NATIONAL OPEN UNIVERSITY OF NIGERIA

APPLIED STATISTICS
ECO 452

FACULTY OF SOCIAL SCIENCES

COURSE GUIDE

Course Developers:
Dr. Adesina- Uthman Ganiyat Adejoke
Department of Economics, Faculty of Social Sciences,
National Open University of Nigeria.
and
Ogunjirin Olakunle
Yaba College of Technology
School of Liberal Studies
Department of Social Sciences.

Course Editor:

Dr. Ogunsakin Sanya

Department of Economics
Senior Lecturer,
Ekiti State University, Ado-Ekiti.
NATIONAL OPEN UNIVERSITY OF NIGERIA
National Open University of Nigeria
Headquarters
Plot 91, Cadastral Zone, University Village,
Nnamdi Azikiwe Expressway, Jabi, Abuja.
Introduction
Statistical economics is a branch of economic that deals with analysis of economic
phenomenon.

What you will learn in the course

In this course, you will be introduced to various analytical tools in statistics; Here, you
will be exposed to the underlying assumptions, formulae and calculations of the topics
under consideration. Also, you will be taught the decision criteria under each topic.

Course Content
This course will expose you to different statistical tools that economist can apply in
economic analysis. This course is built on the foundation of elementary statistics and
elementary economics in the understanding of real life situation.

Course Aims
There are fourteen study units in the course and each unit has its objectives. You are
advised to read through the objective of each them and bear them in mind as you through
each of the unit. In addition these objective is the overall objective which includes;
- Exposing you to basic statistical tools that can be applied in economics,
- Apply these tools to real life situation,
- Expose the students to economic interpretation of all calculated coefficients

Course objectives
The over-all objectives of this course are;
- to expand the learning horizons of the students
- to understand how to apply statistical tool in economics
Working through This Course
You have to work through all the study units in the course. There are four modules and
fourteen study units in all.

Course Materials
Major components of the course are:
1. Course Guide
2. Study Units
3. Textbooks
4. CDs
5. Assignments File
6. Presentation Schedule

Study Units
The breakdown of the four Modules and the 14 study units are as follows:
Module 1; Statistical Inference
Unit 1; sampling distribution defined
Unit 2; sampling distribution of proportion
Unit 3; sampling distribution of difference and sum of two means
Unit 4; probability distribution

Module 2: Analysis of Variance and Analysis of Covariance

Unit 1: one-way factor analysis of variance
Unit 2: two-way factor analysis of variance
Unit 3: analysis of covariance

Module 3: Multiple Regression Analysis

Unit 1: Estimation of multiple regressions
Unit 2: Partial correlation coefficient
Unit 3: Multiple correlation coefficient and coefficient of determination
Unit 4: Overall test of significance

Module 4: Time Series Analysis

Unit 1: time series and its components
Unit 2; Quantitative estimation of time series
Unit 3: price index

References and Other Resources

Attached to every unit is a list of references and further reading. Try to get as many as
possible of those textbooks and materials listed. The textbooks and materials are meant to
deepen your knowledge of the course.

Assignment File
In this file, you will find all the details of the work you must submit to your tutor for
marking. The marks you obtain from these assignments will count towards the final mark
you obtain for this course. Further information on assignments will be found in the
Assignment File itself and later in this Course Guide in the section on assessment.

Presentation Schedule
The Presentation Schedule included in your course materials gives you the important
dates for the completion of tutor-marked assignments and attending tutorials. Remember,
you are required to submit all your assignments by the due date. You should guard
against falling behind in your work.

Assessment
Your assessment will be based on tutor-marked assignments (TMAs) and a final
examination which you will write at the end of the course.
Tutor Marked Assignments (TMA)
Every unit contains at least one or two assignments. You are advised to work through all
the assignments and submit them for assessment. Your tutor will assess the assignments
and select four which will constitute the 30% of your final grade. The tutor-marked
assignments may be presented to you in a separate file. Just know that for every unit there
are some tutor-marked assignments for you. It is important you do them and submit for
assessment.

Final Examination and Grading

At the end of the course, you will write a final examination which will constitute 70% of
your final grade. In the examination which shall last for two hours, you will be requested
to answer three questions out of at least five questions.

Course Marking Scheme

This table shows how the actual course Marks
marking is broken down. Assessment
Assignments Four assignments, best three marks of
the four count at 30% of course marks
Final Examination 70% of overall course marks
Total 100% of course marks
CONTENT
Introduction
Course Content
Course Aims
Course Objectives
Working through This Course
Course Materials
Study Units
Textbooks and References
Assignment File
Presentation Schedule
Assessment
Tutor-Marked Assignment (TMAs)
Final Examination and Grading
Course Marking Scheme
Course Overview
How to Get the Most from this Course
Tutors and Tutorials
Summary

Introduction
The course advanced statistics (ECO 410) is a first semester course which carries two
credit units for fourth year level economics students in the School of Art and Social
Sciences at the National Open University, Nigeria. The course is a very useful course to
you in your academic pursuit, because it helps gain in-depth insight of the underlining
statistical tools usually used by economists.
This course guide tells you what advanced statistics entails, what course materials you
will be using and how you can work your way through these materials. It suggests some
general guidelines for the amount of time required of you on each unit in order to achieve
the course aims and objectives successfully. It also provides you some guidance on your
tutor marked assignments (TMAs) as contained herein.

Course Content
This course is built on the foundation of what you have learnt in your elementary
statistics. Topics covered include: statistical inference, probability distribution, analysis
of variance, multiple regressions, time series, price index.

Working through the Course

To successfully complete this course, you are required to read the study units, referenced
books and other materials on the course.
Each unit contains self-assessment exercises called Tutor Marked Assessment. At some
points in the course, you will be required to submit assignments for assessment purposes.
At the end of the course there is a final examination. This course should take about
15weeks to complete and some components of the course are outlined under the course
material subsection.

Course Material
The major component of the course, What you have to do and how you should allocate
your time to each unit in order to complete the course successfully on time are listed
follows:
1. Course guide
2. Study unit
3. Textbook
4. Assignment file
5. Presentation schedule
Tutor-Marked Assignments (TMAs)
There are four tutor-marked assignments in this course. You will submit all the
assignments. You are encouraged to work all the questions thoroughly. The TMAs
constitute 30% of the total score.

Assignment questions for the units in this course are contained in the Assignment File.
You will be able to complete your assignments from the information and materials
contained in your set books, reading and study units. However, it is desirable that you
demonstrate that you have read and researched more widely than the required minimum.
You should use other references to have a broad viewpoint of the subject and also to give
you a deeper understanding of the subject.

When you have completed each assignment, send it, together with a TMA form, to your
tutor. Make sure that each assignment reaches your tutor on or before the deadline given
in the Presentation File. If for any reason, you cannot complete your work on time,
contact your tutor before the assignment is due to discuss the possibility of an extension.
Extensions will not be granted after the due date unless there are exceptional
circumstances.

Final Examination and Grading

The final examination will be of two hours' duration and have a value of 70% of the total
course grade. The examination will consist of questions which reflect the types of self-
assessment practice exercises and tutor-marked problems you have previously
encountered. All areas of the course will be assessed

Use the time between finishing the last unit and sitting for the examination to revise the
entire course material. You might find it useful to review your self-assessment exercises,
tutor-marked assignments and comments on them before the examination. The final
examination covers information from all parts of the course.
Course Marking Scheme
The table presented below indicate the total marks (100%) allocation.

Assessment Marks
Assignment (Best three assignment out of the four marked) 30%
Final Examination 70%
Total 100%

Course Overview
The table presented below indicate the units, number of weeks and assignments to be
taken by you to successfully complete the course, advanced statistics (ECO 410).

How to Get the Most from This Course

In distance learning the study units replace the university lecturer. This is one of the great
advantages of distance learning; you can read and work through specially designed study
materials at your own pace and at a time and place that suit you best.
Think of it as reading the lecture instead of listening to a lecturer. In the same way that a
lecturer might set you some reading to do, the study units tell you when to read your
books or other material, and when to embark on discussion with your colleagues. Just as
a lecturer might give you an in-class exercise, your study units provides exercises for you
to do at appropriate points.

Each of the study units follows a common format. The first item is an introduction to the
subject matter of the unit and how a particular unit is integrated with the other units and
the course as a whole. Next is a set of learning objectives. These objectives let you know
what you should be able to do by the time you have completed the unit.
You should use these objectives to guide your study. When you have finished the unit
you must go back and check whether you have achieved the objectives. If you make a
habit of doing this you will significantly improve your chances of passing the course and
getting the best grade.

The main body of the unit guides you through the required reading from other sources.
This will usually be either from your set books or from a readings section. Some units
require you to undertake practical overview of events. You will be directed when you
need to embark on discussion and guided through the tasks you must do.
The purpose of the practical overview of some certain practical issues are in twofold.
First, it will enhance your understanding of the material in the unit. Second, it will give
you practical experience and skills to evaluate economic propositions, arguments, and
conclusions. In any event, most of the critical thinking skills you will develop during
studying are applicable in normal working practice, so it is important that you encounter
them during your studies.

Self-assessments are interspersed throughout the units, and answers are given at the ends
of the units. Working through these tests will help you to achieve the objectives of the
unit and prepare you for the assignments and the examination. You should do each self-
assessment exercises as you come to it in the study unit.

The following is a practical strategy for working through the course. If you run into any
trouble, consult your tutor. Remember that your tutor's job is to help you. When you need
help, don't hesitate to call and ask your tutor to provide it.

1. Read this Course Guide thoroughly.

2. Organize a study schedule. Refer to the `Course overview' for more details. Note
the time you are expected to spend on each unit and how the assignments relate to
the units. Important information, e.g. details of your tutorials, and the date of the
first day of the semester is available from study centre. You need to gather
together all this information in one place, such as your dairy or a wall calendar.
 Whatever method you choose to use, you should decide on and write in your own
dates for working breach unit.
3. Once you have created your own study schedule, do everything you can to stick to
it. The major reason that students fail is that they get behind with their course
work. If you get into difficulties with your schedule, please let your tutor know
before it is too late for help.
4. Turn to Unit 1 and read the introduction and the objectives for the unit.
5. Assemble the study materials. Information about what you need for a unit is given
in the `Overview' at the beginning of each unit. You will also need both the study
unit you are working on and one of your set books on your desk at the same time.
6. Work through the unit. The content of the unit itself has been arranged to provide
a sequence for you to follow. As you work through the unit you will be instructed
to read sections from your set books or other articles. Use the unit to guide your
reading.
7. Up-to-date course information will be continuously delivered to you at the study
centre.
8. Work before the relevant due date (about 4 weeks before due dates), get the
Assignment File for the next required assignment. Keep in mind that you will
learn a lot by doing the assignments carefully. They have been designed to help
you meet the objectives of the course and, therefore, will help you pass the exam.
Submit all assignments no later than the due date.
9. Review the objectives for each study unit to confirm that you have achieved them.
If you feel unsure about any of the objectives, review the study material or consult
your tutor.
10. When you are confident that you have achieved a unit's objectives, you can then
start on the next unit. Proceed unit by unit through the course and try to pace your
study so that you keep yourself on schedule.
11. When you have submitted an assignment to your tutor for marking do not wait for
it return `before starting on the next units. Keep to your schedule. When the
 assignment is returned, pay particular attention to your tutor's comments, both on
the tutor-marked assignment form and also written on the assignment. Consult
your tutor as soon as possible if you have any questions or problems.
12. After completing the last unit, review the course and prepare yourself for the final
examination. Check that you have achieved the unit objectives (listed at the
beginning of each unit) and the course objectives (listed in this Course Guide).

Tutors and Tutorials

There are some hours of tutorials (2-hours sessions) provided in support of this course.
You will be notified of the dates, times and location of these tutorials. Together with the
name and phone number of your tutor, as soon as you are allocated a tutorial group.

Your tutor will mark and comment on your assignments, keep a close watch on your
progress and on any difficulties you might encounter, and provide assistance to you
during the course. You must mail your tutor-marked assignments to your tutor well
before the due date (at least two working days are required). They will be marked by your
tutor and returned to you as soon as possible.

Do not hesitate to contact your tutor by telephone, e-mail, or discussion board if you need
help. The following might be circumstances in which you would find help necessary.
Contact your tutor if.
• You do not understand any part of the study units or the assigned readings
• You have difficulty with the self-assessment exercises
• You have a question or problem with an assignment, with your tutor's comments on an
assignment or with the grading of an assignment.

You should try your best to attend the tutorials. This is the only chance to have face to
face contact with your tutor and to ask questions which are answered instantly. You can
raise any problem encountered in the course of your study. To gain the maximum benefit
from course tutorials, prepare a question list before attending them. You will learn a lot
from participating in discussions actively.

Summary
On successful completion of the course, you would have developed critical thinking skills
with the material necessary for efficient and effective use of statistical tools economics.
However, to gain a lot from the course please try to apply anything you must have learnt
in the course to practice by doing the calculation on paper yourself. We wish you success
with the course and hope that you will find it both interesting and useful.
MODULE ONE; Statistical Inference
Unit 1; Sampling distribution defined
Unit 2; Sampling distribution of proportion
Unit 3; Sampling distribution of difference and sum of two means
Unit 4; Probability distribution
UNIT ONE; Sampling Distribution
CONTENT
1.0 Introduction
2.0 Objectives
3.0 Main Content
3.1 Sampling Distribution, Population and Sample Defined
3.2 Sampling Distribution of Parameter Estimate
3.3 Estimate of Sample Statistics
3.4 Estimators for Mean and Variance
3.5 The Role and Significant of Statistics in Social Sciences
4.0 Conclusion
5.0 Summary
6.0 Tutor-Marked Assignment
7.0 References/ Further Readings

1.0 INTRODUCTION
Generally statistical data are studied in order to learn something about the broader field
which the data represents. In order to make statistical work meaningful, statistician
generalize from what we find in the figure at hand to the wider phenomenon which they
represent. In technical language we regard a set of data as a sample drawn from a larger
“universe”. We analyze the data of the sample in order to draw conclusion about the
corresponding universe or population.
In a sense universe actually exists and it is theoretically possible to study the universe
completely. But in another sense the universe is broader and in a sense less tangible.
This unit happens to be one of the four units in this module, for proper understanding of
the topics in this unit a thorough knowledge of elementary statistics is required.

2.0 OBJECTIVES
At the end of this unit you should be able to understand the following;
- Sample
- Population
- Sampling theory
- Parameter estimation
- Estimate sample mean, population mean etc.

3.0 MAIN CONTENT

3.1 Sampling Theory, Population and Sample Defined
Statistical inference is defined as the process by which on the basis of sample we draw
conclusion about the universe from which sample is drawn. It can as well be defined as a
process by which conclusion are drawn about some measure or attribute of a population
based upon analysis of sample. Samples are taken and analyzed in order to draw
conclusion about the whole population.
Sampling theory is a study of relationships existing between a population and samples
drawn from the population. Sampling theory is also useful in determining whether the
observed differences between two samples are due to chance variation or whether they
are really significant.
In general, a study of the inference made concerning a population by using sample drawn
from it together with indication of accuracy of such inferences by using probability
theory is called statistical inference. Population of a variable X is usually defined to
consist of all the conceptually possible values that the variable may assume. Some of
these values may have already been observed, others may not have occurred, but their
occurrence is conceivably possible. The number of conceptually possible values of a
variable is called size of the population. This size varies according to the phenomenon
being investigated.
A population may be finite, when it consists of a given number of values or it may be
infinite, when it includes an infinite number of values of the variable.
In most cases values of population are hardly known, what we usually have is a certain
number of values that any particular variable has assumed and which have been recorded
in one way or the other. Such data form a sample from the population.
Sample refers to a collection of observation on a certain variable. The number of
observations included in the sample is called the size of the sample.
The main object of the theory of statistics is the development of method of drawing
conclusion about the population (unknown) from the information provided by a sample.
In order to facilitate the study of population and sample, statisticians have introduced
various descriptive measures that is various characteristics values that describes the
important features of the sample or the population. The most important of these
characteristics are the mean, variance and the standard deviation. To distinguish between
sample and populations statistician use the term parameter for the basic descriptive
measure of population while statistics is usually used for the basic descriptive measure of
a sample.

Table M1.1.1
Basic Descriptive Measure of Population and Sample
Population parameters Symbol Sample statistics Symbol
I Population mean  Sample mean x̅
2
Ii Population variance σx Sample variance Sx2
Iii Population standard σx Sample standard sx
deviation deviation

Note E(x) =  = x1 + x2 + …… xn
n
SELF ASSESMENT EXERCISE
What are descriptive measures that can be used in describing a sample or population?

3.2 Sampling Distribution of Parameter & Sample Estimates

The population mean is usually referred to as the expected value of the population and it
is conventionally denoted as E(x) or . But for a discrete random variable the expected
value is computed by the sum of the product of value of X1 multiplied by their various
probabilities.
E (x) =  = ∑𝑛𝑖=1 xf(X1)
Where Xi is the probability of variable x.
The variance of a population is defined as the expected value of the squared deviations of
the value of x from their expected mean value.
Var (x) = σx2 = ∑ (X – E(x))2 = ∑ (x - )2
n n
Where E (x) = population mean value
This shows the various ways in which the various value of random variable x is
distributed around their expected mean values. The smaller the variance, the closer and
cluster of the values of x around the population mean.
The standard deviation of a population is defined as the square root of the population
variance. This is denoted as:
σx = (x-(Ex))2 =  (x - )2
n n
The standard deviation is a measure that describes how dispersed the values of x is
around the population mean.
COV(XY) = (XY) - X Y
Worked Example
Given the population 11, 12, 13, 14, 15 calculate the mean, standard deviation, and the
variance of the given population.

