CHAPTER 3:

STATISTICAL
INFERENCES (WEEK 8)
3.1 Introduction
3.2 Sampling distribution WEEK 8

3.3 Inference for single population

3.4 Inference for two populations WEEK 9

3.5 Nonparametric statistics WEEK 10

3.1 INTRODUCTION
Descriptive Statistics Inferential Statistics
Summarize, organize and display a Using sample for making a
set of data numerically or generalization about the population
graphically
Statistical inference is the process of drawing conclusions about
populations by analysis of data. The population is assumed to be
larger than the observed data set (i.e the observed data is assumed
to be sampled from a larger population).
Inferential statistical analysis infers properties about a population.
This includes parameter estimation and hypothesis testing.
3.2 SAMPLING DISTRIBUTION

Sampling distribution:
A sampling distribution is a probability distribution of a statistic
obtained through a large number of samples drawn from a specific
population.

Why we do sampling?
1. Sample only contains portion of a population, thus save time
(less time consumes), less costly and less uses of resources.
2. It is more practical than studying the entire population.
 Relationship between population distribution & sampling distribution
of the sample mean

1. The mean of the sample means is exactly equal to the population mean.
2. The dispersion of the sampling distribution is narrower than the original
distribution.
3. The sampling distribution of the sample means tend to become bell-shaped
and
approximately normal.

Central limit theorem

If the sample size is sufficiently large, the sampling distribution of a sample

statistic (eg: sample mean) is approximately normal
Regardless of the shape of the original population.
3.3 INFERENCE FOR SINGLE POPULATION

Definitions:
Estimation
A process of estimating the value of population parameter that is obtained from
sample information.

Point estimate
A single value estimate of a parameter.

Example: Point estimate of population mean,  is the sample mean, x .

An Interval Estimate
In interval estimation, an interval is constructed around the point
estimate and it is stated that this interval is likely to contain the
corresponding population parameter.
Confidence Level and Confidence Interval
Each interval is constructed with regard to a given confidence level
and is called a confidence interval. The confidence level associated
with a confidence interval states how much confidence we have that
this interval contains the true population parameter. The confidence
level is denoted by (1 −  )100%.
 Confidence Interval for Mean, 
The (1 −  )100% Confidence Interval of Population Mean, 

 MARGIN ERROR, E
Known Unknown

  
x  z  n  30 n  30
2

n
s
Margin error tells the
 s  x  t
maximum difference between
x  z   2
,v
n
the point estimate of a 2 n
parameter and the actual value
of the parameter
v = n −1
Confidence Interval Estimates for Population Mean
The (1 −  )100% Confidence Interval of Population Mean, 
     
(i) x  z     x − z    x + z 
2 n  2 n 2 n

 s   s s 
(ii) x  z     x − z    x + z 
2 n  2 n 2 n

s  s s 
(iii) x  tn−1,v   x − tn−1,    x + tn−1, 
n  2 n 2 n
 Determining Sample Size for population mean problems

Determining
Sample Size for the mean

From CI of the mean,

determine the sampling error


Sampling error (margin of error Sample size for
population:
X  Z / 2 
@ error of estimation)
 z /2 
2
n E = Z / 2 n= 
n  E 
 Example 1
If a random sample of size n = 20 from a normal population with the
variance  2 = 225has the mean x = 64.3, construct a 95% confidence
interval for the population mean,  .
Solution:
It is known that, n = 20,  = x = 64.3 and  2 = 225 thus  = 15
For 95% CI,
95% = 100(1 –  )%
1 – = 0.95
 = 0.05

= 0.025  z = z0.025 = 1.96 11
2 2
   
Hence, 95% CI = x  z  
2  n 
 15 
= 64.3  1.96  
 20 
= 64.3  6.57
= [57.73, 70.87]
@
57.73    70.87

Thus, we are 95% confident that the mean of random variable

is between 57.73 and 70.87
12
Example 2 :
A publishing company has just published a new textbook. Before the company decides
the price at which to sell this textbook, it wants to know the average price of all such
textbooks in the market. The research department at the company took a sample of 36
comparable textbooks and collected the information on their prices. This information
produced a mean price RM 70.50 for this sample. It is known that the standard deviation
of the prices of all such textbooks is RM4.50. Construct a 90% confidence interval for the
mean price of all such college textbooks.

