Statistical Inference &

Quantitative Research

Topic:
Confidence Interval Estimation:
Further Topics

Arshad Ali Bhatti, PhD

Fall_2023
 Lecture Goals
After completing this lecture, you should be able to:
◼ Form confidence intervals for the difference between
two independent population means (standard deviations
known or unknown)
◼ Compute confidence interval limits for the difference
between two independent population proportions

Dr. Arshad Ali Bhatti/ SOE, IIIE Ch. 8-2

 Estimation: Additional Topics
Lecture Topics

Confidence Intervals

Population Population
Means, Means, Population
Dependent Independent Proportions
Samples Samples
Examples:
Same group Group 1 vs. Proportion 1 vs.
before vs. after independent Proportion 2
treatment Group 2

Dr. Arshad Ali Bhatti/ SOE, IIIE Ch. 8-3

8.2
Difference Between Two Means:
Independent Samples
Confidence Interval Estimation of the
Population means, Difference Between Two Normal
independent Population Means: Independent Samples
samples

Goal: Form a confidence interval

for the difference between two
population means, μ1 – μ2

Dr. Arshad Ali Bhatti/ SOE, IIIE Ch. 8-4

 Difference Between Two Means:
Independent Samples
(continued)

Population means, Goal: Form a confidence interval

independent for the difference between two
samples population means, μ1 – μ2
◼ Different data sources
◼ Unrelated

◼ Independent

◼ Sample selected from one population has no effect on the

sample selected from the other population
◼ The point estimate is the difference between the two
sample means:
x1 – x2
Dr. Arshad Ali Bhatti/ SOE, IIIE Ch. 8-5
 Difference Between Two Means:
Independent Samples
(continued)

Population means,
independent
samples

σ12 and σ22 known Confidence interval uses z/2

σ12 and σ22 unknown

σ12 and σ22

assumed equal Confidence interval uses a value
from the Student’s t distribution
σ12 and σ22
assumed unequal
Dr. Arshad Ali Bhatti/ SOE, IIIE Ch. 8-6
 σ12 and σ22 Known

Population means, Assumptions:

independent
samples ▪ Samples are randomly and
independently drawn
σ12 and σ22 known
* ▪ both population distributions
σ12 and σ22 unknown are normal

▪ Population variances are

known

Dr. Arshad Ali Bhatti/ SOE, IIIE Ch. 8-7

 σ12 and σ22 Known
(continued)

When σx and σy are known and

Population means,
both populations are normal, the
independent
samples variance of x1 – x2 is
σ1 2 σ2 2
σ2x −x = +
σ12 and σ22 known
* n1 n2
1 2

…and the random variable

σ12 and σ22 unknown
(𝑋ത1 −𝑋ത2 ) − (μ1 − μ2 )
Z=
σ12 σ22
+
n1 n2

has a standard normal distribution

Dr. Arshad Ali Bhatti/ SOE, IIIE Ch. 8-8
 Confidence Interval,
σx2 and σy2 Known

Population means,
independent
samples

σ12 and σ22 known

* The confidence interval for
μ1 – μ2 is:
σ12 and σ22 unknown
σ12 σ12
(𝑋ത1 −𝑋ത2 ) ± zα/2 +
n1 n2

Dr. Arshad Ali Bhatti/ SOE, IIIE Ch. 8-9

 Example 8.3 Comparison of GPAs

Dr. Arshad Ali Bhatti/ SOE, IIIE Ch. 8-10

Dr. Arshad Ali Bhatti/ SOE, IIIE Ch. 8-11
 σx2 and σy2 Unknown,
Assumed Equal

Population means, Assumptions:

independent
▪ Samples are randomly and
samples
independently drawn

σ12 and σ22 known ▪ Populations are normally

distributed
σ12 and σ22 unknown
▪ Population variances are
σ12 and σ22
assumed equal * unknown but assumed equal

σx2 and σy2

assumed unequal
Dr. Arshad Ali Bhatti/ SOE, IIIE Ch. 8-12
 σx2 and σy2 Unknown,
Assumed Equal
(continued)

Population means, Forming interval

independent estimates:
samples
▪ The population variances
σ12 and σ22 known are assumed equal, so use
the two sample standard
deviations and pool them to
σ12 and σ22 unknown
estimate σ
σ12 and σ22
assumed equal * ▪ use a t value with
(n1 + n2 – 2) degrees of
σx2 and σy2 freedom
assumed unequal
Dr. Arshad Ali Bhatti/ SOE, IIIE Ch. 8-13
 σx2 and σy2 Unknown,
Assumed Equal
(continued)

Population means,
independent
samples
The pooled variance is
σ12 and σ22 known

σ12 and σ22 unknown (n − 1)s 2

+ (n − 1)s 2
1 1 2 2
sp2 =
n1 + n2 − 2
σ12 and σ22
assumed equal *
σ12 and σ22
assumed unequal
Dr. Arshad Ali Bhatti/ SOE, IIIE Ch. 8-14
 Confidence Interval,
σx2 and σy2 Unknown, Equal

σ12 and σ22 unknown

σ12 and σ22

assumed equal * The confidence interval for
μ1 – μ2 is:
σ12 and σ22
assumed unequal
sp2 sp2
(𝑋ത1 −𝑋ത2 ) ± t α,n +
2 1 +n2 −2 n1 n2

(n1 − 1)s12 + (n2 − 1)s22

sp2 =
Where n1 + n2 − 2

Dr. Arshad Ali Bhatti/ SOE, IIIE Ch. 8-15

 Pooled Variance Example

You are testing two computer processors for speed.