Table M1.1.2
Table of Analysis for Sample Mean, Standard Deviation and Variance
X X- (X - )2
X – E(X) (X – E(x)2
11 1 1 – 13 = 2 4
12 12 – 13 = 1 1
13 13 – 13 = 0 0
14 14 – 13 = 1 1
15 15 – 13 = 2 4
n=5 10

x  = 11 + 12 + 13 + 14 + 15 = 65
5 5 = 13
Var (X) =  (X – E(x))2 =  (x - )2
Var (X) = 10
2 =5
δX = √2
δX = 1.4142

SELF-ASSESSMENT EXERCISE
Define standard deviation of a population

3.3 Estimation of Sample Statistics

As it has been said before now that, the term statistics is usually used in describing the
features of a sample. The basic statistic of a sample corresponding to the parameters of
the population are sample mean usually denoted by x̅, sample variance denoted by Sx2
and sample standard deviation denoted by Sx.
Sample mean is defined as the average value in the sample it is denoted by x̅. The sample
arithmetic mean is calculated by adding up the observation of the sample and then
dividing by the total number of observations.
̅=
X ∑n 𝑋
i=o

n
Sample variance as it has been said before now, it is a measure of dispersion of the value
of x in the sample around their average value. This is denoted as

Sx2 = ∑ni=o(x − x̅)2 = x2 - nx̅ 2 = x2 - x̅2

n n n
The sample standard deviation is denoted by Sx this is taken to be the square root of the
variance.
Sx = √ (x − x̅)2
n
Covariance; this statistics usually involves two variable. The covariance is defined as the
sum of the product of the deviation of variable x and y from the various means.
̅
COV(XY) = ∑𝑛𝑖=1(𝑥 − 𝑥̅ ) ( y – y)
n
Question
From the information of population supplied in the preceding subsection i.e. 11, 12, 13,
14, 15
Table M1.1.3
Sample Statistics Table of Analysis
Possible samples x̅ = mean of each sample
(11,12) 11+12
= 11.5
2

(11,13) 11+13
= 12
2

(11,14) 11+14
= 12.5
2

(11,15) 11+15
= 13
2

(12,13) 12+13
= 12.5
2

(12,14) 12+14
= 13
2

(12, 15) 12+15

= 13.5
2

(13,14) 13+14
= 13.5
2

(13,15) 13+15
= 14
2

(14,15) 14+15
= 14.5
2

n = 10

Sample mean = 11.5+12+12.5+13+12.5+13+13.5+13.5+14+14.5

10
130
= = 13
10

All the information about the population and possible samples can be summarize in a
frequency distribution as depicted in table 1.3 below.
Table M1.1.14
Table of Possible Samples
X F
11 1
11.5 1
12 1
12.5 2
13 2
13.5 2
14 1
14.5 1
15 1

Variance of sample mean = ∑ni=1(x − x̅)2

N
Sx = (11-13) + (11.5-13) + (12-13)2 + 2(12.5-13)2 + 2 (13-13)2 + 2 (13.5-13)2 +
2 2 2

(14-13)2 + (14.5-13)2
9
+ (15-13)2
9
Sx2 = (-2)2+(1.5)2+(-1)2+2(-0.5)2+2(0)2+2(0.5)2+(1)2+(2)2+(1.5)2
9
Sx2 = 4 + 2.25 + 1 + (0.25)2 + 2(0) + 2 (0.25) + 1 + 4 + 2.25
9
Sx2 = 4 + 2.25 + 1 + 0.5 + 0 + 0.5 + 1 + 4 + 2.25
9
Sx2 = 15.5
9
Sx2 = 1.722
Sx2 ≅ 2
Sx = √1.722
Sx = 1.31233
From the foregoing analysis it would be observed that given X1, X2 ……Xn of any
random sample of size n from any infinite population with population mean u and σ 2 then
1
with sample mean x̅ = ∑ni=1 x we have
n

(i) E(x̅) = 
𝜎2
(ii) Var (x) =
n

SELF-ASSESSMENT EXERCISE
Define the sample variance of any given population

3.4 Estimators for Mean and Variance

Given that X1, X2, X3 …. Xn is a random sample of size n from normal population with
1
mean  and variance σ2 i.e. (X  N (, σ2), then the statistics x̅ = ∑ni=1 x
n
x̅− 
Therefore Z =  N (0,1)
σ

√n
This is a general case whereby sampling is specifically taken from a normal distribution.

Worked Example
Given a random sample of 20 taken from a normal distribution with mean 90 and
variance 25 find the probability that the mean is greater than 101.
Solution
x̅  20 (90, 25)
x̅− 
Z=  N (0,1)
σ

√n
= 101 – 90
25
√20
= 11
25
√20
= 11
25
4.47213
11
=
5.590169

= 1.967739
≅ 1.968

SELF-ASSESSMENT EXERCISE
What are the assumptions of a normal distribution?

3.5. The Role and Significant of Statistics in Social Sciences

It is interesting to know that accuracy, validity, reliability, objectivity, analysis, efficiency

are all characteristics of the roles expected of statistical research in decision making and
policy formulation for societal development. Do you know that social statistics are
necessary in information gathering about socio-economics variables that are indices of
economic growth and development? It started with what is known as the “statists” social
research” and later grow to be known as “statistics”, a new term for quantitative
evidence. Social sciences’ statistics is very significant because it assist in quantifying
scientific developments and data on them therefore, making information on scientific
studies more concise and precise. Social statistics is usually conducted to prove
something for instance, how many women are affected by malaria compare to men in the
society? How many people in the society are able to afford living in a duplex, flat, one-
room apartment, face-to-face room or under the bridge? Consequently, it is significant to
note that adequate cautions are usually put into stepwise data gathering, accuracy, and
analysis for efficiency. The role of statistics in social sciences and its significant cannot
be overemphasized.

Self-Assessment Question

Do you think that statistics in social sciences has role to play in societal problem solving?

4.0 CONCLUSION
It has been established that given a random sample of X1, X2, …. Xn with population
mean  and standard variance r2.
(i) (x̅) = 
(ii) Var(x) = σ2/n

5.0 SUMMARY
In this unit, we have attempted the definition of population, sample, sample distribution
theory, so also estimation of parameter estimate and sample statistics had been attempted,
so also it has been proved from our calculation that the mean of sample must always
equal to the population mean it’s representing and that the variance of the population and
sample estimate are equal.
6.0 TUTOR MARKED ASSIGNMENT
Explain the descriptive measure of a sample statistics.

8.0 REFERENCE/FURTHER READING

- Adedayo, O.A. (2006): Understanding statistics. JAS publishers Akoka, Yaba.

- Dominick, S. and Derrick P. (2011): (Schaum outline series) Statistics and

Econometrics (second edition) MCGRAW HILL, New York.
- Edward, E.L. (1983): Methods of statistical analysis in Economics and Business.
HOUGHTON MIFFLIN COMPANY BOSTON.

- Esan F.O. and Okafor, R.O. (2010): Basis statistical method (revised edition)
Toniichristo Concept Lagos.

- Koutsoyianis, A. (2003): Theory of Econometrics (second edition). Palgrave

publishers Ltd (formerly Macmillan press Ltd), London and Basic stoke.

- Murray R. S. and Larry J. S. (1998): (Schaum outlines series). Statistics (Third

edition) MCGRAW HILLS.

- Olufolabo, O.O. and Talabi, C.O (2002): Principles and practice of statistics.
HASFEM Nig Enterprises, Shomolu, Lagos.

- Oyesiku, O.K. and Omitogun, O. (1999). Statistics for social and management
sciences. Higher Education Books Publisher Lagos.
UNIT TWO; SAMPLING DISTRIBUTION OF PROPORTION
CONTENT
1.0 Introduction
2.0 Objectives
3.0 Main Content
3.1 Sampling Distribution of proportion defined Sampling Distribution of
Parameter Estimate
3.2 Standard Error
3.3 Sampling Distribution of differences and sum of means
4.0 Conclusion
5.0 Summary
6.0 Tutor-Marked Assignment
7.0 References/ Further Readings

1.0 INTRODUCTION
This unit is an extension of unit one of this module. In this unit we are going to look at
sampling distribution of proportion, sampling distribution of sum and difference and
standard error. Since this unit is an offshoot of the unit one of this module, most of the
statistical term used in unit one will be implied here.

2.0 OBJECTIVE
At the end of our discussion of this unit, you should be able to calculate:
- Sampling distribution of proportion
- Sampling distribution of sum
- Sampling distribution of difference and
- Standard error
3.0 MAIN CONTENT
3.1 Sampling Distribution of Proportion Defined
Samples are usually embedded in a population, each time attribute is sampled, the
concept of proportion is coming in. the estimation here is concentrating on the proportion
of the population that has a peculiar characteristics. This sampling distribution is like of
binomial distribution, where an event is divided into been a success represented with p or
been a failure represented with q or 1 – p.
Given an infinite population consisting of sample size n. The sampling distribution of
proportion is said to have a mean of np.
and variance
P (1−p) pq
var (p) = var (p) = =
n n

It is to be noted at this juncture that the sample proportion is also an unbiased estimator
of the population proportion i.e. (p) = P
Example
A coin is tossed 120 times, find the probability that head will appear between 45% and
55%.

Solution
From the above the prob(head) = ½ = p
Prob(not obtaining ahead) = ½ = q = 1 – p
45% of tosses = 45 x 120
100
= 54
55
While 55% of tosses gives x 120
100

= 66
Mean p = np = 120 x ½
= 60
0.25
S.D = √npq = x 120
120

= 0.00208333
0.25
S.D. √
120

= 0.4564
1 1
S.D. = √npq = √120 ( ) ( )
2 2

= √30
= 5.477225575
54−60 66−60
Prob (54 < p < 78) = p ( <z< )
5.5 5.5
6 6
=p( <z< )
5.5 5.5

= p (- 1.0909 < z < 1.091) = (0.3621) x 2

= 0.7242

SELF-ASSESSMENT EXERCISE
What is the symbolic definition of standard deviation of a sample proportion?

3.2 Standard Error

Standard error usually represented by S.E. is defined as the square root of the population
variance written as √var p
pq p (1−p)
note var(p) = =
n n

( )
:. √p 1 − p
n
From the example in subsection 3.2 above
p=½=q
n = 120
 0.5 (0.5)
:. S.E = √
120
0.25
S.E = √
120
S.E = √0.0020833
S.E = 0.0456

SELF-ASSESSMENT EXERCISE
What does S.E stands for?

4.0 CONCLUSION
During the course of our discussion of this unit we have talked about;
- Sampling distribution of proportion
- Standard error

5.0 SUMMARY
In the course of our discussion we defined the mean of a sampling distribution of
proportion as np.
i.e. mean = np
variance (p) = P(1-P)
σ(p) = √npq

6.0 TUTOR MARKED ASSIGNMENT

A coin is tossed 90 times, find the probability that tail will appear between 35% and 55%.
7.0 References

7.0 REFERENCE/FURTHER READING

- Adedayo, O.A. (2006): Understanding statistics. JAS publishers Akoka Yaba.

- Esan, F.O. and Okafor, R.O. (2010): Basis statistical method (revised edition)
Toniichristo Concept, Lagos.
- Murray, R.S. and Larry, J. S. (1998): (Schaum outlines series). Statistics (Third
edition) MCGRAW HILLS.

- Olufolabo, O.O. and Talabi, C.O. (2002): Principles and practice of statistics.
HASFEM Nig Enterprises Shomolu Lagos.

- Oyesiku, O.K. and Omitogun, O. (1999): Statistics for social and management
sciences. Higher Education Books Publisher Lagos.
UNIT THREE; SAMPLING DISTRIBUTION OF SUM AND DIFFERENCE OF
TWO MEANS
CONTENT
1.0 Introduction
2.0 Objectives
3.0 Main Content
3.1 Sampling Distribution of difference and sum of two means defined
3.2 Worked example sampling distribution of sum of two means
3.3 Worked example sampling distribution of sample differences of two
means
4.0 Conclusion
5.0 Summary
6.0 Tutor-Marked Assignment
7.0 References/ Further Readings

1.0 INTRODUCTION
This unit is an extension of unit one and unit two of this module. In this unit we are going
to look at sampling distribution of sum and difference of two means. Since this unit is an
offshoot of the unit one of this module, most of the statistical term used in unit one will
be implied here.

2.0 OBJECTIVE
At the end of our discussion of this unit, you should be able to calculate:
- Sampling distribution of sum of two means
- Sampling distribution of difference and
3.0 MAIN CONTENT
3.1 Sampling Distribution of Difference of Two Means and Sum (𝐱̅ − 𝐱̅)
If two independent random sample of sizes n1 and n2 are selected from 2 different
population of size N1 and N2 with population means 1 and 2 respectively and
population variance σ12 and σ22 respectively, then the sampling distribution of the
difference of two means (x̅1 − x̅2 ) = p1 -p2 and standard deviation of the sample
distribution is written as
σx1 – x2 = σ12 + σ22
n1 n2

Also the sampling distribution of sum of means is as defined below:

p1 + p2 = p1 +p2 and the standard deviation
σ2p1 + p2 = σ12 + σ22
n1 n2

SELF-ASSESSMENT EXERCISE
Sampling distribution of the difference of two means is defined as-------

3.2 Worked Example of Sampling Distribution of Sum of Two Means

Given that p1 = (30,50) and p2 = (40,70) show that p1 +p2 = p1 + p2;
(ii) p1- p2 = p1 - p2 and (iii) σ2p1 + p2 = σ2p1 + σ2p2 for a sample drawn from each other.
Solution
Sampling sum
Possible sample combination = (30, 40), (30,70) (50,40) (50,70)
Sample sum = 30 + 40 = 70; 30 + 70 = 100; 50 + 40 = 90; 50 + 70 = 120
:. p1 + p2 = 70 + 100 + 90 + 120
4
380
p1 + p2 =
4

p1 + p2 = 95
Considering the 1st population p2 (30,50)
30+50
p1 =
2
80
p1 =
2

p1 = 40
Considering the 2nd population (40, 70)
40+70
p2 =
2
110
p2 = = 55
2

:. p1 + p2 = 55 + 40
p1 + p2 = 95
Note p1 + p2 = 95
p1 + p2 = 95
:. p1 + p2 = p1 + p2

SELF-ASSESSMENT EXERCISE
What is the population and sample mean of P1 = (70,90), P2 = (60,80)

3.3 Worked Example of Sample Differences of Two Means

p1 - p2 = 40 – 55 from our calculation of means above
p1 - p2 = 15
Taking the differences of possible =sample p1 – p2
p1 - p2 = (30-40) + (30-70) + (50-40) + (50-70)
4
p1 – p2 = 10 + -40 + 10 – 20
4

p1 – p2 = -10 – 40 + 10 – 20
4
 − 60
p1 – p2 =
4

p1 – p2 = - 15
:. p1 – p2 = p1 - p2
- 15 = - 15

(iii) σ2p1 + p2 = variance of 70,10, 90 & 120

Note population mean = 95
σ2p1 + p2 = (x -x̅)2
n
:. σ2p1 + p2 = (70-95)2 + (100-95)2 + (90-95)2 + (120-95)2
4
σ2p1+ p2 = -252 + 52 + -52 + 252
4
σ2p1 + p2 = 625 + 25 + 25 + 625
4
1300
σ2p1 + p2 =
4
2
σ p1 + p2 = 325
Considering the population independently
σ2p1 = variance of (30,50)
σ2p1 = (30 - 40)2 + (50 – 40)2
2
Where 40 = p1 = mean of population 1
σ2p1 = (-10)2 + (10)2
2
σ2p1 = 100 + 100
2
200
σ2p1 =
2

= 100
Considering the 2nd population
σ2p2 = variance of (40,70)
σ2p2 = (40 – 55)2 + (70 – 55)2
2
Where 55 = mean of population = p2
σ2p2 = 152 + 152
2
σ2p2 = 2252 + 2252
2
σ2p2 = 450
2
σ2p2 = 225
σ2p1 + σp2 = 225 + 100 = 325
σ2p1 + p2 = 325

SELF-ASSESSMENT EXERCISE
Sampling distribution of the difference of 2 mean x̅1 & x̅2 is usually written as?

4.0 CONCLUSION
In the course of our discussion of this unit you have learnt about
- Sampling distribution of difference of two means
- Sampling distribution of sum of two means

5.0 SUMMARY
In the course of our discussion on this unit we defined sampling distribution of the
difference of two mean as p1 - p2 and standard deviation of the difference as

rx1 – r2 = r12 + r22

n1 n2
6.0 TUTOR MARKED ASSIGNMENT
Given the following population p1 = (10,20) p2 = (30,40) show that
(i) p1 + p2 = p1 + p2
(ii) p1 – p1 = p1 - p2.

7.0 REFERENCE/FURTHER READING

- Adedayo, O.A. (2006): Understanding statistics. JAS publishers, Akoka, Yaba.

