POSTS AND TELECOMMUNICATIONS INSTITUTE OF

TECHNOLOGY
——————–o0o——————–

Chapter 5: Estimation

PROBABILITY & STATISTICS FOR ECONOMICS

Department of Mathematics, Faculty of Fundamental

Science 1

Hanoi - 2026

1 / 32 Chapter 5: Estimation
Chapter 5: Estimation

1 5.1 Point Estimation

2 5.2 Interval Estimation

2 / 32 Chapter 5: Estimation
Chapter 5: Estimation

1 5.1 Point Estimation

2 5.2 Interval Estimation

3 / 32 Chapter 5: Estimation
Point Estimators
Let (X1 , X2 , . . . , Xn ) be a random sample of size n taken from a
population with an unknown parameter θ. If a statistic
Θ̂ = h(X1 , X2 , . . . , Xn ) is used instead of θ, then Θ̂ is called a
point estimator of θ.
After the sample has been selected, Θ̂ takes on a particular
numerical value θ̂ called the point estimate of θ.

4 / 32 Chapter 5: Estimation
Example 1. Suppose that the random variable X is normally
distributed with an unknown mean µ.
The sample mean X is a point estimator of µ.
After the sample has been selected, the numerical value x is the
point estimate of µ. Thus, if x1 = 25, x2 = 30, x3 = 29, and
x4 = 31, the point estimate of µ is
25 + 30 + 29 + 31
x= = 28.75
4

5 / 32 Chapter 5: Estimation
Unbiased Estimators
A statistic Θ̂ is called an unbiased estimator of the parameter θ if

E(Θ̂) = θ.

Otherwise, Θ̂ is called a biased estimator of θ.

Example 2. Suppose that X is a random variable with mean µ and

variance σ 2 . Let (X1 , X2 , . . . , Xn ) be a random sample of size n from
the population represented by X. Then,
the sample mean X is an unbiased estimator of µ,
the sample variance S 2 is an unbiased estimator of σ 2 .

6 / 32 Chapter 5: Estimation
Consistent Estimators
A statistic Θ̂ is called a consistent estimator of the parameter θ if

lim P (|Θ̂ − θ| < ε) = 1, ∀ε > 0.

n→∞

7 / 32 Chapter 5: Estimation
Efficient Estimators
If we consider all unbiased estimators of a parameter θ, the one with
the smallest variance is called an efficient estimator of θ.

We often need to estimate:

1) The mean µ of a single population.
2) The variance σ 2 (or standard deviation ) of a single population.
3) The proportion p of items in a population that belong to a class of
interest.

8 / 32 Chapter 5: Estimation
Reasonable point estimates of these parameters are as follow:
1) The best point estimate for µ is µ̂ = x.
2) The best point estimate for σ 2 is σ̂ 2 = s2 .
3) The best point estimate for p is p̂ = x/n, where x is the number of
items in a random sample of size n that belong to the class of
interest.

9 / 32 Chapter 5: Estimation
Chapter 5: Estimation

1 5.1 Point Estimation

2 5.2 Interval Estimation

10 / 32 Chapter 5: Estimation
5.2.1 Introduction

Suppose that L = L(X1 , X2 , . . . , Xn ), U = U (X1 , X2 , . . . , Xn ) are

two statistics from a random sample (X1 , X2 , . . . , Xn ), θ is a
population parameter, α ∈ (0, 1).
The interval [L, U ] is called a confidence interval for θ with a
confidence level 1 − α if

P (L ≤ θ ≤ U ) = 1 − α.

U –L is called the width of the confidence interval.

11 / 32 Chapter 5: Estimation
 If we have selected the sample:

X1 = x1 , X2 = x2 , . . . , Xn = xn

and computed L = ℓ, U = u, a 1 − α confidence interval for θ is

ℓ ≤ θ ≤ u.

ℓ: the lower confidence limit,

u: the upper confidence limit.

12 / 32 Chapter 5: Estimation
5.2.2 Confidence Interval for a Population Mean

Problem
Find a 1–α confidence interval for a population mean µ.

Let X1 , X2 , . . . , Xn be a random sample of size n taken from the

population.

We consider the problem in three cases:

1. The population has a normal distribution with variance σ 2 known.
2. The population has an arbitrary distribution, large-sample.
3. The population has a normal distribution with variance σ 2
unknown.

13 / 32 Chapter 5: Estimation
Case 1: The population has a normal distribution with
variance σ 2 known

Use the notation zα to represent the value of Z such that the area
to its right under the standard normal curve is α; that is,

P (Z > zα ) = α.

z0.05 = 1.645, z0.025 = 1.96, z0.01 = 2.33, z0.005 = 2.575.

In Excel, zα = [Link](1 − α).

14 / 32 Chapter 5: Estimation
If the population is normally distributed with mean µ and standard
deviation σ, then
X −µ
Z= √
σ/ n
is standard normally distributed.

15 / 32 Chapter 5: Estimation
  
X −µ
P −zα/2 ≤ √ ≤ zα/2 = 1 − α
σ/ n
 
σ σ
⇔ P X − zα/2 √ ≤ µ ≤ X + zα/2 √ =1−α
n n

16 / 32 Chapter 5: Estimation
Case 1: The population has a normal distribution with variance
σ 2 known
A 1–α confidence interval for µ is
σ σ
x − zα/2 √ ≤ µ ≤ x + zα/2 √ .
n n

Example 3. An economist is interested in studying the incomes of

consumers in a particular region. A random sample of 50 individuals
resulted in an average income of $15000. Assume that the income of
consumers has a normal distribution with a standard deviation of
$1000. Construct a 95% confidence interval for the mean income of
consumers in this region.