Form a confidence interval for the difference in CPU
speed. You collect the following speed data (in Mhz):

CPU1 CPU2
Number Tested 17 14
Sample mean 3004 2538
Sample std dev 74 56

Assume both populations are

normal with equal variances,
and use 95% confidence

Dr. Arshad Ali Bhatti/ SOE, IIIE Ch. 8-16

 Calculating the Pooled Variance
The pooled variance is:
2 2
2
n1 − 1 S1 + n2 − 1 S2 17 − 1 742 + 14 − 1 562
Sp = = = 4427.03
(n1 − 1) + (n2 − 1) (17−1) + (14 − 1)

The t value for a 95% confidence interval is:

t n1 +n2 −2 , α/2 = t 29 , 0.025 = 2.045

Dr. Arshad Ali Bhatti/ SOE, IIIE Ch. 8-17

 Calculating the Confidence Limits

◼ The 95% confidence interval is

sp2 sp2
(𝑋ത1 −𝑋ത2 ) ± t α,n +
2 1 +n2 −2 n x ny

4427.03 4427.03
(3004 − 2538)  (2.054) +
17 14

416.69 < μ1 − μ2 < 515.31

We are 95% confident that the mean difference in

CPU speed is between 416.69 and 515.31 Mhz.
Dr. Arshad Ali Bhatti/ SOE, IIIE Ch. 8-18
 Give it a try?

◼ Two separate groups of subjects were tested. The

experimental group (X1) had 5 subjects; the control group
(X2) had 4 subjects. The data are given below; the scores are
assumed to be normally distributed.

◼ X1: 12, 13, 16, 14, 15

◼ X2: 10, 13, 14, 12

◼ Find the 95% confidence interval for the difference of the

means.

Dr. Arshad Ali Bhatti/ SOE, IIIE Ch. 8-19

8.3
Two Population Proportions
Confidence Interval Estimation of the
Difference Between Two Population
Population Proportions (Large Samples)
proportions

Goal: Form a confidence interval for

the difference between two
population proportions, P1 – P2

Dr. Arshad Ali Bhatti/ SOE, IIIE Ch. 8-20

8.3
Two Population Proportions

Goal: Form a confidence interval for

Population the difference between two
proportions population proportions, P1 – P2

Assumptions:
Both sample sizes are large (generally at
least 40 observations in each sample)

The point estimate for pො 1 − pො 2

the difference is

Dr. Arshad Ali Bhatti/ SOE, IIIE Ch. 8-21

 Two Population Proportions
(continued)

◼ The random variable

Population
proportions
(ොp1 − pො 2 ) − (p1 − p2 )
Z=
pො 1 (1 − pො 1 ) pො 1 (1 − pො 2 )
+
n1 n2

is approximately normally distributed

Dr. Arshad Ali Bhatti/ SOE, IIIE Ch. 8-22

 Confidence Interval for
Two Population Proportions

Population The confidence limits for

proportions
P1 – P2 are:

pො 1 (1 − pො 1 ) pො 2 (1 − pො 2 )
(ොp1 − pො 2 ) ± z𝛼/2 +
n1 n2

Dr. Arshad Ali Bhatti/ SOE, IIIE Ch. 8-23

 Example:
Two Population Proportions

Form a 90% confidence interval for the

difference between the proportion of
men and the proportion of women who
have college degrees.

◼ In a random sample, 26 of 50 men and

28 of 40 women had an earned college
degree

Dr. Arshad Ali Bhatti/ SOE, IIIE Ch. 8-24

 Example:
Two Population Proportions
(continued)
26
Men: pො 1 = = 0.52
50

28
Women: pො 2 = = 0.70
40

pො 1 (1 − pො 1 ) pො 2 (1 − pො 2 ) 0.52(0.48) 0.70(0.30)
+ = + = 0.1012
n1 n2 50 40

For 90% confidence, Z/2 = 1.645

Dr. Arshad Ali Bhatti/ SOE, IIIE Ch. 8-25

 Example:
Two Population Proportions
(continued)

The confidence limits are:

pො 1 (1 − pො 1 ) pො 2 (1 − pො 2 )
(ොp1 − pො 2 ) ± Zα/2 +
n1 n2
= (0.52 − 0.70) ± 1.645(0.1012)

so the confidence interval is

-0.3465 < P1 – P2 < -0.0135

Since this interval does not contain zero we are 90% confident that the
two proportions are not equal
Dr. Arshad Ali Bhatti/ SOE, IIIE Ch. 8-26
 Give it a try

◼ In a random sample of 600 adults and 400 teenagers

who watched a certain T.V program, 300 adults and
100 teenagers indicated that they liked it. Construct
99% confidence interval of the difference in
proportions of all adults and all teenagers who
watched the program and liked it.

Dr. Arshad Ali Bhatti/ SOE, IIIE Ch. 8-27

◼ n1=600, x1=300 ; p1^=300/600=0.5, q1^=0.5
◼ n2=400, x2=100 ; p2^=100/400=0.25,
q2^=0.75
◼ SE(p1^-p2^)=sqrt{(0.5)(0.5)/600
+(0.25)(0.75)/400}=0.0297
◼ Z(0.99)=+-2.58
◼ (0.5-0.25)+-2.58(0.0297)
◼ 0.25+-0.076
◼ 0.173___ 0.326
Dr. Arshad Ali Bhatti/ SOE, IIIE Ch. 8-28
 Lecture Summary
◼ Compared two independent samples
◼ Formed confidence intervals for the difference
between two means, population variance known,
using z
◼ Formed confidence intervals for the differences
between two means, population variance unknown,
using t
◼ Formed confidence intervals for the differences
between two population proportions

Dr. Arshad Ali Bhatti/ SOE, IIIE Ch. 8-29

◼ Thanks

Dr. Arshad Ali Bhatti/ SOE, IIIE Ch. 8-30