- Esan F.O. and Okafor, R.O. (2010): Basic statistical method (revised edition)
Toniichristo Concept, Lagos.

- Murray, R. S. and Larry J. S. (1998): (Schaum outlines series). Statistics (Third

edition) MCGRAW HILLS.

- Olufolabo, O.O. and Talabi, C.O. (2002): Principles and practice of statistics.
HASFEM Nig Enterprises Shomolu, Lagos.

- Oyesiku, O.K. and Omitogun, O. (1999): Statistics for social and management
sciences. Higher Education Books Publishers, Lagos.
UNIT FOUR; PROBABILITY DISTRIBUTION
CONTENTS
1.0 Introduction
2.0 Objectives
3.0 Main Content
3.1 Probability defined
3.2 Probability distribution of a random variable (Binomial distribution)
3.3 Poisson distribution
3.4 Probability distribution of a continuous variable (normal distribution)
4.0 Conclusion
5.0 Summary
6.0 Tutor-Marked Assignment
7.0 References/ Further Readings

1.0 INTRODUCTION
Probability Defined
For thorough understanding of this unit, it is assumed that you must have familiarized
yourself with introductory statistics and unit one of this module. The main thrust of this
unit is to introduce to you the concept of probability distribution, its discussion,
calculation and interpretation of result. This unit is fundamental to the understanding of
subsequent modules. This is because other unit and module will be discussed on the basis
of the fundamentals concept explained here.

2.0 OBJECTIVES
At the end of this unit you should be able to understand the following:
i. Concept of probability
ii. Different probability distribution
iii. Calculate the different probability distribution
3.0 MAIN CONTENT
3.1 Probability Defined
Statisticians spends quality time measuring data and drawing conclusions based on his
measurement sometimes, all the data is available to the statisticians and the measurement
are bound to be accurate in such circumstances, it can be said that the statistician has
perfect knowledge of the population.
There are a-times whereby this will not be the usual situation. In most cases, the
statistician will not have the details he wants about the population and will be unable to
collect the information he wants because of cost and labour involved.
However, because the entire population has not been examined, the statistician can never
be completely sure of the result, so when quoting conclusion based on sample evidence, it
is usual to state how confident the statistician is about his result. So you will often see
estimates quoted with 85% confidence. This is simply talking about the probability that
the estimate is right is 85%.
The probability of a value X of a random variable is usually referred to as the limiting
value of the relative frequency of that value as the total number of observation on the
variable approaches infinity, the value which the relative frequency assumes at the limit
as the number of observations tends to infinity. This can be written as
f
P(x) = limn
∑Fx

SELF-ASSESSMENT EXERCISE
What is another name that probability can be called?

3.2 Probability Distribution of a Random Variable

If a variable is discrete, if its value are distinct i.e. they are separated by finite distance.
To each we may assign a given probability. If x is a discrete random variable which may
assume the values X1, X2 …..Xn with respective probabilities f(x1), f(x2) ……, f(xn). Then
the entire set of pairs of permissible value together with their respective probabilities is
called probability distribution of a random variable x.
A random variable is a variable whose values are associated with the probability of being
observed. A discrete random variable is one that can assume only finite and distinct
value.
One of the discrete probability is the binomial distribution. This distribution is used to
find the probability of X number of occurrences or success of an event, P(x) in n-trials of
same experiment.
Binomial distribution is usually use to predict occurrence of events that are mutually
exclusive in other words Binomial distribution is useful for problem that are concerned
with determining the number of times an event is likely to occur or not occur during a
given number of trials and consequently the probability of it occurring or not occurring.
Symbolically it is written as;
P(x) = nCx Pxqn-x

Alternatively
𝑛!
P(x) = px (1-p)n-x
𝑋!(𝑛−𝑥)

Where P = probability of a success in a simple trial probability of one event

q = 1-P, probability of the alternative to the event (failure)
n = number of times the event can occur in number trials
x = number of successes in n-trials
Mean of the binomial distribution is  = np and standard deviation is
σ = √np (1 − p) or
σ = √npq
Example: What is the probability of obtaining 3 heads in 5 toss of a balanced coin. (b)
What is the probability of obtaining less than 3 heads in 5 toss of coin.

Solution
Probability of obtaining a head = 1/2 = p
Probability of not obtaining ahead = 1-p = q = ½
X = 3, n = 5
𝑛!
(a) P(x) = p.x (1-p)n-x
𝑋!(𝑛−𝑥)!
𝑛!
= p.x qn-x
𝑋!(𝑛−𝑥)!

P(x) = 5x4x3x2x1 ½.3 ½5-3

3x2x1 (5-3)!
P(x) = 5x4x3x2x1 ½.3 ½2
3x2x1 (5-3)!
10 1 1
P(x) = x x
1 8 4
10
P(x) =
32

P(x) = 0.3125
:. The probability of obtaining 3 heads from 5 tosses of coin = 0.3125

(b) Probability of obtaining less than 3 heads = P(0) + P(1) + P(2)

5!
:. P(0) = ½0 . ½ 5-0
0!(5−0)!
5𝑥4𝑥3𝑥2𝑥1 1
= .1 .
5𝑥4𝑥3𝑥2𝑥1 32

= 1/32
= 0.03125

1 5-1
P(1) = 5! 1 1
1! (5-1)! 2 2
5𝑥4𝑥3𝑥2𝑥1
= ½ . ½4
4𝑥3𝑥2𝑥1𝑥1
5 1
= x½x
1 16
5
= = 0.15625
32

P(2) = 5! 12 1 5-2

2! (5-2)! 2 2
5𝑥4𝑥3𝑥2𝑥1 1
= ¼.
2𝑥3𝑥2𝑥1𝑥1 8
 10 1
P(2) = x
1 32
10
P(2) =
32

P(2) = 0.3125
P(<3) = P(0) + P(1) + P(2)
P (<3) = 0.03125 + 0.15625 + 0.3125
P (<3) = 0.49625

Mean =  = np = 5 (½)
np = 5/2
np = 2.5 heads
Standard deviation = σ = √npq

1 1
σ= √5 ( ) ( )
2 2

5
σ= √
4

σ= √1.25
S.D. = σ = 1.1180339887499
= 1.12 heads

SELF-ASSESSMENT EXERCISE
What do you understand by the word a random variable?

3.3 Poisson Distribution

Poisson distribution is another discrete probability distribution useful in describing the
number of events that will occur in a specific period of time. It is usually used in
determining the probability of a designated number of successes per unit of time. When
the event or successes are independent and the average number of successes per unit of
time remains constant. Symbolically it is written as;
P(x) = x e-
 X!
Where P(x) = probability of x number of successes
X = number of success (0,1,2 ….)
 = average or mean number of success or event that occur in a given internal
e = natural logarithms base whose value equal 2.71828
note  = mean & variance of poisson distribution
σ = √
Example
A study shows that an average number of 6 customers per hour stop for fueling at a
filling station.
(a) What is the probability of 3 customers fuelling at any hour?
(b) What is the probability of less than 3 customers, fueling in any hour?
Solution
(a) note mean = variance =  = 6
𝑒 = 2.71828
x =3
p(x) = x e-
x!
P(x=3) = 63.(2.71828)-6
3x2x1
P(x=3) = 216 x 0.00248
6
P(x=3) = 0.53568
6
P (x=3) = 0.08928
(b) P(x < 3) = Prob(0) + prob(1) + prob(2)
P(x=0) = 6o x 2.71828-6
0!
 P(x=0) = 1 x 0.00248
1
Note 0! = 1
P(x=0) = 0.00248

P(x=1) = 61 x 2.71828-6
1!
P(x =1) = 6 x 0.00248
1
P(x=1) = 0.01488
P(x=2) = 62 x 2.71828-6
2!
P(x =2) = 36 x 0.00248
2x1
P(x=2) = 18 x 0.00248
P(x=2) = 0.04464

:. Prob (x>3) = P(0) + P(1) + P(2)

P (x<3) = 0.00248 + 0.01488 + 0.04464
P (x<3) = 0.062
Mean = variance = 6 =
S.D=δ. = √6
S.D.=δ = 2.449489743

SELF-ASSESSMENT EXERCISE
Define standard deviation of Poisson distribution?
3.4 Probability Distribution of a Continuous Variable (Normal Distribution)
If a variable is continuous, it can assume an infinite number of values within a given
interval. An important feature of probability distribution is that the areas under these
curve represents probabilities. The total area under the curve of a probability distribution,
being the sum of individual probabilities is equal to unity (1).
The normal distribution as a continuous probability distribution and the most commonly
used distribution in statistical analysis. The normal curve is bell-shaped and symmetrical
about its mean. Usually, it extends indefinitely in both directions, but most of the area
(probability) is clustered around the mean.
To find the probabilities for problems involving the normal distribution, first convert the
x value into corresponding z value using
X− 
Z=
σ

Where  = mean value

σ = standard deviation
Then check up the corresponding value from the normal distribution table.

Example
Given that family incomes are normally distributed with  = N14,000 and δ = 4000.
What is the probability that a family picked a random will have;
(a) Between N13,000 and N16,000 ?
(b) Below N13,000?
(c) Above N16,000 ?
(d) Above N18,000?
X− 
Z=
σ

(a) Here x = 13,000 & 16,000

When x = 13,000; Z1 = 13,000 – 14,000
4,000

Z1 = - 1,000
 4,000
Z1 = - 0.25
When X = 16,000; Z2 = 16,000 - 14,000
4,000
= 2000
4000
= 0.5
Z1 = 0.25 ; Z2 = 0.5
ZT1 = 0.0987 ; ZT2 = 0.1915
Where ZT1 and ZT2 represents the table value for Z1 and Z2
:. Prob (13,000  x  16,000) = 0.0987 + 0.1915
:. Prob (13,000  x  1,6000) = 0.2902
= 29%
(b) Prob (x < 13,000) = 0.5 – 0.0987
= 0.4013
≅ 40%
(c) Prob (x > 16,000) = 0.5 – 0.1915
= 0.3085
≅ 30.85%
(d) Prob (x > 18,000)
(x = 18,000)
Z = 18,000 – 14,000
4,000
Z = 40,000
4000
Z =1
ZT = 0.3413
:. Prob (x > 18,000) = 0.5 – 0.3413
= 0.1587
 Prob (x > 18,000) = 15.8%
≅ 16%

SELF-ASSESSMENT EXERCISE
Explain the attributes of a normal distribution curve

4.0 CONCLUSION
From our discussion so far you have learnt about:
- Probability
- Probability distribution
- Different probability distribution, the binomial, Poisson, and normal distribution.

5.0 SUMMARY
In the course of our discussion of this unit, we have defined the different probability
distributions binomial distribution is defined as
P(x) = nCx Px qn-x
Alternatively
𝑛!
P(x) = Px (1-p)n-x
𝑋!(𝑛−𝑥)!

Where P = Probability of success

q = 1 – P = probability of failure
mean = np, S.D = σ = √npq
Poisson distribution is defined as
P(x) = x e-ʎ
x!
ʎ = mean = variance

√ = standard deviation
Normal distribution
 X− 
Z=
σ

6.0 TUTOR MARKED ASSIGNMENT

A study shows that 40% of the people entering a supermarket make a purchase. Using
(a) binomial distribution, (b) Poisson distribution find the probability that out of 30
people entering the supermarket 10 or more will make a purchase.

7.0 REFERENCE/FURTHER READING

- Adedayo, A.O. (2006): Understanding Statistics. JAS Publishers, Lagos.

- Dominick, S. and Derrick, R. (2011): Statistics and Econometrics. (Schaum’s

outlines) McGraw Hill, New York.

- Esan, E.O. and Okafor, R.O. (2010): Basic Statistical Methods (Revised
Edition) Tonichristo Concept.

- Koutsoyianis, A. (2003): Econometric Methods (second edition). Palgrave

publishers Ltd (formerly Macmillan press ltd), London and basin stoke.

- Murray, R. S. and Larry, J. S. (1998): Statistics (Schaum outlines). McGraw Hill.

- Olufolabo, O.O. and Talabi, C.O. (2002): Principles and practice of statistics,
HASFEM (NIG) ENTERPRISES, Somolu, Lagos.

- Oyesiku, O.K. and Omitogun, O. (1999): Statistics for social and management
sciences. Higher Education Book Publishers Lagos.

- Owen, F. and Jones R. (1983): Statistics. Polytech Publishers Ltd. Stockport.

MODULE TWO: Analysis of variance and analysis of covariance
Unit 1: One-way factor analysis of variance
Unit 2: Two-way factor analysis of variance
Unit 3: Analysis of covariance
UNIT ONE: ONE-WAY ANALYSIS OF VARIANCE
CONTENTS
1.0 Introduction
2.0 Objectives
3.0 Main Content
3.1 Logic of analysis of variance
3.2 Assumption and steps involved in analysis of variance
3.3 Computation
3.4 Worked example
4.0 Conclusion
5.0 Summary
6.0 Tutor-Marked Assignment
7.0 References/ Further Readings

1.0 INTRODUCTION
A detailed knowledge and understanding of introductory statistics is assumed, it is also
expected that students would have familiarized themselves with hypothesis testing. This
unit is one of the four units in module 2 of the course.

2.0 OBJECTIVE
At the end of this unit, you should be able to understand and be able to calculate:
 Total sum of square
 Sum of square between groups
 Sum of square within the group
 Mean square
3.0 LOGIC OF ANALYSIS OF VARIANCE (ANOVA)
Analysis of variance (Anova) is usually used to test null hypothesis that the means of two
or more populations are equal versus the alternative that at least one of the means is
different. The null hypothesis (Ho) tested in the case of ANOVA is that the means of the
population from which the sample is drawn are all equal i.e. Ho, 1 = 2 = 3 = ……… =
n while the alternative hypothesis says that Ho taken as a whole is not true i.e. H1; 1≠
2 ≠ 3.
It is to be noted that each time ANOVA is used, all we are trying to do is to analyze or
test the variances in order to test the null hypothesis about the means (i.e. H o; 1 = 2 =
3). The ANOVA procedure is based on mathematical theory that the independent sample
data can be made to yield two independent estimate of the population variance namely;
(i) Within group variance (or error) this is variance estimate which deals with how
different each of the values in a given sample is from other values in the same
group.
(ii) Between group variance this is estimate that deals with how the means of the
various samples differs from each other.

SELF-ASSESSMENT EXERCISE
State the null hypothesis of analysis of variance?

3.1 Assumptions of Anova

(i) Observations are independent and value of any of observation should not be
related to the value of another observation.
(ii) Homogeneity of sample variance, it should be assumed that the variance are equal
for all treatment populations.
(iii) The values in the population are normally distributed.
SELF-ASSESSMENT EXERCISE
State one assumption of analysis of variance

3.2 Steps involved in Anova Analysis

(i) Estimate the population variance from the variance between sample means (MSA)
(ii) Estimate the population variance from the variance within the samples (MSE)
(iii) Compute the fisher ratio. This is given as F = MSA
MSE
i.e. F = Variance of between the sample mean
Variance of within the sample
(iv) Compute the various degree of freedom i.e. the degree of freedom for between,
within and total groups.
Degree of freedom for the sum between group is given as C – 1
Degree of freedom within group is writer as (r – 1) c
Total degree of freedom as r – 1
Where c = no of samples
R = no of observations
(v) The next thing is to obtain the critical value of F statistics using the F-table in the
table, we have the horizontal row which is for degree of freedom of the sum
between group numerator. While, the vertical column is meant for within group,
check the between degree of freedom along the horizontal axis and within group
along vertical axis. This can be checked at either at 0.05 (5%) level of significance
or 0.01(1%) level of significance.
(vi) Compare the F- statistic value with the critical value if the calculated value is less
than the tabulated value, accept the null hypothesis (Ho) and concluded that the
difference is not significant. If the calculated value is greater the critical value
reject Ho and accept Hi the alternative hypothesis and conclude that the difference
is significant.
(vii) The result is expected to be summarized on an ANOVA table.
Table M2.1.1
Analysis of Variance Table
Sources of variation Sum of squares Degree of Mean square I ratio
freedom
Between the means 𝑆𝑆𝐴 = 𝑟 (𝑥̅𝑗 − 𝑥̿ )2 C–1 MSA = SSA MSA
(examples by Factor C–1 MSE
A)
Within the sample 𝑆𝑆𝐸 =  (𝑥𝑖𝑗 − ̅̅̅̅)
𝑥𝑖𝑗
2 (r – 1)c MSE = SSE -
(error or (r-1)c
unexplained)
Total 2
𝑆𝑆𝑇 =  (𝑥𝑖𝑗 − x̿ ) = rc – 1 - -

SSA + SSE

𝑥𝑖𝑗
Where 𝑥̅𝑗 = mean of sample j composed of r observations =
r

𝑖 𝑗 𝑥𝑖𝑗
x̅ = 𝑔𝑟𝑎𝑛𝑑 𝑚𝑒𝑎𝑛 𝑜𝑓 𝑎𝑙𝑙 𝑐 𝑠𝑎𝑚𝑝𝑙𝑒𝑠 =
rc
SSA = Sum of square explained by factor A = 𝑟 (x̅ − 𝑥̿ )2
SSE = Sum of square of error unexplained by factor A =  (𝑥𝑖𝑗 − 𝑥 ̅ )2
2
SST = Total Sum of squares = SSA + SSE =  (𝑥𝑖𝑗 − 𝑥̿𝑗 )
Where c = no of samples
r = no of observations in each sample

SELF-ASSESSMENT EXERCISE
State the fisher ratio
3.4 Worked Example `
The information below relates to quantities of plastic produced by a plastic industry in 3
sections (morning, afternoon and evening) for 5 weeks. The production data are normally
distributed with equal variance.

