Professional Documents
Culture Documents
Quantitative Techniques
CERTIFIED PUBLIC
ACCOUNTANTS UGANDA
Paper- 2
LEVEL-1 STUDY
TEXT
2022-2023
Preface
CONTENTS
1
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
COMPLETE SYLLABUS
Chapter 1 Introduction to statistical data,
presentations & measures
Chapter2 Sampling and sampling methods
Chapter3 Data classification and tabulation
Chapter4 Measures of central tendency
Chapter5 Measures of dispersion
Chapter6 Introduction to probability
Chapter7 Decision making
Chapter8 Permutations and combinations
Chapter9 Discrete probability distribution
Chapter10 Continuous probability distribution
Chapter11 Estimation and hypothesis testing
Chapter12 Chi – squared tests
Chapter13 Control charts
Chapter14 Linear algebra and calculus
Chapter15 Linear Programming
Chapter16 Correlation and regression
Chapter17 Forecasting: Time series
Chapter18 Net work analysis
Chapter 19 Index numbers
FACILITATOR
MBAZIIRA HASSAN
hassanmbaziira@gmail.com
0703- 859 320
2
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Chapter-1
Introduction to statistical data, presentation and measures
1.1 INTRODUCTION
Most of the quantitative methods mentioned in this course fall under the general heading of
statistics which is often used to refer simply to a set of data. For example, we can refer to a
Uganda's unemployment statistics (which might be presented in a table or chart showing the
country's unemployment rates each year for the last few years, and might be broken down by
gender, age, region and/or industrial sector, etc.).
However, we can also use the term "Statistics" (preferably with a capital letter) to refer to the
academic discipline concerned with the collection, description, analysis and
interpretation of numerical data. As such, the subject of Statistics may be divided into two
main categories; descriptive and inferential statistics.
B. STATISTICAL INFERENCE This is concerned with analyzing data and then interpreting
the results (attempting to go "beyond the data"). The main way in which this is done is by
collecting data from a sample and then using the sample results to infer conclusions about
the population. For example, prior to general elections in Uganda and many other
countries, statisticians conduct opinion polls in which samples of potential voters are asked
which political party they intend to vote for. The sample proportions are then used to
predict the voting intentions of the entire population. Of course, before any descriptive
statistics can be calculated or any statistical inferences made, appropriate data has to be
collected. We will start the course, therefore, by seeing how we collect data. This study
unit looks at the various types of data, the main sources of data and some of the
numerous methods available to collect data.
A. QUANTITATIVE VARIABLES, to which we shall restrict discussion here, are those for
which observations are numerical in nature. Qualitative variables have non-numeric
observations, such as color of hair, although of course each possible non-numeric value
may be associated with a numeric frequency.
B. A CONTINUOUS VARIABLE may take any value between two stated limits (which may
possibly be minus and plus infinity). Height, for example, is a continuous variable, because
a person's height may (with appropriately accurate equipment) be measured to any minute
fraction of a millimeter. A discrete variable however can take only certain values
occurring at intervals between stated limits. For most (but not all) discrete variables, these
intervals are the set of integers (whole numbers). For example, if the variable is the
number of children per family, then the only possible values are 0, 1, 2, ... etc., because it 3
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
is impossible to have other than a whole number of children. However in Uganda shoe
sizes are stated in half-units, and so here we have an example of a discrete variable which
can take the values 1, 1½, 2, 2½, etc. You may possibly see the difference between
continuous and discrete variables stated as "continuous variables are measured, whereas
discrete variables are counted". While this is possibly true in the vast majority of cases, you
should not simply state this if asked to give a definition of the two types of variables.
ADVANTAGES OF INTERVIEWING
It is the most appropriate method when studying attitudes, beliefs, values and motives of the
respondents. It is also appropriate in studies of irregular, private behavior as compared to say,
observation, it is very flexible, allows assessment of non-verbal behavior, ideal method to get
information from incapable people, i.e. those who cannot read and write, high response rate,
possible to get deeper into the details of the subject of the interview, possibility of cross
checking whether the information provided is correct, more complex questions can easily be
administered /asked through the interview method, e.g. questions full of graphs, charts and
detailed instructions, allows to attain spontaneous answers, i.e. non specific structured
answers, it is possible to study and exploit unknown events which are only known to the
respondents and the interview ensures that all questions are answered.