Solution:
It is known that, n = 36
x = RM70.50
 = RM 4.50 13
For 90% CI,
90% = 100(1 –  )%
1 – = 0.90
 = 0.1

= 0.05  z = z0.05 = 1.65
2 2

  
Hence, 90% CI = x  z  
2  n
 4.50 
= 70.50  1.65  
 36 
= 70.50  1.24
= [ RM 69.26, RM 71.74]
Thus, we are 90% confident that the mean price of all such college
14

textbooks is between RM69.26 and RM71.74.

Example 3:
A team of efficiency experts intends to use the mean of a random sample of size n=150 to
estimate the average mechanical aptitude of assembly-line workers in a large industry (as
measured by a certain standardized test). If, based on experience, the efficiency experts can
assume that  = 6.2 for such data, what can they assert with probability 0.99 about the
maximum error of their estimate?
Solution:
Substituting n = 150,  = 6.2, and for 0.99 probability
1 – = 0.99
 = 0.01

= 0.005  z = z0.005 = 2.575
2 2

into the expression for the maximum error, we get

z  2.575(6.2)
E =  /2 = = 1.30
n 150
Thus, the efficiency experts can assert with probability 0.99 that their
15

error will be less than 1.30.

 Example 4:
• If  = 45, what sample size is needed to estimate the mean within ± 5 with 90% confidence.
Solution:
For 90% CI,
90% = 100(1 –  )%
1 – = 0.90
 = 0.1

= 0.05  z = z0.05 = 1.65
2 2

Z 2 σ2 (1.65) 2 (45) 2
n= 2
= 2
= 219.19 (Always round up)
E 5
• So the required sample size is n = 220
 Confidence Interval Estimates for Population Proportion

The (1 −  )100% Confidence Interval for p for Large Samples

(n  30) :

pˆ qˆ x
pˆ  z where pˆ = and qˆ = (1 − pˆ )
2 n n
or

pˆ qˆ pˆ qˆ
pˆ − z  p  pˆ + z
2 n 2 n 17
 Determining Sample Size for population proportion problems

Determining
Sample Size for the proportion

From CI of the mean,

determine the sampling error

Sample size for

Sampling error (margin of error
pˆ (1 − pˆ )
population:

pˆ  z
@ error of estimation)
( Z / 2 ) pˆ (1 − pˆ )
2

n ˆ (1- p
p ˆ) n=
E = Z / 2
2
E2
n
Example 1:
According to the analysis of Women Magazine in June 2005, “Stress has become a common
part of everyday life among working women in Malaysia. The demands of work, family and
home place an increasing burden on average Malaysian women”. According to this poll,
40% of working women included in the survey indicated that they had a little amount of
time to relax. The poll was based on a randomly selected of 1502 working women aged 30
and above. Construct a 95% confidence interval for the corresponding population
proportion.
Solution:

Let p be the proportion of all working women age 30 and above, who have a limited
amount of time to relax.

Let pˆ be the corresponding sample proportion.

n = 1502 , pˆ = 0.40, qˆ = 1 − pˆ = 1 – 0.40 = 0.60
From the given information: 19
For 95% CI,
95% = 100(1 –  )%
1 – = 0.95
 = 0.05

= 0.025  z = z0.025 = 1.96
2 2

 ˆˆ
pq 
Hence, 95% CI = p
ˆ  z 
 

2  n 
 0.40(0.60) 
= 0.40  1.96 
 

 1502 
= 0.40  0.02478
= [0.375, 0.425] or 37.5% to 42.5%

Thus, we can state with 95% confidence that the proportion of all
working women aged 30 and above who have a limited amount of
20
time to relax is between 37.5% and 42.5%.
Example 2:
How large a sample would be necessary to estimate the true proportion defective in a large
population within ±3%, with 95% confidence?
(Assume a sample yields p = 0.12)

Solution:

For 95% confidence, we have Zα/2 = Z0.025 = 1.96

E = 0.03; p = 0.12
ˆ (1 − p
Z / 2 2 p ˆ) (1.96) 2 (0.12)(1 − 0.12)
n= 2
= 2
= 450.74
E (0.03)

(Always round up)