Table M2.1.2
Table showing production of a plastic industry
Weeks Morning (X1) Afternoon (X2) Evening (X3)
1 85 77 90
2 83 81 92
3 79 75 84
4 81 82 82
5 82 80 87
Is there any significant difference due to production session?
Test at 5% level of significance.

Solution
Ho; 1 = 2 = 3
Hi; 1 ≠ 2 ≠ 3
Note let the quantities produced in morning be represented by X1, afternoon X2, evening
X3 .
X1 = 410
𝑥̅1 = X1 = 410 = 82
r 5
where r = number of weeks
X2 = 395
𝑥2 = X2 = 395 = 79  79
̅̅̅
r 5
X3 = 435
𝑥3 = X3 = 435 = 435 = 87  87
̅̅̅
r r 5
410 + 395 + 435
̅
x =
(5)(3)
= 1240 = 82.66667 = 82.67
15
 83
SSA = 5[(82 – 82.67)2 + (79 – 82.67)2 + (87 – 82.67)2]
= 5[(-0.67)2 + (-3.67)2 + (4.33)2]
= 5(0.4489 + 13.4689 + 18.7489)
= 5(32.667)
= 163.3335
2
SSE =  (𝑥𝑖𝑗 − 𝑥̅𝑗 )
= (85 – 82)2 + (83 – 82)2 + (79 – 82)2 + (81 – 82)2 + (82 – 82)2 +(77 – 79)2
+ (81 – 79)2 + (75 – 79)2 + (82 – 79)2 + (80 – 79)2 + (90 – 87)2 +(92 – 87)2
+ (84 – 87)2 + (82 – 87)2 + (87 – 87)2
= (3)2 + (1)2 + (-3)2 + (-1)2 + 02 +(-2)2 +(2)2 + (-4)2 (3)2 + (1)2 + (3)2 + (5)2 +(-
3)2 + (-5)2 + 0
= 9 + 1 + 9 + 1 + 0 + 4 + 4 + 16 + 9 + 1 + 9 + 25 + 9 + 25 + 0
= 122

SST = (85 – 82.67)2 + (83 – 82.67)2 + (79 – 82.67)2 + (82 – 82.67)2 +

(77 – 82.67)2 + (81 – 82.67)2 + (75 – 82.67)2 + (82 – 82.67)2 +
(80 – 82.67)2 + (90 – 82.67)2 + (92 – 82.67)2 + (84 – 82.67)2 +
(82 – 82.67)2 + (87 – 82.67)2
 = (2.33)2 + (0.33)2 + (-3.67)2 + (1.67)2 + (0.67)2 + (-5.67)2 + (1.67)2 + (-7.67)2
+ (0.67)2 + (2.67)2 + (7.33)2 +(9.33)2 + (1.33)2 + (0.67)2 + (4.33)2
= 5.4289 + 0.1089 + 13.4689 + 2.7889 + 0.4489 + 32.1489 + 58.8289 +
2.7889 + 0.4489 + 7.1289 + 53.7289 + 87.0489 + 1.7689 + 0.4489 +
18.7489
= 285.3335

Table M2.1.3
One-Way Analysis of Variance Table
Sources of variation Sum of squares Degree of Mean square I ratio
freedom
Explained variation 𝑆𝑆𝐴 = 163.3335 3-1 =2 MSA = 163.335
(between column) 2
= 81.66675 81.66675
Unexplained 10.167
variation or error 𝑆𝑆𝐸 = 122 (5 – 1)3 = (4)3 MSE = 122 = 8.0325
(within column) = 12 12
= 10.167
Total 285.3335 rc – 1 = 14 -

Note Sum of Square

𝑆𝑆𝐴 = 𝑟 (𝑥̅𝑗 − 𝑥̿ )2
2
𝑆𝑆𝐸 =  ( 𝑥𝑖𝑗 − ̅̅̅̅
𝑥𝑖𝑗 )
2
𝑆𝑆𝑇 =  (𝑥𝑖𝑗 − x̿ )

Degree of Freedom
Explained variation = c – 1 Where c = number of samples
Unexplained variation = (r – 1) c r = number of weeks
Total variation = rc – 1
MEAN SQUARE
MSA = SSA
c–1
MSE = SSE
(r-1)c
F-ratio = MSA
MSE
F0.05(2,12) = 3.88 (Critical value)
Source: F distribution table

Decision
Accept Hi, reject Ho because Fcal > Ftab which implies that there is significant difference
between the mean of production sessions.
Self-assessment exercise
State the formulae for sum of square?

4.0 CONCLUSION
In the course of our study of one-way analysis of variance you must have learnt about;
 Explained variation
 Unexplained variation
 Total variation

5.0 SUMMARY
In the course of our discussion of one-way analysis of variation the following definitions
were inferred
𝑆𝑆𝐴 = 𝑟 (𝑥̅𝑗 − 𝑥̿ )2
 2
𝑆𝑆𝐸 =  ( 𝑥𝑖𝑗 − ̅̅̅̅
𝑥𝑖𝑗 )
2
𝑆𝑆𝐽 =  (𝑥𝑖𝑗 − x̿ )
6.0 TUTOR MARKED ASSIGNMENT
Submit a one page essay on the definition of degree of freedom for explained variation,
unexplained variation and total variation.

7.0 REFERENCES/FURTHER READIDINGS

- Adedayo, O. A. (2006): Understanding Statistics: JAS Publishers, Akoka, Lagos.

- Dominick, S. and Derrick, R. (2011): Statistics and Econometrics, (Schaum

Outlines) McGraw-Hill Company, New York.

- Edward, E.L. (1983): Statistical analysis in Economics and Business. Houghton

Mufflin Company, Boston.

- Olufolabo, O.O. and Talabi, C.O. (2002): Principles and Practice of Statistics;
HAS-FEM ENTERPRISES Somolu, Lagos.

- Owen, F. and Jones, R. (1978): Statistics, Polytech Publishers Ltd, Stockport.

UNIT TWO: TWO-WAY ANALYSIS OF VARIANCE
CONTENTS
1.0 Introduction
2.0 Objectives
3.0 Main Content
3.1 Two way analysis of variance defined
3.2 Two-way classification
3.3 Computations
3.4 Worked example
4.0 Conclusion
5.0 Summary
6.0 Tutor-Marked Assignment
7.0 References/ Further Readings

1.0 INTRODUCTION
This unit is an extension of unit one, the difference between them is that, here, we
can test for two (2) null hypothesis, one for factor A and the other for factor B.

2.0 OBJECTIVE
At the end of this unit, you should be able to test for two null hypothesis.
Ho; Ua1 = Ua2 = Ua3
Ho; Ub1≠ Ub2 ≠ Ub3

3.0 MAIN CONTENT

3.1 Two- Way Analysis of Variance Defined
For two way analysis, the set of observation involved are classified into two (2)
factors or criteria; treatment factor or criteria and block or homogenous factor or
criteria.
 As we have discussed in one factor- analysis of variance, the total variation is
divided or splitted into 3 components.
 Variation between treatment
 Variation between blocks and
 Residual or error variation

SELF-ASSESSMENT EXERCISE
State the divisions into which total variation is divided into?

3.2 Two-way Classification

Table M2.2.1
Two-way classification table
Treatment (Factor A)
1 2 3 …………… t Total
Block factor 1 Y11 Y12 Y13 ………………… Y1j B1
B
2 Y21 Y22 Y23 ………………… Y2t B2
3
4
5
“
“
“
“
B Yb1 Yb2 Yb3 …………………. Ybt Bb

3.3 The Formulars

𝑥𝑖𝑗
(i) Column means is given by
r
𝑥𝑖𝑗
Row means of given
c

Grand mean is given by x̿ =  ̅𝑖 =  ̅𝑖

𝑥. 𝑥
r c

The subscripted dot signifies that more than one factor is under consideration.
 2
SST =  (𝑥𝑖𝑗 − 𝑥̿𝑗 )
SSA = 𝑟 (xij
̅̅̅ − 𝑥̿ )2 between column variation
SSB = 𝑐 (𝑥̅𝑖 − x̿ )2 between row variation
SSE = SST – SSA – SSB
Degree of freedom of SSA = c – 1
Degree of freedom of SSB = r – 1
Degree of freedom of SSE = (r-1) (c – 1)
Degree of freedom of SST = rc – 1
Mean Square
MSA = SSA
c–1
MSB = SSB
r–1
MSE = SSE
(r-1)(c-1)
F- statistics
F-ratio for factor A = MSA
MSE
F-ratio for factor B = MSB
MSE
It is to be noted that; two (2) separate null hypothesis is considered.
(i) Ho; There is no difference between mean of treatment
(ii) Ho; There is no difference between mean of block.

SELF-ASSESSMENT EXERCISE
State the formulae for column mean?
3.4 Worked Example
Samples taken involving two (2) interactive factors A & B in a two analysis of
variance experience gives the result below:
TableM2.2.2
Table showing interactive factors A and B
Treatment A
Block (B) 22 11 10 5
13 10 8 6
7 9 6 2

You are carry out a 2-way analysis of variance at 0.05 level of significance?

Solution
Hypothesis
1. Ho; 1 = 2 3 = 4; H1; 1 ≠ 2 = 3 = 4
2. Ho; 1 = 2 = 3; H1; 1 ≠ 2 ≠ 3
Table M2.2.3
Two-Way Classification Table
Treatment A Total Sample
mean
Block B 22 11 10 5 48 𝑥̅ 1 = 12
13 10 8 6 37 𝑥̅ 2 = 9.25
7 9 6 1 23 𝑥̅ 3 = 5.75
Total 42 30 24 12 108 𝑥̅ i = 27
Sample 42/3 30/3 24/3 12/3 𝑥̿ = 9
mean x.1 = 14 x.2 = 10 x.3 = 8 x.4 = 4

2
SST =  (𝑥𝑖𝑗 − 𝑥̿𝑗 )
(22 – 9)2 = (13)2 = 169; (11 – 9)2 = (2)2 = 4; (10 – 9)2 = (1)2 = 1
(13 – 9)2 = (4)2 = 16; (10 – 9)2 = (1)2 = 1; (8 – 9)2 = (-1)2 = 1
(7 – 9)2 = (-2)2 = 4; (9 – 9)2 = (0)2 = 0; (6 – 9)2 = (-3)2 = 9
= 189 =5 = 11
(5 – 9)2 = (-4)2 = 16;
(6 – 9)2 = (-3)2 = 9;
(1 – 9)2 = (-8)2 = 64
= 89

:. SST = 189 + 5 + 11 + 89 = 294

SSA = 𝑟 (x̅ − 𝑥̿ )2 where r = no of column
= 3 [(14 – 9)2 + (10 – 9)2 + (8 – 9)2 + (4-9)2]
= 3 [52 + (1)2 + (-1)2 + (-5)2
= 3 (25 + 1 + 1 + 25)
= 3 (52)
= 156

SSB = 𝑐 (𝑥̅𝑖 − x̿ )2 Where c = number of row

= 4[(12 – 9)2 + (9.25 – 9)2 + (5.75 – 9)2]
= 4 [(3)2 + (0.25)2 + (-3.25)2]
= 4 (9 + 0.0625 + 10.5625)
= 4 (19.625)
= 78.5

SSE = SST – SSA – SSB

= 294-156 – 78.5
= 59.5

Degree of Freedom
SSA = c – 1 = 4 -1 = 3
 SSB = r – 1 = 3 – 1 = 2
SSE = (r-1) (c – 1) = (3 – 1) (4 – 1) = (2) 3 = 6
SST = rc – 1 = (4 x 3) – 1 = 12 – 1 = 11
Mean Square
MSA = SSA = 156 = 156 = 52
c–1 4 -1 3
MSB = SSB = 78.5 = 78.5 = 39.25
r–1 3-1 2
MSE = SSE = 59.5 = 59.5 = 59.5 = 9.916666667
(r-1)(c-1) (3-1)(4-1) (2)(3) 6

F-ratio
MSA = 52 = 5.243697303
MSE 9.916667
MSB = 39.25 = 3.95798318
MSE 9.9166667
Table M2.2.4
Two-ways / Two Factor Analysis of Variance
Sources of variation Sum of squares Degree of Mean square E ratio
freedom
Explained variation
by factor A (between 𝑆𝑆𝐴 = 156 C–1=3 MSA = 5.24370
column) MSA = 52 MSE
Explained variation
by factor B (between 𝑆𝑆𝐵 = 78.5 r–1=2 MSB = 3.95798
rows) MSB = 39.25 MSE
Unexplained variation 𝑆𝑆𝐸 = 59.5 MSE= -
or error (r – 1)(c-1) = 6 9.91667
Total 294 11 - -

Decision Criteria Test 1

1. Factor A Critical Value
 F3,6 = 4.76
Because Fcal. > Ftab. Reject Ho and accept H1 meaning that the mean of factor A are
not equal.
Test II
2. Factor B Critical Value
F2,6 = 5.14
Since Fcal. < Ftab. Accept Ho and reject H1 conclude that the mean of factor B are all
equal.

SELF-ASSESSMENT EXERCISE
State the decision criteria for accepting or rejecting hypothesis?

4.0 CONCLUSION
In the course of our discussion on two-way analysis of variance, we have learnt
about:
(i) Sum of square of Factor A
(ii) Sum of square of Factor B
(iii) Sum of square of the error term
(iv) Mean square of Factor A
(v) Mean square of Factor B
(vi) F-ratio of both Factor A and Factor B
(vii) Sum of Square of total variation.
5.0 SUMMARY
In our discussion the following definition were inferred to:
2
(i) SST =  (𝑥𝑖𝑗 − 𝑥̿𝑗 )
(ii) SSA = 𝑟 (xj
̅ − 𝑥̿ )2
(iii) SSB = 𝑐 (𝑥̅𝑖 − x̿ )2
(iv) SSE = SST – SSA – SSB
 (v) MSA = SSA
c–1
(vi) MSB = SSB
r-1
(vii) MSE = SSE
(r-1)(c-1)
(viii) F-ratio for
Factor A = MSA
MSE
Factor B = MSB
MSE

6.0 TUTOR MARKED ASSIGNMENT

Submit a one page essay on the definition of MSE, SST and F-ratio.

7.0 REFERENCES/FURTHER READING

- Adedayo, O.A. (2006): Understanding Statistics: JAS Publishers, Akoka, Lagos.

- Dominick, S. and Derrick, R. (2011): Statistics and Econometrics, (Schaum’s

Outlines) McGraw-Hill Company, New York.

- Edward, E. L. (1983): Statistical analysis in Economics and Business. Houghton

Mifflin Company. Boston.

- Olufolabo, O.O. and Talabi, C.O. (2002): Principles and Practice of Statistics;
HAS-FEM ENTERPRISES Somolu, Lagos.

- Owen, F. and Jones, R. (1978): Statistics, Polytech Publisher Ltd, Stockport.

UNIT THREE: ANALYSIS OF COVARIANCE
CONTENTS
1.0 Introduction
2.0 Objectives
3.0 Main Content
3.1 Analysis of covariance defined
3.2 Assumption of analysis of covariance
3.3 Estimation of analysis of covariance
3.4 Worked example
4.0 Conclusion
5.0 Summary
6.0 Tutor-Marked Assignment
7.0 References/ Further Readings

1.0 INTRODUCTION
In general, research is conducted for the purpose of explaining the effect of the
independent variable on the dependent variable, and the purpose of research design is to
provide a structure for the research. In the research design, the researcher identifies and
controls independent variable that can help to explain the observed variation in the
dependent variable which in turn reduces error variables (unexplained variation).
In addition to controlling and explaining variation through research design, it is also
possible to use statistical control to explain the variation in the dependent variable,
statistical control is usually used when experimental control is difficult, if not impossible,
can be achieved by measuring one or more variable in addition to the independent
variable of primary interest and by controlling the variation attributed to these variables
through statistical analysis rather than through research design. The analysis procedure
employed in this statistical control is analysis of covariance (ANCOVA).
2.0 OBJECTIVE
At the end of our discussion on analysis of covariance, you should be able to;
- Define analysis of variance
- Define covariate
- Define adjusted Yis
- Develop table of analysis of covariance
- Calculate the various terms that may be needed on the computation of ANCOVA
Table

3.0 MAIN CONTENT

3.1 Analysis of Variance Defined
Analysis of covariance is an extension of the one-way analysis of variance that added
quantitative variable (covariate) when used, it is assumed that their inclusion will reduce
the size of the error variance and thus increase the power of the design. Analysis of
covariance (ANCOVA) is a statistical test related to analysis of variance (ANOVA). It
tests whether there is a significant difference between groups after controlling for
variance explained by a covariate.
A covariate is a continuous variable that correlates with the dependent variable. This
means that you can, in effect, “partial out” a continuous variable and run an ANOVA on
the result.
This is one way that you can run a statistical test with both categorical and continuous
independent variables.
The purpose of analysis of covariance is to remove one or more unwanted factor or
variables in the analysis. A variable whose effect one wishes to eliminate by means of a
covariance analysis called a covariate sometimes called concomitant variable.

