Statistics for Business and Economics

Chapter 3. Sample Statistics - Point and Interval Estimations

(Reference: Chapter 7, 8, 10)

Dr. Ha Van Hieu

Faculty of Economic Mathematics, University of Economics and Law

Outline

Point Estimation

Interval Estimation - one population

Interval Estimation - two populations

Random Sample

▶ A simple random sample of size n from a population of size N is a sample selected

such that: each possible sample of size n has the same probability of being
selected.
▶ A random sample of size n from an infinite population is a sample selected such
that: (1) Each element selected comes from the same population, and (2) Each
element is selected independently.

Sample statistic
To estimate the value of a population parameter, we calculate a corresponding
characteristic of the sample, referred to as a sample statistic.
Exercise: Method 1 (p. 309).
Point Estimate

Data: Status of the Salary and Training Program of 2500 EAI Managers.

We refer to the sample mean x as the point estimator of the population mean µ.
Sampling Distributions

If we consider the process of selecting a simple random sample as an experiment, the

sample mean x is a random variable. The probability distribution of x is called the
sampling distribution of x.
The sampling distribution of x
The histogram of x

The sampling distribution of x is the

probability distribution of all possible
values of the sample mean x. Note that in
practice, we select only one simple
random sample from the population.
▶ Expected value:

E(x) = µ

▶ Sample size: n, population size: N.

Finite
√ population Infinite population
( )
Standard Deviation: N−n σ σ
σx = √ σx = √
N−1 n n
In the case where the population size is large, we may use the later formulate to
calculate the standard deviation of x.
Central Limit theorem

Central Limit Theorem

In selecting random samples of size n from a population, the sampling distribution of
the sample mean x can be approximated by a normal distribution as the sample size
becomes large.

( ( )2 )
σ
x ∼ N µ, √
n

Exercises: Methods 19,

21 (p.324).
Sampling Distribution of p

The sampling distribution of p is the probability distribution of all possible values of

the sample proportion p.
▶ The expected value of p is equal to the population proportion:

E(p) = p

▶ The standard deviation

Finite population
√ Infinite (or √
large) population
√
N − n p(1 − p) p(1 − p)
σp = σp =
N−1 n n
▶ The sampling distribution of p can be approximated by a normal distribution
whenever np ≥ 5 and n(1 − p) ≥ 5.
Exercises: Methods 31, 32, 33, 34 (p.330).
Other Sampling methods

▶ In stratified random sampling (chọn mẫu phân tổ), the elements in the population
are first divided into separate groups (for example, department, age, etc.). After
that, a simple random sample is taken from each group.
Other Sampling methods

▶ In cluster sampling (chọn mẫu cả khối), the elements in the population are first
divided into separate groups called clusters. Each element of the population
belongs to one and only one cluster. A simple random sample of the clusters is
then taken. All elements within each sampled cluster form the sample.
Other Sampling methods
▶ Systematic Sampling (chọn mẫu ngẫu nhiên hệ thống): every k-th element is
chosen from the population.

▶ Convenience Sampling (chọn mẫu thuận tiện): a nonprobability sampling

technique. The sample is identified primarily by convenience.
▶ Judgment Sampling (chọn mẫu phán đoán): the person most knowledgeable on
the subject of the study selects elements of the population that he or she feels are
the most representative of the population.
Interval Estimation

▶ The general form of an interval estimate of a population mean is x ± error.

▶ The general form of an interval estimate of a population proportion is p ± error.
Case 1.1 Estimate µ when σ is known.
Case 1.2 Estimate µ when σ is unknown.
Case 2 Estimate p.
Case 3 Estimate µ1 − µ2 .
Case 4 Estimate p1 − p2 .
Population mean: σ known

As we know,
x ∼ N(µ, σx2 ),
√
where σx = σ/ n.
Example: Lloyd Data: infinite population with mean µ needed to estimate and
σ = 20, a sample of size n = 100 with x = 82. This is a point estimate of µ.

Remark: 95% of all x values

obtained using a sample size of
n = 100 will be within ±3.92 of
the population mean µ.
Example - Lloyd Data
Population: σ = 20. Sample: n = 100, x = 82. Need to estimate µ in the form
x ± error. Let’s consider the case where there were three different samples with values
x1 , x2 , and x3 .
Suppose we choose the error equal to 3.92.
Using x ± 3.92 to construct the interval
estimate, we obtain 82 ± 3.92. Because
95% of all the intervals constructed using
x ± 3.92 will contain the population mean,
we say that we are 95% confident that the
interval 78.08 to 85.92 includes the
population mean µ. 95% is referred to as
the confidence level, 0.95 is referred to as
the confidence coefficient, and
(78.08, 85.92) is called the 95% confidence
interval.
Interval Estimate of population mean: σ known
Therefore,
( )
1 − confidence coefficient σ
ϵ = f−1 √
2 n

The interval estimate:

σ
x ± zα/2 √
n

where
0.95 = 1 − 2f(1.96) ▶ 1 − α is the confidence coefficient.
( )
1 − 0.95 ▶ f is the density function of the
1.96 = f−1
2 standard normal distribution.
σ ▶ zα/2 is the z-value. That is
σx = √
n f(zα/2 ) = α/2
error: ϵ = 1.96σx
Common Z values

Exercises: Methods 1, 2, 3, 4 (pp.352-353)

Populations mean: σ unknown

When σ is unknown, we use t-distribution instead of normal distribution.