4
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
DISADVANTAGES OF INTERVIEWING
Very costly/expensive especially if samplers are large, it is time consuming because each
respondent needs his/her own time to be interviewed, too much reliance on memory because
the interviewer is expected to do a lot of things at the same time. Also the memory of the
respondents is not reliable. No opportunity to consult records, difficulty in persuading
respondents to respond as required, problem of finding trained personnel to conduct the
interview, personal emotions of the interviewer and the respondents, Language barrier, the
“interact” between the interviewer and the respondents may sometimes bias the answers, e.g.
race, religion, sex and age differences, etc.
Obviously, the interview method is time consuming and not suitable when we have large
groups of respondents. Here one drafts a questionnaire.
ADVANTAGES
Enables the researcher to study the behavior as it occurs. This guards against the difference
between reported behavior and actual behavior that are common when interviewing. Only
method to gather information about non-verbal behavior. There the researcher records the
silent features of behavior. Observation is free from non-response errors. Errors arising from
misinterpretation of question thus giving distorted answers. It is possible to adjust goals and
objectives of the study as data is collected which is not applicable in the interview or even the
questionnaire method. Research is carried out in the natural environment so the results got are
based on existing reality.
DISADVANTAGES
Time consuming. The researcher has to wait4 until the phenomenon/event occurs. Poor
methods when the researcher is interested in attitudes or opinion studies. It is a costly
technique in terms of man-hours spent in the field. Besides, it is tricky/difficult to delegate
research assistants in the observation method. Behavior that extends over a long period of
time is difficult to study under this method. Bias from the researcher or observer since he/she
may be influenced by particular moods or prevailing physiological status of the
researcher/observer. Lack of control on other natural factors that may affect our data. Only
deals with smaller sample sizes since it is not easy to observe large populations. Poor method
when studying sensitive topics issues, e.g. studies/research on sexual practices, abortion,
contraception, etc.
Scanning Published Data. When you examine published data from whatever source, it is
helpful to adopt the following procedure: (a) Overview the whole publication Flip through the
pages so that you get a feel for the document. See if it contains tables only, or if it uses
graphs and tables to describe the various statistics. (b) Look at the Contents pages. A study of
the contents pages will show you exactly what the document contains and give you a good
idea of the amount of detail. It will also show you which variables are described in the tables
and charts. (c) Read the Introduction. This will give a general indication of the origin of the
statistics in the document. It may also describe how the survey which collected the information
was carried out. (d) Look at part of the Document in Detail. Take a small section and study
that in depth. This will give you an appreciation of just what information is contained and in
what format. It will also get you used to studying documents and make you appreciate that
most tables, graphs or diagrams include some form of notes to help explain the data.
6
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
1.4 INTERNAL DATA SOURCES
All types of organization will collect and keep data which is therefore internal to the
organization. More often than not it applies to the organization where you work, but you
should not think of it as meaning just that type of organization. It is important when looking
for some particular types of data to look internally because;
It will be cheaper if the data can be obtained from an internal source as it will save the
expense of some form of survey. Readily available information can be used much more quickly
especially if it has been computerized and can be easily accessed. When the information is
available from within your own organization, it can be understood much more easily as
supporting documentation is likely to be readily available. Overall there are several advantages
from using internal data, although there is a tendency when using this type of data to make do
with something that is nearly right. Companies' annual reports provide a particularly useful set
of data for financial and business research.
The main producer of statistics in Uganda is Uganda Bureau of Statistics (UBOS). UBOS exists
primarily to service the needs of central government and other interested parties. However,
much of the information it produces is eminently suitable for use by the business community
as well, and indeed central government which is increasingly becoming aware of the need to
gear its publications so that they can be used by the business sector.