ANCOVA works by adjusting the total sum of square, group sum of squares and error
sum of square of the independent variable to remove the influence of the covariate.
ASSUMPTIONS OF ANALYSIS OF COVARIANCE
- Variance is normally distributed
- Variance is equal between group
- All measure are independent
- Relationship between dependent variable and the covariate as linear
- The relationship between the dependent variable and the covariate is the same for all
groups.
Self-assessment exercise
Why analysis of covariance?

3.2 Estimation of ANCOVA

Hypothesis for ANCOVA
- Ho and Hi need to be stated slightly different for an ANCOVA than a regular
ANOVA.
Ho: the group means are equal after controlling for the covariate
Hi : the group means are not equal after controlling for the covariate
Below are the lists of notation for the calculation of ANCOVA.
𝐧 𝐧 𝐧
̿)𝟐 = ∑ ∑ 𝐘 𝟐 − 𝐘 𝟐 / 𝐚𝐧
𝐒𝐲𝐲 = ∑ ∑(𝐘 − 𝐘
𝐢 =𝟏 𝐣=𝟏 𝐢−𝟏 𝐣= 𝟏

𝐧 𝐧
̿)𝟐
𝐒𝐱𝐱 = ∑ ∑(𝐗 − 𝐗
𝐢 =𝟏 𝐣= 𝟏
𝐧 𝐧

= ∑ ∑(𝐗 𝟐 − 𝐘𝐢𝐣𝟐 / 𝐚𝐧)

𝐢−𝟏 𝐣= 𝟏

𝐧 𝐧

𝐒𝐱𝐲 = ∑ ∑( 𝐗 𝐢𝐣 − 𝐗̿ ) (𝐘𝐢𝐣 − 𝐘
̿)
𝐢 =𝟏 𝐣=𝟏
 𝐧 𝐧

= ∑∑ 𝐗 𝐢𝐣 𝐘𝐢𝐣 − 𝐗 𝐢𝐣 𝐘𝐢𝐣 / 𝐚𝐧
𝐢 =𝟏 𝐣= 𝟏
𝐧 𝐧

= ∑∑ 𝐗 𝐢𝐣 𝐘𝐢𝐣 − 𝐗 𝐢𝐣 𝐘𝐢𝐣 / 𝐚𝐧
𝐢 =𝟏 𝐣= 𝟏
𝐧

𝐓𝐲𝐲 ̅𝐢 − 𝐘
= ∑( 𝐘 ̿𝐢 )𝟐
𝐢 =𝟏

= ∑𝐢 =𝟏 Yi2 - Y2
n an

̅𝐢 − 𝐗
𝐓𝐱𝐱 = ∑( 𝐗 ̿ .. )𝟐
𝐢 =𝟏

=  Xi2 - X2
n an

̅ 𝐢𝐣 − 𝐗
𝐓𝐱𝐲 = ∑( 𝐗 ̿ 𝐧 ) (𝐘
̅𝐢𝐣 − 𝐘
̿.. )
𝐢 =𝟏

=X
̅Y̅ − XY
an
𝐧 𝐧
̅) 𝟐
𝐄𝐲𝐲 = ∑ ∑( 𝐘𝐢𝐣 − 𝐘
𝐢 =𝟏 𝐣 =𝟏

= Syy – Tyy
𝐧 𝐧
̅ )𝟐
𝐄𝑿𝑿 = ∑ ∑( 𝐗 𝐢𝐣 − 𝑿
𝐢 =𝟏 𝐣 =𝟏

= SXX – TXX
𝐧 𝐧
̅ 𝐢 ) (𝐘𝐢𝐣 − 𝐘
𝐄𝐱𝐲 = ∑ ∑( 𝐗 𝐢𝐣 − 𝐗 ̅𝐢 )
𝐢 =𝟏 𝐣 =𝟏

= Sxy – Txy
S=T+E
 ̅ = mean of X
Where X
̿ = Grand mean of X
X
̿ = Grand mean of Y
Y
a = variable involved
n = no of observations
Where the symbols S,T and E are used to denote sum of square and cross product for
total, treatment and error respectively.

Table M2.3.1
Analysis of Covariance for a Single Factor Experiment with One Covariate
Source of Df Sum of square and Adjusted df Mean square
variation product Regression Y error (MSE)
X XY Y
Treatment a–1 Txx Txy Tyy
Error a (n-1) Exx Exy Eyy SSE = Eyy – (Exy)2 a(n-1)-1 SSE
Exx a(n-1)-1

Total (an-1) Sxx Sxy Syy SS1E = Syy – (Sxy)2 an-2

Sxx
Adjusted SS1E – SSE a-1 SS1E – SSE
Treatment a–1

Fo = Fstatistics = Exy2/Exx
MSE
Fc = (SS’E – SSE) / (a-1)
SSE / (a(n-1)-1)
Which is distribute as
Fa-1,a(n-1)-1
Decision criteria
Reject Ho if Fc > F1, a(n-1)-1

Worked Example
A soft drink distributor is studying the effectiveness of delivery methods. Three different
types of truck have been developed, and an experiment is performed in the company’s
laboratory. The variable of interest is the delivery time in minute (Y): however, delivery
time is also strongly related to the case volume delivered (X). Each truck is used four
times and the data below are obtainable.

Table M2.3.2
Table Showing Delivery Method of a Distributor
Truck Types
1 2 3
Y X Y X Y X
27 24 25 26 40 38
44 40 35 32 22 26
33 35 46 42 53 50
41 40 26 25 18 20
Y1 = 145 Y1 = 139 Y2 = 132 Y2 = 125 Y3 = 133 Y3 = 134

Solution
̅1 = 145 = 36.25
Y ̅1 = 139 = 34.75
X
4 4

̅2 = 132 = 33
Y ̅2 = 125 = 31.25
X
4 4
133 134
̅3 =
Y = 33.25 ̅3 =
X = 33.5
4 4

̿ = 139 + 125 + 134

X
 12 = 33.167
Ho = T1 = T2 = …. = Tn = 0
Hi = T1 ≠ T2 ≠ ….. ≠ Tn = 0
𝐧 𝐧

𝐒𝐲𝐲 = ∑ ∑( 𝐘𝐢𝐣 𝟐 − 𝐘 𝟐 /𝐚𝐧)

𝐢 =𝟏 𝐣 =𝟏

a=3
n=4
Syy = 272 + 442 + 332 + 412 + 252 + 352 + 462 + 262 + 402 + 222 + 532 + 182 – 4102 /3x4
= 729 + 1936 + 1089 + 1681 + 625 + 1225 + 2116 + 676 + 1600 + 484 + 2809 +324 -
(410)2 / 12
Syy = 15,294 – 168,100 /12
Syy = 15294 – 14,008.33
Syy = 1,285.6711

Sxx = 242 + 402 + 352 + 402 + 262 + 322 + 422 + 252 + 382 + 262 + 502 + 202 – (3982
/(3x4))
Sxx = 576 + 1600 + 1225 + 1600 + 676 + 1024 + 1764 + 625 + 1444 + 676 + 2500 + 400
– (158404/12)
Sxx = 14,110 – 13,200.333
Sxx = 909.6666711

𝐧 𝐧

𝐒𝐱𝐲 = ∑ ∑( 𝐗𝐘 − 𝐗𝐘/𝐚𝐧)
𝐢 =𝟏 𝐣 =𝟏

Sxy = (27x24) + (44x40) + (33x35) + (41x40) + (25x26) + (35x32) + (46x42) + (26x25)

+ (40x38) + (22x26) + (53x50) + (18x20) – ( (410) (398) /12)
Sxy = 648 +1760 + 1,155 + 1640 + 650 + 1120 + 1932 + 650 + 1520 + 572 + 2650 + 360
– (163180 /12)
Sxy = 14,657 – 163,180 /12
Sxy = 14,657 – 13,598.333
Sxy = 1,058.67
n
Yi (Y)2
Tyy =∑ −
a an
i =1

Tyy = 1452 + 1322 + 1332 - 4102

4 3x4

Tyy = 21,025 + 17,424 + 17,689 - 168,100

4 12
Tyy = 56,138 - 168100
4 12
Tyy = 14,034.5 – 14,008.33
Tyy = 26.1667
n
(X)2
Txx ̅𝟐 −
= ∑(X )
an
i =1

X 2 (X)2
Txx = −
n an
Txx = 1392 + 1252 + 1342 – 3982
4 3x4
Txx = 19,321 + 15,625 + 17,956 – 158,404
4 12
Txx = 52902 - 158,404
4 12
Txx = 13,225.5 - 13,200.333
Txx = 25.1667
𝐧
̅ 𝐢𝐣 𝐘
𝐓𝐱𝐲 = ∑ 𝐗 ̿ 𝐘
̅𝐢𝐣 – ( 𝐗 ̿)
𝐢 =𝟏
 n
XiY XY
Txy = ∑ −
n an
i =1

Txy = (145 x 139) + (132 x 125) + (133 x 134) – (410)(398)

4 12
Txy = 20,155 + 16,500 + 17,822 - 163,810
4 12
Txy = 54,477 - 163,810
4 12
Txy = 13,619.25 – 13598.333
Txy = 20.91667

Eyy = Syy – Tyy

Eyy = 1285.6667 – 26.1667
Eyy = 1259.5

Exx = Sxx – Txx

Exx = 909.667 - 25.1667
Exx = 884.5

Exy = Sxy – Txy

Exy = 1058.67 – 20.9167
Exy = 1037.753

SS1E = Syy- (Sxy)2

Sxx
SS1E = 1285.67 – (1,058.67)2
909.667
SS1E = 1285.67 – 1,120,782.169
909.667
SS1E = 1285.67 – 1,232.08
SS1E = 53.59038
SS1E ≅ 53.59
with (an – 2) df = 12 – 2 = 10df

SSE = Eyy = (Exy)2

Exx
SSE = 1259.5 – (1037.753)2
884.5
SSE = 1259.5 – 1,076,931.912
884.5
SSE = 1259.5 – 1217.560104
SSE = 41.939896
SSE = 41.94
with a (n-1)-1) df = 3(4-1) – 1
= 3(3) – 1
=9–1
= 8 d.f.
SS1E – SSE = 53.59 – 41.94
= 11.65
with a – 1 df = 3 – 1 = 2 .d.f.
All the above calculations can be summarized in an ANCOVA Table, as presented below
Table M2.3.3
Analysis of Covariance (ANCOVA) Table
Source of d.f Sum of square and product Adjusted d.f Mean
variation Regression Square
Error
X XY Y
Treatment (3-1) 25.1667 20.91667 26.1667
2
Error 3(4-1) 884.5 1029.753 1259.5 41.94 3(4-1)-1
9 8 5.2425

Total (12-1) 909.667 1058.65 1285.67 53.59 (12-2)

11 10
Adjusted 11.65 2 5.825
Treatment

Fstatistics = Fc = SS1E – SSE  (a-1) = 11.65/2

SSE  a (n-1) – 1 53.59/8
Fc = 5.825
6.69875
 Fc = 0.869565217
Fc = 0.9
Ftab = F2,8 = 4.446
From the above Fc > Ftab
reject Ho , accept Hi ,: the mean of the delivery time are not equal.
̂ of the regression can be compute from
The estimate B
̂ = Exy = 1037.7533
B
Exx

884.5
̂ = 1.1732265461
B
Test of hypothesis can be carried out on this too, by using the test statistic.
̂ =0
Ho: B
Fc = (Exy)2 /(Exx)
MSE
Fc = (1037.753)2 / 884.5
5.2425
Fc = 1,217.5594
5.2425
Fc = 232.2478588
F0.05,1,8 = 5.32

Decision
Since Fc > Ftab reject Ho and accept Hi, it simply implies that the exists a linear
relationship between the delivery time and volume delivered.
The adjusted treatment can be computed as;
̅1 - B
Adjusted Y1 = Y ̂(X
̅1 -X
̿)
̅2 - B
Y2 = Y ̂ (X
̅2 -X
̿)
̅3 - B
Y3 = Y ̂(X
̅3 -X
̿)
̿ = grand mean of Xiz = X
Where X ̅1 + X
̅2 + X
̅3 = X
̿
̅1,X
X ̅2, X
̅3 = the respective mean of x
̅1,Y
Y ̅2, Y
̅3 = respective mean of Y
̅1 – B
Adjusted Y1 = Y ̂ (X
̅1 - X
̿)
= 36.25 – (1.173265461) (34.75 – 33.167)
= 36.25 – 1.16422 (1.5833)
= 36.25 – 1.857631204
= 34.3923688
≅ 34.40
Adjusted ̅2 – B
Y2 = Y ̂ (X
̅2 - X
̿)
Y2 = 33 – 1.173265461 (31.25 – 33.167)
Y2 = 33 – (1.16422) (-1.917)
Y2 = 33 + 2.249149889
Y2 = 35.24914989
Y2 ≅ 35.249
Adjusted ̅3 – B
Y3 = Y ̂ (X
̅3 - X
̿)
Y3 = 33.25 – 1.173265461 (33.5 – 33.167)
Y3 = 33.25 – 1.173265461 (0.33)
Y3 = 33.25 – 0.387177602
Y3 = 32.8628224
Y3 ≅ 32.86
SELF-ASSESSMENT EXERCISE
Define 𝐒𝐲𝐲 ?

4.0 CONCLUSION
In the course of our discussion on analysis of covariance you have learnt about the
following:
- Definition of analysis of covariance
- Estimation of analysis of covariance
 - Computation of analysis of covariance table
- Adjustment of the dependent variables

5.0 SUMMARY
In the course of our discussion the following were inferred.
𝐧
(𝐄𝐘)𝟐
𝐒𝐲𝐲 = ∑ ∑(𝐘 − 𝟐
=  (𝐘 − 𝐘
̿)
𝐚𝐧
𝐢 =𝟏 𝐣=𝟏
𝐧
(𝐗)𝟐
𝐒𝐱𝐱 ̿) =  (𝐗 𝟐 –
= ∑ ∑(𝐱 − 𝐗
𝐚𝐧
𝐢 =𝟏 𝐣=𝟏
𝐧
𝐗𝐢𝐘)
̿) =  (𝐗 𝐢𝐣 𝐘𝐢𝐣 −
̿)(𝐘 − 𝐘
𝐒𝐱𝐲 = ∑ ∑(𝐱 − 𝐗
𝐚𝐧
𝐢 =𝟏 𝐣=𝟏
𝐧
𝐘 𝐘 𝟐
𝐓𝐲𝐲 ̅ ̿ 𝟐
= ∑(𝐘 − 𝐘) = ∑{ − }
𝐧 𝐚𝐧
𝐢=𝟏
x x2
Txx =  (X
̅1 − X
̿) = −
n an

XY
Txy =  (X
̅1 − X
̿) = (Y ̿) = XY −
̅1 − Y
an
Eyy = Syy – Tyy
Exx = Sxx – Txx
Exy = Sxy – Txy
6.0 TUTOR MARKED ASSIGNMENT
Submit a one page discussion on the definition of analysis of covariance and its
assumption.

7.0 REFERENCES
- Damodar N. G., Dawn C. P. and Sangetha, G. (2012): Basic Econometrics. Tata
McGraw Hill Education Private Ltd. New Delhi India.
- Dominick, S. and Derrick, R. (2011): Statistics and Econometrics, (Schaum’s
Outlines) McGraw-Hill Company, New York.
- Kuotsoyanis, A. (2003): Theory of Econometrics (second edition).Palgrave
publishers Ltd (formerly Macmillan publishers Ltd), Houndmills, Basingstoke,
New York.
- www.youtube.com
MODULE 3: Multiple Regression Analysis
Unit 1: Estimation of multiple regressions
Unit 2: Partial correlation coefficient
Unit 3: Multiple correlation coefficient and coefficient of determination
Unit 4: Overall test of significance

UNIT ONE: MULTIPLE REGRESSIONS

CONTENTS
1.0 Introduction
2.0 Objectives
3.0 Main Content
3.1 Multiple regression
3.2 Assumptions of multiple regression
3.3 Estimation of multiple regression parameters
3.4 Worked example
4.0 Conclusion
5.0 Summary
6.0 Tutor-Marked Assignment
7.0 References/ Further Readings

1.0 INTRODUCTION
Multiple Regressions Defined
In introductory statistic, simple linear regression is one of the topics discussed.
Regression equation is an expression by which you may calculate a typical value of a
dependent variable say Y, on the basis of the values of independent variable(s).
Multiple regression model attempts to expose the relative and combine importance of the
independent variables on dependent variables.
Multiple regression models is one among the commonly used tools in research for the
understandings of functional relationship among multi-dimensional variables. The model
attempts to expose the relative and combine effect of the independent variable on the
dependent variable.
For your success in this course of study it is required that you have a thorough knowledge
of simple regression model, hypothesis testing among others.
2.0 OBJECTIVE
At the end of our discussion on multiple regression you should be to;
(i) Regress the independent variable on the dependent variable
(ii) Understand parameter estimates involved
(iii) You should know how to calculate the values of bo, b1, b2, … bn
(iv) Test of significance
Coefficient of multiple determinations
Test of overall significance of the regression
Partial correlation coefficient

3.0 MAIN CONTENT

3.1 Multiple Regression and Assumptions Defined
Multiple regression analysis is usually used for testing hypothesis about the relationship
between a dependent variable Y and two or more independent variable X and for
prediction or forecasting. Three variable linear regression models is usually written as:
Y = bo + b 1 X 1 + b 2 X 2 + 
Where Y = dependent variable
bo = intercept
b1, b2, bn = partial correlation coefficient or regression coefficient
 = error term or residuals
Assumptions of Multiple Regressions
Multiple regression models has the following assumptions
i. Randomness
ii. Normality
iii. Measurement error
iv. Independent of  and xs
v. Correct specification of model
vi. Multi-colinearity
vii. Homoscedascity
viii. Linearity
ix. Same number of cases and variables

SELF-ASSESSMENT EXERCISE
Define multiple regression model of four variables?