So the required sample size is n = 451

Example 3:
A study is made to determine the proportion of voters in a sizable community who favor the
construction of a nuclear power plant. If 140 of 400 voters selected at random favor the
project and we use pˆ = 140 = 0.35 as an estimate of the actual proportion of all voters in
400
the community who favor the project, what can we say with 99% confidence about the
maximum error?
Solution:
Substituting n = 400, p
ˆ = 0.35, and z0.005 = 2.575 into the formula, we get

ˆ (1 − p
p ˆ) (0.35)(0.65)
E = z / 2 = 2.575 = 0.061
n 400

Thus, if we use p ˆ = 0.35 as an estimate of the actual proportion of voters

in the community who favor the project, we can assert with 99% confidence
22
that the error is less than 0.061.
Example 4:
How large a sample required if we want to be 95% confident that the error in using p̂ to
ˆ = 0.12, find the required sample size.
estimate p is less than 0.05? If p
Solution:


 = 0.05  = 0.025  z0.025 = 1.96, p
ˆ = 0.12, E = 0.05
2
2
 z0.025 
n=  ˆ (1 − p
p ˆ)
 E 
2
 1.96 
=  0.12(0.88)  163
 0.05 
23
HYPOTHESIS TESTING
 HYPOTHESIS TESTING
Everyday, in every aspects of life, problems occur and need to be solved. From the
problems, questions arise and researchers are interested in answering those questions.
For example:
• Is the earth warming up?
• Does a new medication lower blood pressure?
• Does the public prefer a certain color in a new fashion line?
• Is a new teaching technique better than a traditional one?
• Do seat belts reduce the severity of injuries?
These types of questions can be addressed/solved through statistical hypothesis testing,
which is a decision-making process for evaluating claims about a population.

25
Definitions:

Hypothesis testing can be used to determine whether a statement about the value of a
population parameter (such as mean or proportion) should or should not be rejected.

Null hypothesis, H0 : is a claim (or statement) about a population parameter that is

assumed to be true.
(the null hypothesis is either to be rejected or fails to be rejected.)

Alternative hypothesis, H1 : is a claim about a population parameter that will be true

if the null hypothesis is false.
26
Test Statistic is a function of the sample data on which the decision is to be
based. Two common test statistics are:
1. z-test/ z-calc
2. t-test.

There are 3 TYPES of hypothesis testing:

1. RIGHT-TAILED
2. LEFT-TAILED
3.TWO-TAILED TEST.

27
Left-tailed test: The critical value,(CV) is Zα,
separates the critical region from the
noncritical region.

The critical or rejection region is

the range of values of the test value
that indicates that there is a
significant difference and that the
null hypothesis should be rejected.

The noncritical or nonrejection

region is the range of values of the
test value that indicates the difference
Z= - 2.33 was probably due to chance and that
the null hypothesis fail to be rejected.

28
Right-tailed test:
 Two-tailed test:

How do we determine type of hypothesis testing?

It is depending to the null & alternative hypothesis (will be discussing in next slides)
• In reality, the null hypothesis may or may not be true, and a decision is made to reject or
fail to reject it on the basis of the data obtained from a sample.
• A type I error occurs if one rejects the null hypothesis when it is true.
• A type II error occurs if one does not reject the null hypothesis when it is false.

Fail to

37
• The level of significance is the maximum probability of
committing a type I error. This probability is symbolized by α
(alpha). That is,
P(type I error) = α
Likewise,
P(type II error) = β (beta).
• Typical significance levels are:
0.10, 0.05, and 0.01

• For example: when α = 0.10, there is a 10% chance of rejecting a

true null hypothesis. 38
Power of a Statistical Test (1-β)

-Power of test is the probability of rejecting the null hypothesis

when the null hypothesis is false.

Value of β depends on:

1. the value of α, the size of the sample,
2. the population standard deviation,
3. the actual difference between the hypothesized value of the
parameter being tested and the true parameter.
The power of a test measures the sensitivity of the test to detect a real difference in
parameters if one actually exists.
The higher the power, the more sensitive the test is to detecting a real difference between
parameters if there is a difference.
In other words, the closer the power of a test is to 1, the better the test is for rejecting the
null hypothesis if the null hypothesis is, in fact, false.
The power is 1 – β.
When conducting a statistical study, use the test that has the highest power for the data.
There are times when the researcher has a choice of two or more statistical tests to test the
hypotheses. The tests with the highest power should be used. It is important, however, to
remember that statistical tests have assumptions that need to be considered.