Also, the United Nations publications and websites are available, which cover world-wide
statistics in subjects such as population and trade. Most Ugandan companies also provide
useful financial data in their annual reports and accounts, most of which are now available
online
7
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Chapter-2
Sampling and sampling methods
2.1 INTRODUCTION
A sample is a subset of a population, a small sample is less than 30 elements and greater
than 30 elements is a large sample. By studying a sample, one can be able to draw conclusions
that are general to the entire population.
A population is the set of all the individuals or objects which have a given characteristic, e.g.
the set of all persons eligible to vote in a given country. Within an entire population, any single
member is called an element.
A study of all elements in the population is called a census where as a study of a few elements
in the sample is called sampling
Sampling is the process of selecting and inquiring from a representative fraction of the
population to make a conclusion about the whole population. Or sampling is a systematic
approach of selecting a few elements from the entire population.
Sampling unit is an item from which information is obtained. It may be a person, an
organization or an inanimate object such as a tyre. Sampling frame – a list of all the items in a
population
A. METHODS OF SAMPLING
PROBABILITY/ RANDOM SAMPLING TECHNIQUES
SIMPLE RANDOM SAMPLING
Every member of the population has an equal probability of being selected. Example; Crane
Bank manager wants to find out views about their bank services. Given the sampling
frame is the list of depositors in Crane bank. Everyone on the depositors list is given a unique
number from 1 to n, (n being the total number of people in the sampling frame). Each number
is now written on a slip of paper and put in a box. If you want a sample of a thousand people
you mix up these slips thoroughly and draw out a thousand slips. The numbers on these slips
then represent the people to be interviewed. In theory each slip would stand an equal chance
of being drawn out and so would have been chosen in a random manner.
Advantages
1. Always produces an unbiased sample.
2. Simple and easy to apply.
Disadvantages
1. Sampling units may be difficult or expensive to contact (dispersed people) 8
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
STRATIFIED SAMPLING
This is a probability sampling technique wherein the researcher divides the entire population
into different subgroups or strata, then randomly selects the final subjects proportionally from
the different strata.
It is important to note that the strata must be non-overlapping. Having overlapping subgroups
will grant some individuals higher chances of being selected as subject. This completely
negates the concept of stratified sampling as a type of probability sampling. Equally important
is the fact that the researcher must use simple probability sampling within the different strata.
The most common strata used in stratified random sampling are age, gender, socioeconomic
status, religion, nationality and educational attainment.
The sample size of each stratum in this technique is proportionate to the population size of the
stratum when viewed against the entire population. This means that the each stratum has the
same sampling fraction.
The important thing to remember in this technique is to use the same sampling fraction for
each stratum regardless of the differences in population size of the strata. It is much like
assembling a smaller population that is specific to the relative proportions of the subgroups
within the population.
Advantages – the advantage of this method is that the results from such a sample will not be
distorted or biased by undue emphasis on extreme observations.
Disadvantages – the main disadvantage is the difficulty of defining the strata. This method
can also be time-consuming, expensive and complicated to analyze.
QUASI-RANDOM SAMPLING
SYSTEMATIC SAMPLING METHOD
It is a type of probability sampling method in which sample members from a larger population
are selected according to a random starting point and a fixed, periodic interval. This interval,
called the sampling interval, is calculated by dividing the population size by the desired sample
size; I= (N/n). N is the population size and n is the sample size.
Example; Assuming the bank’s population of depositors is so large, he can’t use simple
random sampling. He will use systematic sampling. On his list of depositors, he might decide to
interview every 10th person on the list. As before, each member of the population (depositors) 9
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
will be given a number from 1 to n, the starting number is selected from a table of random
numbers by taking the first number in the table between 1 and 9. Say a 2 was chosen, then
the 2nd, 12th, 22nd, 32nd … etc. person would be selected from the sampling frame. This
method of sampling is often used as it reduces the amount of time that the sample takes to
draw. However, it is not a purely random method of selecting a sample, since once the
starting point has been determined, then the items selected for the sample have also been set
Advantages
1. Quick to select a sample.
2. Also it is sufficiently close to simple random sampling, in most cases, to justify its
widespread use.