3.2 Estimation of the Parameters of the Multiple Regression (bo, b1 …bn)

For the purpose calculation and because of the parameters involved deviation method of
calculating regression will be used. The parameters involve are define as stated below:
b̂o = Y
̅ - b̂1 X
̅1 - b̂2 X
̅2

b̂1 = (x1y) (x22) – (x2y )(x1x2)

(x12) (x22) – (x1x2)2

̂b2 = (x2y) (x12) – (x1y)(x1x2)

(x12) (x22) – (x1x2)2
bo = Calculated bo
b̂1 = Calculated b1
b̂2 = Calculated b2
̅ = Y
Y
n
 ̅ = X
X
n
̅1
x1 = X1 - X
̅
y =Y-Y
̅2
x2 = X2 - X
Self-assessment exercise
Define b2?

3.3 Worked Example

The table below shows the value of expenditure on clothing, total expenditure and the
price of clothing.

Table M3.1.1
Table Showing Expenditure on Clothing, Total Expenditure and Price of Clothing
1990 1991 1992 1993 1994 1995
Price of clothing (X2 3.5 9.8 8.3 7.6 9.3 7.7
Total expenditure (X1 3.5 2.2 2.7 1.6 2.8 4.6
Value of expenditure clothing (Y) 2.0 1.5 1.7 1.6 2.7 3.5
Find the least square regression equation of Y on X1 and X2.

Table M3.1.2
Multiple Regression Table
Years Y X1 X2 y x1 x2 x1x2 x12 x22 x1y x2y y2
1990 20 35 35 -1.7 6 -42 -252 36 1764 -10.2 71.4 2.89
1991 15 22 98 -6.7 -7 21 -147 49 441 46.9 -140.7 44.89
1992 17 27 83 -4.7 -2 6 -12 4 36 9.4 -28.2 22.09
1993 16 16 76 -5.7 -13 -1 13 169 1 74.1 5.7 32.49
1994 27 28 93 5.3 -1 16 -16 1 256 -5.3 84.3 28.09
1995 35 46 77 13.3 17 0 0 289 0 226.1 0 176.89
n=6 Y X1 = X2 = ∑y=0 x1 = 0 x2 = 0 -414 548 2498 341 -7 307.34
=130 174 462
Y 130
̅ =
Y = = 21.7
n 6

̅1 = X1 = 174
X = 29
 n 6
̅2 = X2
X = 462 = 77
n 6
̅
y =Y- Y
̅2
x1 = X1 - X
̅2
x2 = X2 - X
b̂1 = (x1y) (x22) – (x2y)(x1x2)
(x12) (x22) – (x1x2)2
b̂1 = (341) (2498) – (- 7)(- 414)
(548) (2498) – (- 414)2
b̂1 = 851,818 - 2898
1368904 – 171396
b̂1 = 848,920
1197508
b̂1 = 0.70891
b̂1 ≅ 0.71

b̂2 = (x2y) (x12) – (x1y)(x1x2)

(x12) (x22) – (x1x2)2
b̂2 = (- 7) (648) - (341) (- 414)
(548) (2498) – (- 414)2
b̂2 = - 4536 + 141174
1368904 – 171396

b̂2 = 136,638
1197508
b̂2 = 0.1141
b̂2 ≅ 0.11
b̂0 = Y
̅ - b̂1 X
̅1 - b̂2 X
̅2

b̂o = 21.7 – 0.71(29) – 0.11 (77)

b̂o = 21.7 – 20.59 – 8.47
b̂o = - 24.370634778 + 21.7
b̂o ≅ - 7.36

The regression of Y on X1 and X2 is as written below

̂ = -7.36 + 0.71X1 + 0.11X2
Y
The equation above is the multiple regression of value of expenditure on clothing, on
price of clothing and total expenditure.
In multiple regression analysis of four parameters are of great importance from both the
equation and the result, these are;
(i) Partial correlation coefficient (bn)
(ii) Multiple correlation coefficient (R2)
(iii) Coefficient of determination (R2)
(iv) Test of significance
All these will be dealt with in the subsequent unit of this module.
SELF-ASSESSMENT EXERCISE
Define the mean deviation of Y?
4.0 CONCLUSION
In the course of our discussion on multiple regression you have learnt about
- Definition of multiple regression
- Assumptions of multiple regression
- Regression coefficients
- Estimation of Multiple regression equation

5.0 SUMMARY
In this unit, multiple regression model is given as
 Y = Bo + B1X1 + B2X2 + 
̂= Y
B ̅-B
̂1 X
̅1 - B
̂2 X
̅2
̂1 = (x1y) (x22) – (x2y)(x1x2)
B
(x12) (x22) – (x1x2)2
̂2 = (x2y) (x12) – (x1y)(x1x2)
B
(x12) (x22) – (x1x2)2
̂1 measures the change in Y for a unit change in X1 while holding X2 constant
Where B
B2 measure change in Y per units change in X2 holding X1 constant
6.0 TUTOR MARKED ASSIGNMENT
i. The simplest possible multiple regression model is a ---------
ii. Given that Y = B1X1 + B2X2 + B3X3 +  where X1 = 1 this is an example
of –

7.0 REFERENCES
- Damodar, N. G., Dawn, C. P., and Sangetha, G. (2012): Basic Econometrics. Tata
McGraw Hill Education Private Ltd. New Delhi, India.
- Dominick, S. and Derrick, R. (2011): Statistics and econometric (Schaum outline)
(2nd edition) McGraw Hill, New York.
- Oyesiku, O.K. and Omitogun, O. (1999): Statistics for Social and Management
Sciences: Higher Education Books Publishers Lagos.
- Oyesiku, O.O., Abosede, A.J., Kajola, S.O, and Napoleon, S.G.(1999): Basics of
Operation research. CESAP Ogun State University. Ago-Iwoye, Ogun State.
UNIT TWO: PARTIAL CORRELATION COEFFICIENT
CONTENTS
1.0 Introduction
2.0 Objectives
3.0 Main Content
3.1 Partial correlation defined
3.2 Assumptions of multiple regression
3.3 Estimation of multiple regression parameters
3.4 Worked example
4.0 Conclusion
5.0 Summary
6.0 Tutor-Marked Assignment
7.0 References/ Further Readings

1.0 INTRODUCTION
It is assumed that you must have read unit 1 of this module that talks about multiple
regression, a detailed understanding of this will be assumed. this unit is building on the
unit 1 of this module. This unit will be dealing with thorough explanation of the
parameters involved in the regression analysis.

2.0 OBJECTIVE
At the end of our discussion, you should be able to understand the following concepts
such as;
- Partial regression co-efficient
- Estimation of partial regression co-efficient

3.0 MAIN CONTENT

3.1 Partial Correlation Defined
Partial correlation coefficient measures the correlation between the dependent variable
and independent variables in the model. The regression coefficient B1 and B2 are known
as partial regression or partial slope coefficient. B1 measures the change in the mean
value of Y per unit change in X1 after removing the influence of X2 or holding X2
constant, this gives the direct effect or net effect of a unit change in X1 on the value of Y.
B2 coefficient measures the change in the mean value of Y per unit change in X2 holding
X1 constant, this gives the direct effect or net effect of a unit change in X2 on the mean of
Y.
SELF-ASSESSMENT EXERCISE
Define partial correlation?

3.2 Estimation and Explanation of Partial Correlation Coefficient

Correlation coefficient (r) is defined as a measure the degree of linear association
between two variable for three variable regression model we can compute 3 correlation
coefficients ryx1.x2, ryx2.x1, rx1x2
Where;
Ryx1 .x2 = partial correlation coefficient between Y and X1 holding X2 constant
ryx2 .x1 = partial correlation coefficient between Y and X2 holding X1 constant
rx1x2 = partial correlation coefficient between X1 and X2 holding Y constant
The formular for these correlation coefficients are
ryx1 = x1y
x12 y2
ryx2 = x2y
x22 y2
rx1x2 = x2x1
x22 x12
ryx1.x2 = ryx1 – (ryx2)(rx1x2) = ryx1 – (ryx2)(rx1x2)
1– r2x1x2 1 – r2yx2 (1 – r2x1x2) (1 –r2yx2)
ryx2.x1 = ryx2 – (ryx1)(rx1x2) = ryx2 – (ryx1)(rx1x2)
1 – r2x1x2 1 – r2yx1 (1 – r2x1x2)(1–r2yx1)
rx1x2.y = rx1x2 – (ryx1) (ryx2)
1– r2yx1 1 – r2yx2
rx1x2y = rx1x2 – (ryx1) (ryx2)
1 – r2yx1 1–r2yx2

Partial correlation coefficients range in value between -1 and +1. This value(s) is usually
used to determine the relative importance of the different explanatory variables in a
multiple regression.

SELF–ASSESSMENT EXERCISE
Define rx1x2.y

3.3 Worked Examples

Here values of our correlation coefficient will be derived from the table in our unit of this
moodle1i.e from the table of analysis.

Table M3.2.1
Correlation Coefficient Table
Years Y X1 X2 Y x1 x2 x1x2 x12 x22 x1y x2y y2
1990 20 35 35 -1.7 6 -42 -252 36 1764 -10.2 71.4 2.89
1991 15 22 98 -6.7 -7 21 -147 49 441 46.9 -140.7 44.89
1992 17 27 83 -4.7 -2 6 -12 4 36 9.4 -28.2 22.09
1993 16 16 76 -5.7 -13 -1 13 269 1 74.1 5.7 32.49
1994 27 28 93 5.3 -1 16 -16 1 256 -5.3 84.3 28.09
1995 35 46 77 13.3 17 0 0 289 0 226.1 0 176.89
n=6 Y =130 X1 = X2 = ∑y=0 x1 = 0 x2 = 0 -414 648 2498 341 -7 307.34
174 462

∑x1y =341; ∑x2y = -7;∑x12 = 648;∑x22; ∑y2=307.34; ∑x1x2= -414

ryx1 = x1y

√x12 √y 2
= 341
√648 x √307.34
= 341
√648 x 307.34
= 341
√199, 156.32
ryx1 = 341
446.2693357
ryx1 = 0.764112549
ryx1 ≅ 0.76
r2yx1 = 0.583867987
r2yx1 = 0.58
ryx2 = x2y

√x 2 2 √y 2
ryx2 = -7
(√2498) (√307.34)
ryx2 = -7
 (√2498 x 307.34
ryx2 = -7
√767,7353.32
ryx2 = -7
876.2050673
= - 0.007988997395
ryx2 = - 0.008
r2yx2 = 0.0000638 = 0.000064
rx1x2 = x1x2
(√x 2 2) (√x12)
rx1x2 = - 414
(√2498) (√648)
rx1x2 = - 414
√2498 x 648
rx1x2 = - 414
√1,618,704
rx1x2 = - 414
1,272.282987
rx1x2 = - 0.325399305
rx1x2 = - 0.33
r2x1x2 = 0.105884707
r2x1x2 = 0.11

Thus ryx1.x2 = ryx1 – (ryx2) (rx1x2)

1 –r2x1x2 1 – r2yx2
ryx1.x2 = 0.76 – ( - 0.008) (- 0.33)
√(1 − 0.11)(1 − 0.000064)
ryx1.x2 = 0.76 – 0.00264
 √(0.89)(0.999936)
ryx1.x2 = 0.75736
√0.88994304
ryx1.x2 = 0.75736
0.943367924
ryx1.x2 = 0.802825685
ryx1.x2 ≅ 0.802
ryx1.x2 = 80.2%

ryx2.x1 = ryx2 – (ryx1) (rx1x2)

(1 – r2x1x2) ( 1 – r2yx1)
ryx2.x1 = - 0.008 – ( 0.76) (- 0.33)
√(1 − 0.11)(1 − 0.58)
ryx2.x1 = - 008 + 0.2508
√(0.89)(0.42)

ryx2.x1 = 0.2428
√0.3738
ryx2.x1 = 0.2428
0.61 ryx2.x1 1391854
ryx2.x1 = 0.397126651
ryx2.x1 ≅ 0.40
ryx2.x1 = 40%
Therefore, from the calculations above it shows that X1 explains more than X2 and X1is
more important in explaining variation in Y.

SELF-ASSESSMENT EXERCISE
Define rx1x2?
4.0 CONCLUSION
In the course of our discussion on partial correlation coefficient you must have learnt
about the following:
- Partial correlation definition
- Partial correlation between the dependent variable and the independent variable such
as; ryx1.x2 = partial correlation coefficient between variable y and x1 holding variable
x2 constant
ryx2 .x1 = partial correlation coefficient between variable y and x2 holding x1 constant.
rx1x2.y = partial correlation between variable y and x1 variable x2 holding variable y
constant

5.0 SUMMARY
In the course of our discussion the following formulas were made use of;
ryx1 = x1y
x12 y2
ryx2 = x2y
x22 y2
rx1x2 = x2x1
x22 x12
ryx1.x2 = ryx1 – (ryx2)(rx1x2)
(1 – r2x1x2)(1 – r2yx2)
ryx2 .x1 = ryx2 – (ryx1)(rx1x2)
(1 – r2x1x2)(1 – r2yx1)

6.0 TUTOR MARKED ASSIGNMENT

Given that Y = Bo + B1 X1 + B2X2 +  the partial regression coefficient is given by------
REFERENCES /FURTHER READINGS
- Damodar, N. G., Dawn, C. P., and Sangetha, G.(2012): Basic Econometrics. Tata
McGraw Hill Education Private Ltd. New Delhi, India.

- Dominick, S. and Derrick, R. (2011): Statistics and econometric (Schaum outline)

(2nd edition) McGraw Hill. New York.

- Oyesiku, O.K. and Omitogun, O. (1999): Statistics for Social and Management
Sciences: Higher Education Books Publishers, Lagos.

- Oyesiku, O.O., Abosede, A.J., Kajola, S.O. and Napoleon, S.G. (1999): Basics of
Operation research. CESAP Ogun State University, Ago-Iwoye, Ogun State.
UNIT THREE: Multiple Correlation Co-efficient and Coefficient of Determination
CONTENTS
1.0 Introduction
2.0 Objectives
3.0 Main Content
3.1 Partial correlation defined
3.2 Assumptions of multiple regression
3.3 Estimation of multiple regression parameters
3.4 Worked example
4.0 Conclusion
5.0 Summary
6.0 Tutor-Marked Assignment
7.0 References/ Further Readings

1.0 INTRODUCTION
This unit is an extension of unit one and two of this module. This unit requires thorough
knowledge of unit 1 and unit two. In this unit we are going to look at multiple Correlation
Coefficients (R) and multiple coefficient of determination (R2).

2.0 OBJECTIVE
At the end of this unit you should be able to:
- Estimate multiple correlation coefficient (r)
- Estimate coefficient of determination
- Interprete your answer i.e. statistical interpretation
3.0 MAIN CONTENT
3.1 Multiple Correlation Coefficient (R) Coefficient of Determination (R2)
Defined
Multiple correlation coefficients represented by R measures the degree of linear
association between two or more variables. Say variable Y and the entire explanatory
variable jointly. Its value can be positive or negative; multiple correlation coefficients is
always taken to be positive. In practice the multiple correlation coefficients is of little
importance. The more meaningful coefficient is the coefficient of determination R2 or r2.
Coefficient of determination (R2) is defined as the proportion of the total variation in Y
explained by the multiple regression of Y on X1 and X2. It measures goodness of fit of the
regression equation. In a three variable model we are always interested in knowing the
proportion of the variation in Y explained by each of the explanatory variable X1 and X2.
The coefficient of determination is denoted by R2 or r2. Because of the relative
importance of coefficient of determination (R2) we concentrate more on the coefficient of
determination (R2).

SELF-ASSESSMENT EXERCISE
The most important coefficient is --------

3.2 Estimation of Coefficient of Determination

Conceptually, it is often written as:
Coefficient of determination = r2 = R2 = b̂1 y1x1 + b̂2yx2
y2
R = r = ryx1x2 = r2yx1 + r2yx2 – 2ryx1 . ryx2 . rx1x2
1 – r2x1x2
The value of R2 lies between 0 and 1, if it is 1, the fitted regression line explains 100% of
the variation in Y, on the other hand, if it is 0, the model does not explain any of the
variation in Y. typically, however, R2 lies between these two extremes values. The fit is
said to be better, the closer R2 is to 1.
 Self-assessment exercise
The coefficient of determination usually lies between------ and ------

3.3 Worked Example

From our calculation in unit 1 & 2 of this module especially the table in unit 1, we are
going to derive our values from the table in unit 1.

