GENG 200

Probability and Statistics

Lecture Note #10

Dr. Nader Meskin

Electrical Engineering Department
Qatar University
1
 Chapter 9:
Tests of Hypotheses for
a single Sample

2
 Chapter Outline
1. Hypothesis Testing
2. Tests on the Mean of a normal distribution,
variance Known
3. Tests on the Mean of a normal distribution,
variance Unknown

3
 Statistical Hypothesis
In Chapter 8, it is shown how to construct a confidence interval estimate of a parameter
from sample data.

In this chapter, we will investigate how to decide which of two competing claims or
statements about some parameter is true. This decision making process is called
Hypothesis Testing.

Statistical hypothesis testing and confidence interval estimation of parameters are the
fundamental methods used at the data analysis stage of a comparative experiment, in
which the engineer is interested, for example, in comparing the mean of a population to a
specified value.

4
 Example:

An engineer is designing an air crew escape system that

consist of ejection sear and a rocket motor that powers the
seat.

The rocket motor contains a propellant, for the proper

ejection, the propellant should have a mean of burning rate of
50 cm/sec.

The practical engineering question is:

Does the mean burning rate of the propellant

5 equal 50 cm/sec?
Two-sided alternative hypothesis

Null hypothesis

Alternative hypothesis

One-sided alternative hypothesis

One-sided alternative hypothesis

6
 Tests of Statistical Hypotheses

Acceptance Critical Region

Critical Region
Region

The boundary values are called the critical values.

The decision is made as follows:

If 48.5 ≤ x < 51.5, we will not reject the null hypothesis

If either x < 48.5 or x > 51.5, we will reject the null hypothesis

The decision procedure can lead to two wrong conclusions.

7
 Tests of Statistical Hypotheses

8
 Type I Error (significant level, α-error)

Consider the previous exam, when a type I error will occur?

We reject the null hypothesis if either x < 48.5 or x > 51.5

Therefore, a type I occurs if the true mean is μ=50 but we have
x < 48.5 or x > 51.5
Suppose that we know the standard deviation as σ=2.5 .
Using central limit theorem, we know that sample mean is approximately
normal with mean μ=50 and the standard deviation as

σ / n = 2.5 / 10 = 0.79

9
10
How can we reduce α?

Lets consider the critical values 48 and 52. (widening the acceptance region).

Then
" 48 − 50 % " 52 − 50 %
α = P$Z < '+ P $ Z > ' = 0.0114
# 0.79 & # 0.79 &

11
We can also reduce α by increasing the sample size.

If n=16, then

σ / n = 2.5 / 16 = 0.625

The α-error will be

" 48.5 − 50 % " 51.5 − 50 %

α = P$Z < '+ P $ Z > ' = 0.0164
# 0.625 & # 0.625 &

12
 Type II error (β-error)

To calculate β, we must have a specific alternative hypothesis, that is, we must have a
particular value of μ.

β = P(48.5 ≤ X ≤ 51.5 when µ = 52)

13
14
 The probability of making a type II error β increases rapidly as
the true value of μ approach the hypothesized value.

Consider the case where the true value of the mean is μ=50.5 and
the hypothesized value is H0:μ=50. The true value of μ is very
close to 50 and the value of β is

β = P(48.5 ≤ X ≤ 51.5 when µ = 50.5)

15
 Type II error probability is much higher for the case that the true mean
16
is 50.5 than the case where the mean is 52.
 What is the effect of sample size on β?

When n=16, then we have σ / n = 2.5 / 16 = 0.625 and

β = P(48.5 ≤ X ≤ 51.5 when µ = 52)

48.5 − 52 51.5 − 52
β = P( ≤Z≤ ) = 0.2119
0.625 0.625

Increasing the sample size results in a decrease in the

probability of type II error.

17
• Generally, the analyst controls the type I error probability α when
he or she selects the critical values.
• Strong Conclusion: Rejection of the null Hypothesis H0

• What value should be used for α?

18
 Power of a statistical test

The power is computed as 1-β and it can be interpreted as

The probability of correctly rejecting a false null hypothesis

19
 One-Sided and Two Sided Hypotheses

H 0 : µ = µ0
Two-sided Hypothesis:
H1 : µ ≠ µ 0

H 0 : µ = µ0
H1 : µ > µ 0
One-sided Hypothesis:

H 0 : µ = µ0
H1 : µ < µ 0

20
21
22
 Tests on the Mean of a Normal Distribution,
Variance known
It is assumed that a random sample X1, X2,…,Xn has been taken from the normal
distribution. The sample mean X is normal with mean μ and variance σ2 /n.

We wish to test the hypotheses: H 0 : µ = µ0

H1 : µ ≠ µ 0

Test Statistics: X − µ0
Z0 =
σ/ n
From a random sample of size n, we can compute the sample mean x and the
standard z-value
x − µ0
z0 =
σ/ n

23
 For fixed significance level α, we have to determine where to place the critical
region

We should reject H0 if either z0 > zα /2 or z0 < −zα /2

We should fail to reject H0 if

−zα /2 ≤ z0 ≤ zα /2
24
25
26
 Tests on the Mean of a Normal Distribution,
Variance Unknown
It is assumed that a random sample X1, X2,…,Xn has been taken from the normal
distribution with mean μ and variance σ2.

We wish to test the hypotheses: H 0 : µ = µ0

H1 : µ ≠ µ 0

Test Statistics: X − µ0
T=
S/ n

has a t distribution with n-1 degrees of freedom.

From a random sample of size n, we can compute the sample mean x and the
standard t-value
x − µ0
t0 =
σ/ n
27
 For fixed significance level α, we have to determine where to place the critical
region

We should reject H0 if either t0 > tα /2,n−1 or t0 < −tα /2,n−1

We should fail to reject H0 if

−tα /2,n−1 ≤ t0 ≤ tα /2,n−1
28
29
 Reading Assignments

— Chapter 9: Sections 9.1 (except 9.1.4, 9.1.5, 9.1.6), 9.2.1,9.3.1

30