40
Hypotheses
There are two hypotheses :
1. Null hypothesis, H 0
2. Alternative hypothesis, H1
Claim
When a researcher conducts a study, he or she is generally
looking for evidence to support a claim. Therefore, the
claim should be stated as the alternative hypothesis, or
research hypothesis. 41
Three methods used to test hypotheses:
1. The traditional method
2. The confidence interval method
3. The P-value method
Hypothesis and Test Procedures for mean
(Traditional Method)
A standard statistical test of hypothesis consist of :
1. State the Null hypothesis, H 0 and Alternative
hypothesis, H1 . Common phrases of hypotheses.
2. Find the critical value, Zα (refer to table z in Textbook)
3. Calculate the test statistic, Zstat/Zcalc.
4. Determine the rejection region.
45

5. Make the conclusion.

 TEMPLATE ANSWERING HYPOTHESIS TESTING QUESTION

Step 1 State the hypotheses and identify the claim.

HYPOTHESIS SYMBOL

H0  = constant (   constant,   constant )

H1   constant (   constant,   constant )

Step 2 Find the critical value(s) from the appropriate statistical table (Table
Z = ?
6 in Textbook).

value of  depends on the test (right, left, two-tail test),

can be determine from H1. Determine rejection region.
Tails of a Test
Two-Tailed Test Left-Tailed Test Right-Tailed Test
Sign in H0 = =@ =@
Sign in H1   
Rejection Region In both tail In the left tail In the right tail
Rejection Region Z  − z or Z  z
2 2
Z< − z Z  z
t  −t or t  t t  t ,v
2, v 2, v t  −t ,v
v = n −1
Step 3 Compute the test value.
x − x −
Z test = ttest =
s @ s
n n

Step 4 Make the decision to reject or not reject the null hypothesis.

Z test  − z or Ztest  z
2 2
Z test  − z Z test  z
Rejection Region ttest  −t or ttest  t ttest  −t ,v
2, v 2, v
ttest  t ,v
v = n −1
Case left-tailed test:
Example: Since Ztest = − 1.7 < − z = −1.5 thus we decide to REJECT H 0 .
Example: Since Ztest =1.2 > − z = −1.5 thus we decide FAIL to / DO
NOT REJECT H 0 .
Step 5 Summarise the results.

Example of conclusion when rejection H 0 :

There is sufficient/ enough evidence to support the claim that ……………

Example of conclusion when fail/ do not rejection H 0 :

There is insufficient/ not enough evidence to support the claim that
……………………

**Summary/ conclusion is basically answering the question thus it will

be based on the question**
Example 1: H0
The average monthly earnings for women in managerial and professional positions is
RM 2400. Do men in the same positions have the average monthly earnings that are
higher than those for women? A random sample of n = 40 men in managerial and
professional positions showed x = RM 3600 and s = RM400. Test the appropriate
hypothesis using  = 0.01. H 1

H 0 : The average monthly earnings for women in M&P positions is RM 2400

H 0 :  = 2400

H1 : The average monthly earnings for men in M&P positions is higher than
RM 2400 ( women )
60

H1 :   2400
Solution:
1.The hypothesis to be tested are,
H 0 :  = 2400
Right-tailed test
H1 :   2400
2.  =0.01,  is known and n  30, We will use normal distribution, Z
Rejection Region : Z  z ; z = z0.01 = 2.33
3. Test Statistic
x −  3600 − 2400
Z= = = 18.97
s 400
n 40
 falls in the 
4. Since Z test = 18.97 is greater than z0.01 = 2.33 (18.97  2.33) , we reject H 0  .
 rejection region 
5. Thus, we conclude that average monthly earnings for men in managerial and professional
61
positions are significantly higher than those for women
 Example 2: Professors’ Salaries H1
A researcher reports that the average salary of assistant professors is more than $42,000. A
sample of 30 assistant professors has a mean salary of $43,260. At α = 0.05, test the claim
that assistant professors earn more than $42,000 per year. The standard deviation of the
population is $5230.