Table M3.3.1
Multiple Regression Table
Years Y X1 X2 Y x1 x2 x1x2 x12 x22 x1y x2y y2
1990 20 35 35 -1.7 6 -42 -252 36 1764 -10.2 71.4 2.89
1991 15 22 98 -6.7 -7 21 -147 49 441 46.9 -140.7 44.89
1992 17 27 83 -4.7 -2 6 -12 4 36 9.4 -28.2 22.09
1993 16 16 76 -5.7 -13 -1 13 169 1 74.1 5.7 32.49
1994 27 28 93 5.3 -1 16 -16 1 256 -5.3 84.3 28.09
1995 35 46 77 13.3 17 0 0 289 0 226.1 0 176.89
n=6 Y X1 X2 ∑y=0 x1 = 0 x2 = 0 -414 548 2498 341 -7 307.34
=130 = =
174 462

R2 = 0.59 (341) + 0.094(-7)

307.34
R2 = 201.19 – 0.658
307.34
R2 = 200.532
307.34
R2 = 0.652476085
R2 = 0.65
R2 ≅ 65%
This implies that the explanatory variable (x1 and x2) can only account for 65% variation
in variable Y i.e. both x1 and x2 contributes 65.2% to the explanation of the variation in
Y.
ryx1x2 = r2yx1 + r2yx2 – 2yx1 . ryx2 . rx1x2
1 – r2x1x2

ryx1x2 = 0.058 + 0.000064 – 2 (0.7641). (0.008) .0.11

1 – 0.11

ryx1x2 = 0.580064 – 2 (15282) (- 0008) 0.11

0.89

ryx1x2 = 0.580064 – 0.001344816

0.89
ryx1x2 = 0.578719184
0.89

ryx1x2 = 0.650246274

ryx1x2 = 0.806378493
r2 = 0.650246274
r2 ≅ 0.65%

SELF-ASSESSMENT EXERCISE
When r2=0.85, what is the economic interpretation of this?

4.0 CONCLUSION
In the course of our discussion of this unit, you have learnt about the following:
- Concept of multiple correlation
- Coefficient of determination
- Estimation of R2 & r
- Interpretation of r & r2

5.0 SUMMARY
In our discussion of this unit we defined coefficient of determination R2 as
R2 = b̂1 y1x1 + b̂2yx2
y2
r = r2yx1 + r2yx2 – 2ryx1 . ryx2 . rx1x2
1 – r2x1x2
The closer the r2 is to 1, the better

6.0 TUTOR MARKED ASSIGNMENT

i. The measure of proportion or percentage of variation in Y explained by the
explanatory variable x1 … xn jointly is given by -----------
ii. Multiple coefficient of determination measures------------
7.0 REFERENCES
- Damordar, N. G., Dawn, C. P. and Sangeetha, G. (2012): Basic Econometrics (5th
edition) Tata McGraw Hill Education Private Limited. New Delhi, India.

- Dominick, S. and Derrick, R. (2011): Statistics and econometrics (Schaum outline)

(2nd edition). McGraw Hill, New York.

- Oyesiku O.K. and Omitogun O. (1999): Statistics for social and management
science (2nd edition). Higher Education Books Publisher, Lagos.

- Oyesiku, O.O., Abosede, A.J., Kajola, S.O. and Napoleon, S.G. (1999): Basics of
Operation research. CEAP OSU, Ago-Iwoye. Ogun State.
UNIT FOUR: TEST OF SIGNIFICANCE
CONTENTS
1.0 Introduction
2.0 Objectives
3.0 Main Content
3.1 Test of significance defined
3.2 Estimation of test of significance
3.3 Summary of F-statistics
3.4 Worked example
4.0 Conclusion
5.0 Summary
6.0 Tutor-Marked Assignment
7.0 References/ Further Readings:

1.0 INTRODUCTION
This unit completes this module, so it is required that thorough knowledge of unit one to
unit three is very germane. It is important to test for the significance of the value of the
regression, Coefficients, and the level of prediction or explanation given by the
regression equation.

2.0 OBJECTIVE
At the end of this unit the student(s) should be able to calculate and understand:
- The calculation of F.statistics (Fcal)
- Check the corresponding tabulated value of F.statistics through its degree of freedom.
- Compare the F.statistics and Ftab
- Interprete your answer
3.0 MAIN CONTENT
3.1 Test of Significance Defined
Test of significance is a procedure by which sample results are used to verify truity of
falsity of a null hypothesis. The key idea behind test of significance is that of a test
statistics (estimator and the sampling distribution of such a statistics under the null
hypothesis).
The decision to accept or reject Ho is made on the basis of the test statistics obtained from
the data at hand.
The overall significance of the regression can be tested with the ratio of the explained to
the unexplained variance. This follows an F-distribution with k – 1 and n – k degree of
freedom, where n is the number of observations and k is the number of parameters
estimated.
The joint hypothesis can be tested by the analysis of variance (Anova).

SELF-ASSESSMENT EXERCISE
The decision to accept or reject Ho depends on ---------

3.2 Estimation of Test of Significance

The F-statistics or F-ratio for the test of significance can be written as:
Ho: 1 = 2; Hi: 1  2

Fk−1,n−k = y/(k-1) = R2/ (k-1)

e2/(n-k) (1-R2) (n-k)
Also Anova table can as well be used for test of significance.
Table M3.4.1
Anova Tables for 3-Variables Regression
Source of variation Sum of squares DF Mean sum of square
Due to Regression ̂1yx1+B
B ̂2yx2 2 ̂1yx1 + B
B ̂2yx2
(ESS) 2
Due to Residual (RSS) i2 n–3 𝜎2 = i2
n–3
Total yi2 n–1

Note
yi2 = B
̂1yx1 + B
̂2 yx2 + 12
TSS = ESS + RSS
̂1yx1 + B
F-.ratio = (B ̂2yx2)/2

2/n – 3
E 
̂2 ̂2) = σ2
= E (σ
n-3

3.3 Summary of F-Statistic

Table M3.4.2
F-STATISTICS TABLE
Null hypothesis Alternative Criteria Region
Ho Hypothesis Reject Ho if
H1
σ12 = σ22 σ12 > σ22 S12 > F, ndf, ddf(ftab)
S22
σ12 = σ22 σ12  σ22 S12 > F/2, ndf, ddf (ftab)
S22
or <F1-/2, ndf, ddf
If the calculated F-ratio (Fc) exceeds the tabular value of F (Ftab) at the specified level of
significance and degree of freedom, the hypothesis is accepted that the regression
parameters are not all equal to zero and that R2 is significantly different from zero.

If the null hypothesis is true, it gives identical estimates of true σ2. This statement should
not be surprising because if there’s a trivial relationship between y and x1 and x2 the
source of variation in Y will be due to the random forces usually represented by ei or 1.
If however, the null hypothesis is false, that is x1 and x2 actually influence Y; the equality
will not hold. Here, the ESS will be relatively larger than the RSS taking due account of
their respective degree of freedom. Therefore, the F-.ratio provides a test of the null
hypothesis that the true slope coefficients are simultaneously zero.
DECISION CRITERIA; If the F-ratio calculated exceeds the critical F-value from the
table at the  percent level of significance we reject Ho; otherwise do not reject it.
Alternatively if the F-cal of the observed F is sufficiently low accept Ho.

SELF-ASSESSMENT EXERCISE
State the decision criteria for test of significance?

3.4 Worked Example

Ho; = 1 = 2 = 0
note that Fk-1,n-k = Y
̂2/k – 1 = R2 / (k – 1)
2/n – k (1 – R2)(n – k)
From the above it is glaring that calculation of 2 is required, so there’s need to generate
a new table apart from the one generated in unit one of this modules as to get the value
for our 2.
Table M3.4.3
Test of Significance Table
Years Y X1 X2 ̂
Y e e2
1990 20 35 35 21.1294 - 1.1294 1.2755
1991 15 22 98 12.1294 - 4 1294 17.0519
1992 17 27 83 20.694 - 3.694 13.6456
1993 16 16 76 13.6094 2.3906 5.7150
1994 27 28 93 22.2194 4.7806 22.8541
1995 35 46 77 31.3994 3.6006 12.9643
n=6 130 174 462 0 73.506436

Note: The trend equation = Y = - 2.6706 + 0.59x1 + 0.09x2

̂ is arrived at by substituting various values of x1 and x2 into the trend equation
The Y
̂ (ie column 2 – column 5)
e =Y-Y
y12 = 307.34 (from unit one)

Method I
F3-1,6-3 = 307.34 = 307.34
3–1 2
73.506432 73.506432
6–3 3
Fcal = F2,3 = 153.67 = 6.271696061
24.502144
Method II
Table M3.4.4
Anova Table for 3-Variance
Sources of variation Sum of squares DF MSS
ESS 200.56 2 100.28
RSS 73.506436 3 24502
Total 274.066436 5

̂y1x1 + ̂B2y1x2 = 0.59 (341) =- 0.09 (-7)

ESS = B
= 201.19 – 0.63
= 200.56
RSS = e12 = 73.506436
Fcal = ESS/df = 100.28
RSS/df 24.502
= 4.09272
Fcal ≅ 4.1
Ftab = F2,3 = 9.55
Ftab = 9.55
Since Fcal < Ftab, accept Ho and reject H1.

SELF-ASSESSMENT EXERCISE
What will be the decision criteria if Fcal > Ftab?

4.0 CONCLUSION
From our discussion on this unit you have learnt about:
- Definition test of significance
- Estimation of test of significance
- The interpretation of resulting
- Estimation through ANOVA table and otherwise
- Derivation of the error term ui or ei

5.0 SUMMARY
In the course of our discussion the following formulars where discussed
Fk-1,n-k = Y
̂2/(k – 1)

e2/(n – k)
y12 = 𝑏𝑖
̂ 1y1x1 + bi
̂ 2y1x2

12 = (Y-Y
̂ )2

6.0 TUTOR MARKED ASSIGNMENT

Given the regression model Y = B1 + B2X2 + B3 X3+U, how would you state the null
hypothesis to test for test for significance of x1 and x2 on Y.

7.0 REFERENCES
- Damodar, N. G., Dawn, C. P. and Sangetha, G. (2012): Basic Econometrics. Tata
McGraw Hill Education Private Ltd. New Delhi, India.

- Dominick, S. and Derrick, R. (2011): Statistics and econometrics (schaum outline)

(2nd edition) McGraw Hill, New York.

- Oyesiku, O.K. and Omitogun, O. (1999): Statistics for Social and Management
Sciences: Higher Education Books Publishers, Lagos.

- Oyesiku, O.O., Abosede, A.J., Kajola, S.O., and Napoleon, S.G.(1999): Basics of
Operation research. (CESAP Ogun State University), Ago iwOye, Ogun State.
MODULE 4: Time series analysis
Unit 1: Time series and its components
Unit 2; Quantitative estimation of time series
Unit 3: Price index
UNIT ONE: TIME SERIES
CONTENTS
1.0 Introduction
2.0 Objectives
3.0 Main Content
3.1 Time Series defined
3.2 Component of time series
3.3 Graphical representation of trends
3.4 Worked example
4.0 Conclusion
5.0 Summary
6.0 Tutor-Marked Assignment
7.0 References/ Further Readings

1.0 INTRODUCTION
In all the social sciences, and particularly economics and business, the problem of how
condition changes with the passage of time is of utmost importance. For study of such
problems, the appropriate kind of statistical information consist of data in the form of
time series, figures which shows the magnitude of a phenomenon month after month or
year after year. The proper methods for treating such data and thus summarizing the
experience which they represent are indispensable part of the practicing statistician
equipment.
2.0 OBJECTIVE
At the of this unit, you should be able to
- Understand or define time series
- Understand component part of time series
- Understand methods of estimating time series
- Estimation and graphical representation of the trend

3.0 MAIN CONTENT

3.1 Time Series Defined
Time series refers to sequence of observations that gives information on how data has
been behaving in the past.
You might wonder why we should spend so much effort constructing series showing
what has happened in the past. This is history and should we not rather be looking to the
future? As you know the twentieth century is age of planning: government plans the
economy for many years ahead; public corporation plan output and investment; most
state plan to keep the rate of inflation down to an acceptable level.
Good planning is usually based on information and this is where the time series comes
into its own. It provides information about the way in which economic and social variable
have been behaving in the recent past, and provides an analysis of that behaviour that
planner cannot ignore. Naturally, if we are looking into the future, there is certain
assumption we have to make, the most important of which is that the behavioural pattern
that we have found in the past could continue into the future. In looking to the future
there are certain pattern that we assume will continue and it is to help in the
determination of these pattern that we undertake the analysis of the time series.
Time series is usually ordered in time or space. Time series is denoted by sequence (Y t)
where Yt is the observed value at time t.
Essentially, time series is usually applied to economic and business problems whose
purpose of analyses data is to permit a forecast to the future both in the long term and
short term. It may be used as an essential aid to planning. Example of time series data are
volume of sales, the character and magnitude of its cost of production etc. population
figure, price level, demand of a commodity.

SELF-ASSESSMENT EXERCISE
The essence of time series is forecast. (true/false)

3.2.0 Components of Time Series

The nature or variation or type of changes in times series can be categorise into:
- Secular trend or long term movement
- Seasonal variation
- Cyclical variation
- Irregular or residual variation

3.2.1 Secular Trend

This refers to the general direction in which the graph of time series appears to be going
over a long period of time. This explains the growth or decline of a time series over a
long period. Time series is said to contain a trend if the mean or average of series changes
systematically with time. The trend could be upward or downward, this could take any of
the shape below.
GRAPHICAL REPRESENTATION OF SECULAR TREND

Fig. M4.1.1
3.2.2 Seasonal Variation
This refers to short term fluctuation or changes that occur at regular intervals less than a
year. It is usually brought about by climatic and social factor(s), it is usually because of
an event occurring at a particular period of the year. Examples of these are sale of card
during valentine period, sale of chicken during xmas, new year or any festive period(s).

GRAPHICAL REPRESENTATION OF SEASONAL VARIATION

Fig. M4.1.2

3.2.3 Cyclical Variation

This refers to long term variations about the trend usually caused by disruption in
services or socio-economic activities, cyclical variations are commonly associated with
economic cycles, successive boom and slumps in the economy. A good example of this is
business cycle.
GRAPHICAL REPRESENTATION OF CYCLICAL VARIATION
Fig. M4.1.3
3.2.4 Irregular Variation
This refers to time series movement that are not definite this is usually caused by unusual
or unexpected and unpredictable events such as strike, war, flood, disasters. Here, there’s
no definite behavioural pattern.

GRAPHICAL REPRESENTATION OF IRREGULAR VARIATION

Fig. M4.1.4

SELF-ASSESSMENT EXERCISE
The trend of secular trend can either be upward or downward. (true/false)

3.3.0 Measurement of Trend

Basically trend values of a time series can be estimated by any of the following methods:
- Free hand
- Least square method
- Moving average and
- Semi average method

3.3.1 Free Hand Method

This method involves the drawing a scattered diagram of the values with time as the
independent variable on the x-axis and then drawing the trend line by eye. This
 method is condemned because it is subjective and inaccurate method of obtaining a
Trend line.
GRAPHICAL REPRESENTATION OF FREE HAND METHOD

Fig. M 4.1.5
Other quantitative methods will be dealt with in the next unit

4.0 CONCLUSION
In the course of our discussion on time series analysis you have learnt about
- Time series
- Time series data
- Component of time series
- Free hand trend measurement
-
5.0 SUMMARY
Majorly time series decomposes itself into the following;
- Secular trend or long term movement
- Seasonal variation
- Cyclical variation
- Irregular or residual variation
6.0 TUTOR MARKED ASSIGNMENT
Table Showing Sales of a Chemist
Years Sales
2000 85
2001 96
2002 108
2003 123
2004 98

Make a freehand sketch of the above information

7.0. REFERENCES/FURTHER READINGS

- Adedayo, O.A. (2006): Understanding Statistics. JAS Publishers, Akoka, Lagos.

- Dawodu, A.F. (2008): Modern business Statistics 1. NICHO Printing Works,

Agbor, Delta State.

- Esan, E.O. and Okafor, R.O. (2010): Basis Statistical Method. Tony Chriisto
Concept, Lagos.

- Olufolabo, O.O. & Talabi, C.O. (2002): Principles and Practice of Statistics HAS-
FEM (NIG) ENTERPRISES Somolu Lagos.

- Owen, F. and Jones, R. (1978): Statistics. Polytech Publishers Ltd. Stockport.

- Oyesiku, O.K. and Omitogun, O.(1999): Statistics for social and Management
Sciences. Higher Education Books Publisher, Lagos.

UNIT TWO; ESTIMATION OF TIME SERIES

CONTENTS
1.0 Introduction
2.0 Objectives
3.0 Main Content
3.1 Estimation of time Series using least square method
3.2 Estimation of time series using moving average
3.3 Estimation of time series using semi average method.
3.4 Worked example
4.0 Conclusion
5.0 Summary
6.0 Tutor-Marked Assignment
7.0 References/ Further Readings

1.0 INTRODUCTION
This unit is an extension of unit one of this module, here, you are going to learn more
about estimation of time series data, also, a thorough understanding of unit one of this
module is required for proper understanding of this module.

