑝 can be made
without taking a sample, we could use this
value for 𝑝 and then determine 𝑛. Lacking
such an estimate, we could take a
preliminary sample of size 𝑛 ≥ 30 to provide
an estimate of 𝑝.

32
Then, using the formula for computing the
sample size for estimating 𝑝, we could
determine approximately how many
observations are needed to provide the
desired degree of accuracy. Once again,
all fractional values of 𝑛 are rounded up to
the next whole number.

33
How large a sample is required in Example
9.8 if we want to be 95% confident that our
estimate 𝑝 is within 0.02?

34
 2
𝑧𝛼/2 𝑝𝑞
𝑛= 2
𝑒
1.96 2 (0.32)(0.68)
𝑛=
(0.02)2

𝑛 = 2089.83 ≈ 2090

35
Occasionally, it will be impractical to obtain
an estimate of 𝑝 to be used in determining
the sample size for a specified degree of
confidence.

36
If 𝑝 is used as an estimate of 𝑝, we can be at
least (1-𝛼)100% confident that the error will
not exceed a specified amount e when the
sample size is

2
𝑧𝛼/2
𝑛=
4𝑒 2
37
How large a sample is required in Example
9.8 if we want to be 95% confident that our
estimate 𝑝 is within 0.02?

38
Unlike Example 9.9, we now assume that no
preliminary sample has been taken to
provide an estimate of 𝑝.

39
 2
𝑧𝛼/2
𝑛=
4𝑒 2 2
1.96
𝑛=
4(0.02)2

𝑛 = 2401

40
Comparing the results of Examples 9.9 and
9.10, we see that information concerning 𝑝,
provided by a preliminary sample, or
perhaps from past experience, enables
us to choose a smaller sample while
maintaining our required degree of
accuracy.
41
If 𝑝1 and 𝑝2 are the proportion of successes in random samples
of size 𝑛1 and 𝑛2 , respectively, 𝑞1 = 1 − 𝑝1 and 𝑞2 = 1 − 𝑝2 , an
approximate (1-α)100% confidence interval for the difference
of two binomial parameters, 𝑝1 − 𝑝2 , is given by

𝑝1 𝑞1 𝑝2 𝑞2 𝑝1 𝑞1 𝑝2 𝑞2
𝑝1 − 𝑝2 − 𝑧𝛼/2 + < 𝑝1 − 𝑝2 < 𝑝1 − 𝑝2 + 𝑧𝛼/2 +
𝑛1 𝑛2 𝑛1 𝑛2
Where 𝑧𝛼/2 is the z value leaving an area of 𝛼/2 to the right.

42
 A poll is taken among the residents of a city and
the surrounding county to determine the feasibility
of a proposal to construct a civic center. If 2400
of 5000 city residents favor the proposal and 1200
of 2000 county residents favor it, find a 90%
confidence interval for the true difference in the
fractions favoring the proposal to construct the
civic center.

43
 2400
𝑝1 = = 0.48
5000
𝑞1 = 1 − 𝑝1 = 1 − 0.48 = 0.52

1200
𝑝2 = = 0.60
2000
𝑞2 = 1 − 𝑝2 = 1 − 0.60 = 0.40

@90% confidence interval 𝑧𝛼/2 = 𝑧0.05 . = 1.645.

44
 𝑝1 𝑞1 𝑝2 𝑞2 𝑝1 𝑞1 𝑝2 𝑞2
𝑝1 − 𝑝2 − 𝑧𝛼/2 + < 𝑝1 − 𝑝2 < 𝑝1 − 𝑝2 + 𝑧𝛼/2 +
𝑛1 𝑛2 𝑛1 𝑛2
0.48 0.52 0.60 0.40
0.48 − 0.60 − 1.645 + < 𝑝1 − 𝑝2
5000 2000
(0.48)(0.52) (0.60)(0.40)
< 0.48 − 0.60 + 1.645 +
5000 2000

−0.1414 < 𝑝1 − 𝑝2 < −0.0986

45
Since both end points of the interval are negative,
we can also conclude that the proportion of
county residents favoring the proposal is greater
than the proportion of city residents favoring the
proposal.

46