Step 1: State the hypotheses and identify the claim.

H0: μ = $42,000
H1: μ > $42,000 (claim)

Step 2: Find the critical value.

Since α = 0.05 and the test is a right-tailed test, the critical value is
Z0.05 = 1.65. 62
Step 3: Compute the test value.
X − 43260 − 42000
z= = = 1.32
 n 5230 30
Step 4: Make the decision.
Since the test value, 1.32, is less than the critical value, 1.65, and is not
in the critical region, the decision is fail to reject the null hypothesis.

Step 5: Summarize the results.

There is not enough evidence to support the claim that assistant
professors earn more on average than $42,000 per year. 63
CONFIDENCE INTERVAL METHOD
Additional Topics Regarding Hypothesis Testing
• There is a relationship between confidence intervals and
hypothesis testing.
• When the null hypothesis is rejected in a hypothesis-testing
situation, the confidence interval for the mean using the same
level of significance will not contain the hypothesized mean.
• Likewise, when the null hypothesis is not rejected, the
confidence interval computed using the same level of
significance will contain the hypothesized mean.
* Applied to two-tailed hypothesis testing ONLY*

66
 Example: Sugar Production
Sugar is packed in 5-pound bags. An inspector suspects the bags may not contain 5 pounds.
A sample of 50 bags produces a mean of 4.6 pounds and a standard deviation of 0.7 pound.
Is there enough evidence to conclude that the bags do not contain 5 pounds as stated at α =
0.05? Also, find the 95% confidence interval of the true mean.

Step 1: State the hypotheses and identify the claim.

H0: μ = 5 and
H1: μ  5 (claim)

Step 2: Find the critical value.

The critical values are z 0.025= ±1.96.

67
Step 3: Compute the test value.

X − 4.6 − 5.0
z= = = −4.04
 n 0.7 50

Step 4: Make the decision.

Since Z test = −4.04 < z 0.025= –1.96. Thus we reject the null hypothesis.

Step 5: Summarize the results.

There is enough evidence to support the claim that the bags do not weigh 5
pounds.

68
 Example: Sugar Production
Recall that: H0: μ = 5 and H1: μ  5 (claim)

For 95% CI,


 = 0.05  = 0.025  z = z0.025 = 1.96
2 2

 s   0.7 
Hence, 95% CI = x  z   = 4.6  1.96   = 4.6  0.2
2  n  50 
= [4.4, 4.8]

Notice that the 95% confidence interval of  does not contain the
hypothesized value μ = 5. Thus we reject the null hypothesis.

Hence, there is agreement between the hypothesis test and the confidence
interval. 69
P-VALUE METHOD
 Hypothesis Testing
The P-value (or probability value) is the probability of getting
a sample statistic (such as the mean) or more extreme sample
statistic in the direction of the alternative hypothesis when the
null hypothesis is true.
p-value is the probability calculated using the test statistic. The
smaller the p-value, the more contradictory it is the data to H .
0

P-value
P-Value

Test Value
71
 Hypothesis Testing
• In this section, the traditional method for solving hypothesis-
testing problems compares z-values:
• critical value
• test value
• The P-value method for solving hypothesis-testing problems
compares areas:
• alpha
• P-value
72
 Procedures:
Solving Hypothesis-Testing Problems
(P-Value Method)
Step 1 State the hypotheses and identify the claim.
Step 2 Compute the test value.
Step 3 Find the P-value.
Step 4 Make the decision.
Step 5 Summarize the results.

73
 p-value
p-value is the smallest significance level at which the null hypothesis is rejected.

Using the p − value approach, we reject the null hypothesis, H 0 if

p − value   for one − tailed test
p − value   for two − tailed test

and we do not reject the null hypothesis, H 0 if

p − value   for one − tailed test
p − value   for two − tailed test 74
 One-tailed test
Two-tailed test
left tailed right talied

Reject null hypothesis p − value   p − value   p − value  

Do not reject null hypothesis p − value   p − value   p − value  

 Example: Cost of College Tuition

A researcher wishes to test the claim that the average cost of tuition
and fees at a four-year public college is greater than $5700. She
selects a random sample of 36 four-year public colleges and finds the
mean to be $5950. The population standard deviation is $659. Is there
evidence to support the claim at α 0.05? Use the P-value method.