2.0 OBJECTIVE
At the of this unit, you should be able to
- estimate any time series data,
- Understand methods of estimating time series,
- Estimate and do the graphical representation of the trends.

3.0 MAIN CONTENT

3.1.0 Least Square Method of Estimation
As defined in Preceding unit of this module, Time series refers to sequence of
observations that gives information on how data has been behaving in the past.
Estimation has to do with how time series are calculated, in this sub section we shall talk
about three methods of estimation or measurement. These are method of least square,
moving average and semi average method.

3.1.1 Least Square Method

This method is a statistical technique usually used in calculating the line of best fit or line
of goodness that measures the goodness of fit of the curve, this is usually independent of
human judgments, it makes an assumption that the trend line is a straight one. The least
square formular is given as;
Y = a + bx + e
Where a = intercept
b = slope of the curve
e = error term

Formular 1
Trend equation of the least square method is given as
Y = na + bx
Y = ax + bx2
Where  = summation term derived from the data of the problem at hand
x = sum of X values
Y = sum of Y values
xy = sum found by multiplying each Y by corresponding X value and adding the
Products
n = no of items involved in the whole time series
The least square estimates of a and b are the solution to the normal equation above which
can be solve simultaneously.

Formular 2
The general formular is as given below
b = nXY - XY
nX2 – (X)2
̅ – 𝑏̂ x̅
a=Y
̅ = Y
where Y
n
̅ = x
X
n

Worked Example
Given the 7weeks information below about the sales of a company
Table M 4.1.1
Table Showing the Sales of a Company
Wk Sales
1 15
2 25
3 38
4 32
5 40
6 37
7 50

Let X represents the weeks

Y represent the sales value
Table M4.1.2
Least Square Method Table of Analysis
X Y XY X2
1 15 15 1
2 25 50 4
3 38 114 9
4 32 128 16
5 40 200 25
6 37 222 36
7 50 350 49
∑X= 28 ∑Y=237 ∑XY=1079 ∑X2=140

̂ = nXY - XY
B
nX2 – (X)2

n=7 ̅ = 28 = 4
X
7

x = 28 ̅ = 237 = 33.857
Y
7

y = 237
xy = 1079
x2 = 140
̂ = 7(1079) – 28 (237)
B
7 (140) – 282
̂= 7553 – 6636
B
980 – 784
̂ = 917
B
196
̂ = 4.67857
B
̅ -B
a=Y ̂ x̅
a = 33.857 – (4.67857)4
= 33.857 – 18.7142
= 15.1427
The trend equation will be:
Y = 15.1427 + 4.6785x
This trend equation can be used in forecasting into future sales of the company, for
example future sales value for the 10th and 12th week can be known by simply
substituting the week’s value into the trend equation.
i.e. for the 10th week we have;
Y = 15.1427 + 4.6785(10)
Y = 15.1427 + 46.785
Y = 61 – 9277
Y ≅ 62
For the 12th week
Y = 15.1427 + 4.6785(12)
Y = 15.1427 + 56.142
Y = 71.2847
Y = 71

3.2.0 Moving Average Method

A moving average is a simple arithmetic mean. We select a group of figures at the start of
the series e.g. 3,4,5,7 and average them to obtain our first trend figure. Then you drop the
first figure and include the next item in the series to obtain a new group. The average of
this group gives the second trend figure. You continue to do this until all figures in the
series is exhausted.
There is no doubt that the trend eliminates the large scale fluctuations found in the
original series moving average smoothing is a smoothing technique used to make the
long-term trend of a time series cleared.
Example 2
The table below contained information about the actual sales of a company
Table M4.1.3
Table Showing the Sales of a Company
Month Sale (units)
Jan 350
Feb 340
Mar 360
April 310
May 280
June 300
July 270
August 260
Sept 310
Oct 350
Nov 370
Dec 390

Prepare a 3 month moving average forecast

Solution
Table M4.1.4
3- Month Moving Average Method Table of Analysis
Months Sales 3months Moving total 3months moving average trend
Jan 350
Feb 340 1050 350
Mar 360 1010 336.7
April 310 950 316.7
May 280 890 296.7
June 300 850 283.3
July 270 830 276.7
Aug 260 840 280
Sept 310 920 306.7
Oct 350 1030 343.3
Nov 370 1110 370
Dec 390

Column 1 on the table represents the months

Column 2 represents the sale’s figure
Column 3 is arrived at by adding the sales figure in 3s i.e
Jan + Feb + Mar = 1050
Feb + Mar + April = 1010
Mar + April + May = 950
Column 4 is arrived at by dividing the column 3 by the n which happen to be the moving
average. This Rs called the trend.
3.2.1 GRAPHICAL REPRESENTATION OF MOVING AVERAGE TREND

Fig. M4.1.6

Example 2
From the time series data below determine the trend on sales of a company
Table M4.1.5
Table showing sales of a company per quarter
Quarter
Years 1 2 3 4
1982 600 820 400 720
1983 630 840 420 740
1984 670 900 430 760
Prepare a 4-quarter moving average
Solution
Table M 4.1.6
4-point moving average table of analysis
Year Quarter Sales 4 point moving 4-point average 2 point total Moving
total moving or 4 or centre average
quarterly total (trend)
average
1982 1 600 - - - -
2 820 - - - -
2540 635 -
3 400 1277.5 638.75
2570 642.5 - -
4 720 1290 645
2590 647.5 - -
1983 1 630 1300 650
2610 652.5 - -
2 840 1310 655
2630 657.5 - -
3 420 1325 662.5
2670 667.5 - -
4 740 1350 675
2730 682.5
1984 1 670 1367.5 683.75
2740 685 - -
2 900 1375 687.5
2760 690 - -
3 430 - - - -
4 760 - - - -

Column 1 represent years

Column 2 represents quarter periods
Column 3 represents sales values
Column 4 is arrived at by adding the sales value in 4s
Column 5 is derived by dividing column 4 by the nos of quarters
Column 6 is the total of column 5 when taken in 2s
Column 7 is arrive at by dividing column 6 by 2
GRAPHICAL REPRESENTATION OF FOUR QUARTER MOVING AVERAGE
TREND

Fig. M4.1.7

3.3.0 Semi Moving Average Method

This method is usually used to estimate trends by separating or dividing that data into two
equal parts and averaging the data each part, thus, obtaining two points on the graph of
time series. A trend is then drawn between these two points and trend value can be
determined. If the number of years is odd, the middle year is deleted and the group can
then be divided into two equal parts.

Example
Table M4.1.8
Semi- Moving Average Method Table of Analysis
Years Quarter Y sales X Semi Average Semi average method
Total trend
1992 1 600 -6
2 820 -5
3 400 -4
4 720 -3 4010 668.33
1993 1 630 -2
2 840 -1
3 420 1
4 740 2
1994 1 670 3
2 900 4
3 430 5
4 760 6 3,920 653.3
Column 4 represents the total of the 1st half and 2nd half.
Column 5 is arrived at by dividing the column 4 by 6 this represents the trend value,
when plotted in a graph it gives the trend line.

3.3.1 GRAPHICAL REPRESENTATION OF SEMI MOVING AVERAGE TREND

Fig. M4.1.8

SELF-ASSESSMENT EXERCISE
State the least square equation of a time series data?

4.0 CONCLUSION
In the course of our discussion on estimation of time series, you have learnt about
- least square method
- moving average method
- semi average method.

5.0 SUMMARY
The least square trend equation is written as
Y= a+bx+e
̅ - b̂ x̅
Where a = intercept = Y
b̂ = slope = nXY - x(Y)
nx2 – (x)2
For moving average develop the following
 - n – moving total
- determine the moving average
- plot the trend value to know the trend line

For semi average

- divide the data into 2 equal part
- when you have an odd data given, eliminate or delete the data in the middle
- get the half way total of each division
- divide the half way total by n depending on data supplied

6.0 TUTOR MARKED ASSIGNMENT

Table 4.1.9
Table Showing the Number of Prescriptions Dispensed by a Chemist
Year Quarters
1 2 3 4
2000 - - 60 71
2001 69 67 62 69
2002 73 66 62 68
2003 72 66 65 67
2004 75 - - -

Prepare a 4-point moving average of the above information?

7.0 REFERENCES/FURTHER READINGS
- Adedayo, O. A. (2006): Understanding Statistics. JAS Publishers, Akoka Lagos.

- Dawodu, A.F. (2008): Modern business Statistics 1. NICHO Printing Works.

Agbor, Delta State.

- Esan, E.O. and Okafor, R.O. (2010): Basic Statistical Method. Tony Christo
Concept, Lagos.

- Olufolabo, O.O. & Talabi, C.O. (2002): Principles and Practice of Statistics HAS-
FEM (NIG) ENTERPRISES Somolu Lagos.

- Owen F. and Jones, R. (1978): Statistics. Polytech Publishers Ltd, Stockport.

- Oyesiku, O.K. and Omitogun, O. (1999): Statistics for social and Management
Sciences. Higher Education Books Publisher, Lagos.
UNIT THREE: PRICE INDEX
CONTENTS
1.0 Introduction
2.0 Objectives
3.0 Main Content
3.1 Price index defined
3.2 Computation of price index
3.3 Worked example
4.0 Conclusion
5.0 Summary
6.0 Tutor-Marked Assignment
7.0 References/ Further Readings

1.0 INTRODUCTION
In introductory statistics a lot of meaning has be given to the average, this (average) has
been confirmed not to be necessarily representative of the data it describes. statisticians
have constructed a device that attempt to measure the magnitude of economic change
over time, a device called index number. This device is also used for international
comparison of economic data. This device called index number is what this unit shall be
looking at, we shall examine the basic principles by which index numbers are
constructed.

2.0 OBJECTIVE
At the end of this unit, you should be able to:
- Define index number
- Calculate the index number through different methods
- Use(s) of index number
- Relevance of index number
3.0 MAIN CONTENT
3.1 Index Number Defined
In statistical analysis of one very large and important class of problems, we must
combine different set of data into a single measure e.g. we may wish to study the
behaviour of wholesale prices and to do this, we calculate an index number which
describes the changes, not in the various individual prices in which we are interested but
in the group of prices taken as a whole.
The relevance of this statistical device is shown by the fact that governmental and other
agencies devotes very substantial amount of money every year to the work of collecting
appropriate data performing the necessary calculations for the construction of index
numbers. The most widely known of these measure is the consumer price index or cost of
living index.

In general, index numbers are used in the study of prices (wholesale, retail, form, export
etc), output (manufacturing mining). The purpose of such measures is to get a summary
of a whole range of similar activities, thereby, one will be able to investigate problem on
relatively broad basis.

SELF-ASSESSMENT EXERCISE
What is the basis for an index number?

3.2 Computation of Index Number

Under this subsection we are going to look at the different index number available, how
the index number can be calculated through different methods.

Price Relative Index Number

This method is usually in use where just one commodity is involved. It measures the rate
of change in single commodity.
 Pn
Price relative = x 100
Po

Where Pn refer to price of current year and Po represents the price of the base period or
reference period.

Simple Price Index Number

Simple price index number is defined as the sum total of the price of related items
divided by the sum total of its base or reference period.
Pn
SPI = x 100
Po

Weighted Price Index Number

Here the concept weight is introduced to index number. These weights indicates the
importance of the particular commodity depending on whether we use base year, given
year or typical year quantities denoted by Qo,Qn. We are going to look at the works and
Marshall edge-worth Laspeyres, Paasche and Fisher on index number.
Laspeyres gave its own index number as
LPI = Pnqo x 100
Poqo 1
LQI =∑ PoQn x 100
∑PoQo
Paasche gave its own as
PPI = PnQn x 100
PoQn

PQI = ∑ PnQn x 100

∑PnQo
Fisher’s Ideal Price Index
Fisher defined its own index number as the square root of the works of both Paashe and
Laspeyres.
 F = √(Laspeyres Index)x (Paasche Index)

= PnQo PnQo x 100

PoQo PoQn

Marshall Edge Worth Price Index

Marshall edge-worth defined its own index as
MEPI = Pn (Qo + Qn) x 100
Po (Qo + Qn)
MEQI = Qn (Po + Pn) x 100
Qo (Po + Pn)

Where MEPI = Marshall edge-worth price index and

MEOI = Marshall edge-worth quantity index

3.3.0 Worked Example

Given the following about Open University, you are to compute the various price index
numbers for 1990 using 1986 as base year.
Table M4.2.1
Table Showing Information about Open University
Commodity Quantities Prices
1986 1990 1986 1990
A 30 70 75 360
B 40 100 160 300
C 50 150 250 960
D 15 33 180 291
Solution
From the above 1986 values is the base year which represents the P o values and Pn value
is represented by 1990 values.
Table M4.2.2
Laspeyre, Paasche and Fisher’s Table of Analysis
Price Quantities
Comm. 1986 1990 1986 1990
Po Pn Qo Qn PnQn PoQo PnQo PoQn
A 75 360 30 70 25,200 2,250 10,800 5,250
B 160 300 40 100 30,000 6,400 12,000 16,000
C 250 960 50 150 144,000 12,500 48,000 37,500
D 180 291 15 33 9,603 2,700 4,365 5,940
665 1,911 208,803 23,850 75,165 64,690

𝑃𝑛
SPI = x 100
𝑃𝑜
1911
= x 100
665

= 287.36
This is simply implying that cost of the commodity had risen by 287.4% between 1986
and 1990.
LPI = pnqo x 100
poqo
75,165
LPI = x 100
23,850

LPI = 315.157

𝑃𝑜𝑄𝑛
LQI =
𝑃𝑜 𝑄𝑜
 64,690
LQI =
23,850

LQI = 271.2369
≅ 271%
Using Laspeyres price index it is showing the rate of rise in price as by 315.16% between
1986 and 1990.
Where LPI = Laspeyres Price Index
LPI = Laspeyres Quantity Index
Paasche method
𝑃𝑛 𝑄𝑛
PPI = x 100
𝑃𝑜 𝑄𝑛
208,803
PPI = x 100
64,690

PPI = 322.77%

PQI = pnqn x 100

pnqo

208,803
PQI =
75,165

PQI = 277.79
PQI ≅ 278%
Using Paasche price index the rate of increase in price is 322.8% between 1986 and 1990.

Fisher’s Ideal Price Index

FPI = √LPI x P. P I
= √315.16 x 322.77
= √101724.1932
= 318.942

FQI = √LQI x PQI

= √271.2369 x 277.79286
FQI = √75,347.67419
FQI = 274.495
≅ 275%

Marshall Edge-Worth Index Number

TableM4.2.3
Marshal Edge-Worth Table of Analysis
Po Pn Qo Qn Po+ Pn Qo + Qn Pn(Qo+ Qn) Po(Qo+ Qn) Qo(Po+ Pn) Qn(Po+Pn)
75 360 30 70 435 100 36,000 7,500 13,050 30,450
160 300 40 100 460 140 42,000 22,400 18,400 46,000
250 960 50 150 1201 200 192,200 50,000 60,500 181,500
180 291 15 33 471 48 13,968 8,640 7,065 15,543
283,968 88,540 99,015 273,493

Using Marshall edge-worth price index

M.E. Price Index = Pn (Qo + Qn) x 100
Po (Qo + Qn)
283,968
= x 100
88,540

= 320.722
≅ 320.7%
M.E Quantity Index = Qn (Po + Pn) x 100
Qo (Po + Pn)
273,493
MEQI =
99,015

MEQI = 276.213
MEQI = 276%

SELF-ASSESSMENT EXERCISE
Define Fisher’s Ideal formular for price index?
4.0 CONCLUSION
In the course of our discussion on this unit you have learnt about
- Definition of price index
- Calculation about:
Simple price index
Weighted price index; where we talked about
Laspeyres index number
Paasche index number
Fisher’s ideal index number
Marshall edge-worth index number

5.0 SUMMARY
Below is the summary of all the price indices we talked about in this unit.
Pn
Price relative = x 100
Po
Pn
Simple price index = x 100
Po
Pn Qo
Laspeyres index = x100
Po Qo
Pn Qn
Paasche index = x100
Po Qo

Fisher’s ideal price index = √(Laspeyre Index)x (Paasche′ s Index)

Marshall Edge-worth Index

MEPI = Pn (Qo + Qn)
Po (Qo + Qn)
MEQI = Qn (Po + Pn)
Qo (Po + Pn)
6.0 TUTOR MARKED ASSIGNMENT
Explain the weighted price index?

7.0 REFERENCES
- Adedayo, O.A. (2006): Understanding statistics. JAS Publishers, Lagos.

- Dawodu, A.F. (2008): Modern Business Statistics NIICHO Printing Works,

Agbor, Delta State.

- Edward, E.L. (1983): Statistical Analysis in Economic and Business. (2nd edition)
Houghton Mifflin Company, Boston.

- Esan, E.O. and Okafor, R.O. (2010): Basic statistical methods (Revised Edition)
Toniichristo Concept, Lagos.

- Olufolabo, O.O. and Talabi, C.O. (2002): Principle and practice of statistics.
HASFEM (NIG) Enterprises, Lagos.

- Owen, F. and Jones, R. (1978): Statistics. Polytech publishers Ltd, Stockport.

- Oyesiku, O.K. and Omitogun, O. (1999): Statistics for social and management
sciences. (2nd edition) HEBP, Lagos.