The interval estimate:
s
x ± tα/2 √
n

▶ s is the sample standard deviation,

√
∑
(xi − x)2
s=
n−1

▶ 1 − α is the confidence coefficient,

Exercises: Methods 11, 12, 13, 14 (p. ▶ tα/2 is the t value with n − 1 degrees
361).
of freedom.
Sample Size

The interval estimate is or

σ ()2
x ± zα/2 √ zα/2 σ 2
n n=
E2
Therefore, the sample size n needed to
provide any desired margin of error E is: σ known σ unknown
( )2 ( )2
σ zα/2 σ 2 tα/2 s2
E = zα/2 √ n= n=
n E2 E2

Exercises: Methods 23, 24 (p.365).

Population Proportion
▶ The general form of an interval ▶ The interval estimate is
estimate of a population proportion p √
is p(1 − p)
p ± zα/2
p ± error n

▶ The sampling distribution of p can be ▶ The sample size for an interval

approximated by a normal estimate with a desired margin of
distribution whenever np ≥ 5 and error E is
n(1 − p) ≥ 5:
(zα/2 )2 p∗ (1 − p∗ )
( ) n=
p ∼ N p, σp2 E2

√ where p∗ is a planning value for p.

p(1 − p)
where σp = . ▶ Exercises: Methods 31, 32, 33, 34
n
(p.370).
Interval estimation of µ1 − µ2

▶ µ1 , and σ1 are the mean, and the standard deviation of population 1, respectively,
▶ µ2 and σ2 are the mean, and the standard deviation of population 2, respectively,
▶ x is the sample mean of a simple random sample from population 1,
▶ y is the sample mean of a simple random sample from population 2.
▶ The point estimator of the difference between two populations means:

x − y.

▶ Interval estimator of µ1 − µ2 :

x − y ± error . (1)
Interval Estimator of µ1 − µ2 : Independent Samples
▶ The interval estimator of µ1 − µ2 when σ1 and σ2 are known.
√
σ12 σ22
x − y ± zα/2 + (2)
n1 n2
▶ Interval estimator of µ1 − µ2 when σ1 and σ2 are unknown.
√
s21 s2
x − y ± tα/2 + 2 (3)
n1 n2

where the degrees of freedom is And s1 and s2 are variance of sample 1

( )2 an 2, respectively.
s21 s22 √
+
n1 n2 1 ∑
df = ( 2 )2 ( 2 )2 s1 = (xi − x)2
1 s1 1 s2 n1 − 1
+
n1 − 1 n1 n2 − 1 n2
Interval Estimator of µ1 − µ2 : Matched Samples

In the matched samples, two methods are tested under similar conditions. It results in
a smaller sampling error. In this case, two samples have the same size n.
Interval Estimator of µ1 − µ2 : Matched samples

▶ The interval estimator of µ1 − µ2 is

Sd
x − y ± tα/2 √
n

where the degree of freedom is df = n − 1, and

√
∑
(xi − yi − (x − y))2
Sd =
n−1
Interval Estimation of p1 − p2
▶ p1 and p2 are the proportions of population 1, and population 2, respectively.
▶ p1 and p2 are the proportions of sample 1 (selected from population 1) and
sample 2 (selected from population 2), respectively,
▶ The point estimator of the difference between two population proportions:

p1 − p2 .

▶ Interval estimator of p1 − p2 :

p1 − p2 ± error . (4)

▶ The mean and the standard deviation of the sampling distribution p1 − p2 is

E(p1 − p2 ) = p1 − p2
√
p1 (1 − p1 ) p2 (1 − p2 )
σp1 −p2 = +
n1 n2
The interval estimate of p1 − p2

▶ The interval estimate of the difference between two population proportions is

√
p1 (1 − p1 ) p2 (1 − p2 )
p1 − p2 ± zα/2 +
n1 n2

Exercises: Methods 19abc, and 20 (p.463), Methods 28 (p.469)

Teamwork for bonus point - Chapter 7

▶ Team 1: Applications 3 - 10 (p. 309 - 310).

▶ Team 2: Applications 13 - 17 (p. 313 - 314).
▶ Team 3: Applications 22 - 30 (p. 325 - 326),
▶ Team 4: Applications 35 - 41 (p. 330 - 332),
▶ Team 5: Supplementary Exercises 42 - 48 (p. 339 - 341),
▶ Team 6: Supplementary Exercises 49 - 54 (p. 339 - 341),
▶ All teams: the case problem (p. 342).
Teamwork for bonus point - Chapter 8

▶ Team 1: Applications 5 - 10 (p. 353 - 354).

▶ Team 2: Applications 15 - 22 (p. 362 - 363).
▶ Team 3: Applications 25 - 30 (p. 365 - 366),
▶ Team 4: Applications 35 - 43 (p. 370 - 372),
▶ Team 5: Supplementary Exercises 44 - 52 (p. 374 - 376),
▶ Team 6: Supplementary Exercises 53 - 60 (p. 374 - 376),
▶ All teams: one of the three case problems (p. 377-380).
Teamwork for bonus point - Chapter 10

▶ Team 1: Applications 4 - 8 (p. 450 - 451).

▶ Team 2: Applications 12 - 18 (p. 458 - 460).
▶ Team 3: Applications 21 - 27 (p. 463 - 465),
▶ Team 4: Applications 30 - 37 (p. 470 - 472),
▶ Team 5: Supplementary Exercises 38 - 43 (p. 474 - 476),
▶ Team 6: Supplementary Exercises 47 - 47 (p. 474 - 476),
Thank you for your attention!