17 / 32 Chapter 5: Estimation
Interpreting a Confidence Interval

Figure 2.1: Repeated construction of a confidence interval for µ

If an infinite number of random samples are collected and a 1 − α

confidence interval for µ is computed from each sample, 100(1 − α)% of
these intervals will contain the true value of µ.

18 / 32 Chapter 5: Estimation
 Figure 2.2: Error in estimating µ with x

σ
|x − µ| ≤ zα/2 √ .
n
σ
B = zα/2 √ is called the bound on the error of estimation.
n

19 / 32 Chapter 5: Estimation
Determining the Sample Size
If x is used as an estimate of µ, we can be (1 − α) confident that the
error |x − µ| will not exceed a specified amount B when the sample size
is
 z σ 2
α/2
n= . (1)
B
If the right-hand side of Equation (1) is not an integer, it must be
rounded up.

20 / 32 Chapter 5: Estimation
Example 4. A medical statistician wants to estimate the average
weight loss of people who are on a new diet plan. In a preliminary
study, he guesses that the standard deviation of the population of
weight losses is about 10 pounds. How large a sample should he take to
estimate the mean weight loss to within 2 pounds, with 90%
confidence?

21 / 32 Chapter 5: Estimation
Case 2: The population has an arbitrary distribution,
large-sample (n > 30)
A 1–α confidence interval for µ is
s s
x − zα/2 √ ≤ µ ≤ x + zα/2 √ .
n n

Example 5. A random sample of 100 students from a large college

showed an average IQ score of 112 with a standard deviation of 10.
Find a 99% confidence interval for the mean IQ score of all students at
this college.

22 / 32 Chapter 5: Estimation
t Distribution
A random variable X is said to have a t distribution with k degrees of
freedom if its probability density function is
 
k+1
Γ − k + 1
x2

2 2
fX (x) = √   1+ , x ∈ R,
k k
kπΓ
2
Z +∞
where Γ(x) = tx−1 e−t dt.
0

23 / 32 Chapter 5: Estimation
Figure 2.3: Probability density functions of several t distributions, k is the
number of degrees of freedom

24 / 32 Chapter 5: Estimation
Note:
If the number of degrees of freedom is large enough, the t
distribution approximates the standard normal distribution.
If T has a t distribution with k degrees of freedom, then tα,k is the
value satisfying
P (T > tα,k ) = α.

Table 4 in Appendix B lists the values of tα,k .

In Excel, tα,k = [Link].2T(2α, k).

25 / 32 Chapter 5: Estimation
Theorem
If (X1 , X2 , . . . , Xn ) be a random sample from a normal distribution
with unknown mean µ and unknown variance σ 2 , then the random
variable
X − µ√
T = n
S
has a t distribution with n − 1 degrees of freedom.

26 / 32 Chapter 5: Estimation
Case 3: The population has a normal distribution with variance
σ 2 unknown
A 1–α confidence interval for µ is
s s
x − tα/2,n−1 √ ≤ µ ≤ x + tα/2,n−1 √ .
n n

Example 6. The amount that a sample of 15 customers spent on

lunch ($) at a fast-food restaurant are

7.42, 6.29, 5.83, 6.50, 8.34, 9.51, 7.10, 6.80

5.90, 4.89, 6.50, 5.52, 7.90, 8.30, 9.60

Assume that the population is normally distributed. Construct a 95%

confidence interval for the mean amount spent on lunch by customers
at this restaurant.

27 / 32 Chapter 5: Estimation
5.2.3 Large-Sample Confidence Interval for a Population
Proportion

Approximate confidence interval for a population proportion

A 1–α confidence interval for the proportion p of items in a population
that belong to a class of interest is
r r
p̂(1 − p̂) p̂(1 − p̂)
p̂ − zα/2 ≤ p ≤ p̂ + zα/2
n n

28 / 32 Chapter 5: Estimation
Example 7. A Pew Research Center survey of 4,787 adults found that
4,178 had bought something online. Of these online shoppers, 789 are
weekly online shoppers.
a) Construct a 95% confidence interval for the proportion of adults
who had bought something online.
b) Construct a 98% confidence interval for the proportion of online
shoppers who are weekly online shoppers

29 / 32 Chapter 5: Estimation
Determining the Sample Size
If we want to be 1–α confident that the error in estimating p by p̂ is no
more than B, the appropriate sample size is
z 2
α/2
n= p̂(1 − p̂).
B
When there is no information concerning the value of p̂, the sample size
is  z 2
α/2
n= 0.25.
B

30 / 32 Chapter 5: Estimation
Example 8. Of 1000 randomly selected cases of lung cancer, 823
resulted in death within 10 years.
a) Using the point estimate of p obtained from the preliminary
sample, what sample size is needed to be 95% confident that the
error in estimating the true value of p is no more than 0.03.
b) How large must the sample be if we want to be at least 95%
confident that the error is no more than 0.03, regardless of the
true value of p.

31 / 32 Chapter 5: Estimation
Exercises for Chapter 5

10.16, 10.26, 10.28 (page 325, 326)

10.54, 10.58, 10.60 (page 331, 332)

12.24, 12.27, 12.28 (page 384, 385)

12.95, 12.97, 12.98 (page 408)

32 / 32 Chapter 5: Estimation