Step 1: State the hypotheses and identify the claim

H0: μ = $5700
H1: μ > $5700 (claim)
76
Step 2: Compute the test value.
X − 5950 − 5700
z= = = 2.28
 n 659 36

Step 3: Find the P-value.

Using Table 5 (Textbook), find the area for 0<z < 2.28.
The area is 0.4887.
Subtract from 0.5 to find the area of the tail.
Hence, the P-value is 0.5 – 0.4887 = 0.0113.
77
Step 4: Make the decision.
Since the P-value = 0.0113 < α = 0.05, the decision is to reject
the null hypothesis.

Step 5: Summarize the results.

There is enough evidence to support the claim that the tuition
and fees at four-year public colleges are greater than $5700.
Note: If α = 0.01, the null hypothesis would not be rejected.
78
 Example: Wind Speed
The average wind speed in a certain city is 8 miles per hour. A
researcher claims that the average wind speed in a certain city may
not be 8 miles per hour. A sample of 32 days has an average wind
speed of 8.2 miles per hour. The standard deviation of the
population is 0.6 mile per hour. At α = 0.05, is there enough
evidence to reject the claim? Use the P-value method.
Step 1: State the hypotheses and identify the claim.
H0: μ = 8
H1: μ ≠ 8 (claim)
X − 8.2 − 8
Step 2: Compute the test value. z = = = 1.89
 n 0.6 32
79
Step 3: Find the P-value.
The area for 0<z <1.89 is 0.4706.
Subtract: 0.5 – 0.4706 = 0.0294.
Since this is a two-tailed test, the area of 0.0294 must be doubled
to get the P-value. Thus, P-value is 2(0.0294) = 0.0588.
Step 4: Make the decision.
Since the P-value = 0.0588 > α = 0.05, the decision is do not reject
the null hypothesis.

Step 5: Summarize the results.

There is no enough evidence to support the claim that the
average wind speed is 8 miles per hour.
80
 Exercise:

In a packing factory, average time taken for an old machine to pack

boxes is 432 seconds. Machine operator believed that the new
machine may affect the average time taken to pack the boxes. A
sample of 40 boxes from new machine are taken and the average
time taken to pack boxes is recorded to be 421 seconds and the
standard deviation is 64 seconds. By using significance level 0.05,
test the hypothesis from the machine operator by using the three
METHODS of hypothesis testing:
1. Confidence interval method
2. Traditional method
3. P-value method
81
Hypothesis and Test Procedures for proportion
(Traditional Method)
A standard statistical test of hypothesis consist of :
1. State the Null hypothesis, H 0 and Alternative
hypothesis, H1 .
2. Find the critical value, Zα (refer to table 6 in Textbook)
3. Calculate the test statistics, Zstat/Zcalc.
4. Determine the rejection region. 84

5. Make the conclusion.

Testing Hypothesis on the single Population Proportion, p
Null Hypothesis : H 0 : p = p0

ˆ − p0
p
Test Statistic : Z test =
p0 q0
n

Alternative hypothesis Rejection Region

H1 : p  p0 Z  − z 2 or Z  z 2

H1 : p  p0 Z  z
H1 : p  p0 Z< − z
85
Example 1:

When working properly, a machine that is used to make chips for calculators produce
4% defective chips. Whenever the machine produces more than 4% defective chips it
needs an adjustment. To check if the machine is working properly, the quality control
department at the company often takes sample of chips and inspects them to
determine if the chips are good or defective. One such random sample of 200 chips
taken recently from the production line contained 14 defective chips. Test at the 5%
significance level whether the machine needs an adjustment.

86
Solution:

1. The hypotheses to be tested are:

H 0 : p = 0.04
H1 : p  0.04

2. Critical value: z = z0.05 = 1.65

Rejection Region : Z  z

3. Test statistic is
ˆ − p0
p 0.07 − 0.04
Z test = = = 2.17
p0 q0 0.04(0.96)
n 200 87
4. Since Z test = 2.17  z0.05 = 1.65, falls in the rejection region,
thus we can reject H 0 .

5. We conclude that there is enough evidence to support the claim

that the machine needs an adjustment.