Disadvantages
1. A major disadvantage occurs if the sampling frame is arranged so that sampling units
with a particular characteristic occur at regular intervals, causing over-representation or
under-representation of this characteristic in the sample. For example, if you are
choosing every tenth house in a street and the first randomly chosen number is 8, the
sample consists of numbers 8, 18, 28, 38 and so on. These are all even numbers and
therefore are likely to be on the same side of the street. It is possible that the houses
on this side may be better, more expensive houses than those on the other side. This
would probably mean that the sample was biased towards those households with a
high income. A sample chosen by systematic sampling must always be examined for
this type of bias.
MULTISTAGE SAMPLING
This refers to a sampling technique where sampling is carried out in stages using smaller and
smaller sampling units at each stage. Or a more complicated form of cluster sampling in which
larger clusters are further subdivided into smaller, more targeted groupings for the purposes of
surveying.
Despite its name, multi-stage sampling can in fact be easier to implement and can create a
more representative sample of the population than a single sampling technique. Particularly in
cases where a general sampling frame requires preliminary construction, multi-stage sampling
can help reduce costs of large-scale survey research and limit the aspects of a population
which needs to be included within the frame for sampling.
Advantages – the advantages of this method are that at each stage the samples selected are
small and interviews are carried out in 50 small areas instead of in 5,000 scattered locations,
thus economizing on time and cost. There is no need to have a sampling frame to cover the
whole country. The sample is effectively a simple random sample.
Disadvantages – the main disadvantages are the danger of introducing interviewer bias and
of obtaining different levels of accuracy from different areas. The interviewers must be well
chosen and thoroughly trained if these dangers are to be avoided.
10
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
NON- PROBABILITY/ NON-RANDOM SAMPLING TECHNIQUES
QUOTA SAMPLING
It is a non-probability sampling technique wherein the assembled sample has the same
proportions of individuals as the entire population with respect to known characteristics, traits
or focused phenomenon. OR; a type of non-probability sample in which the researcher selects
people according to some fixed quota. That is, units are selected into a sample on the basis of
pre-specified characteristics so that the total sample has the same distribution of
characteristics assumed to exist in the population being studied.
For example, if you are a researcher conducting a national quota sample, you might need to
know what proportion of the population is male and what proportion is female as well as what
proportions of each gender fall into different age categories, race or ethnic categories,
educational categories, etc. The researcher would then collect a sample with the same
proportions as the national population. In quota sampling, the researcher aims to represent
the major characteristics of the population by sampling a proportional amount of each.
Example; let’s say, for example, that you want to obtain a proportional quota sample of 100
people based on sex.
First you would need to find out the proportion of the population that is men and the
proportion that is women. If you found out the larger population is 40% women and 60%
men, you would need a sample of 40 women and 60 men for a total of 100 respondents. You
would start sampling and continue until you got those proportions and then you would stop.
So, if you’ve already got 40 women for the sample, but not 60 men, you would continue to
sample men and discard any legitimate women respondents that came along. You don’t need
them because you have already “met your quota.” The difficulty here is that you have to
decide in advance the specific characteristics on which you will base the quota. Will it be by
gender, age, education race, religion, etc.?
Definition; the method consists of dividing the sampling area into a number of small
concentrations or clusters of sampling units. Some of these clusters are chosen at random, and
every unit in the cluster is sampled.
For example, suppose you decided to carry out the bank survey using the list of all the
customers as the sampling frame. If you wished to avoid the cost of simple random sampling,
you could take each branch of the bank as a cluster of customers. Then you select a
number of these clusters randomly, and interview every customer on the basis of the branches
chosen. As you interview all the customers at the randomly selected branches, the sum of all
interviews forms a sample which is representative of the sampling frame, thus fulfilling your
major objective of a random sample of the entire population.
A variation of this method is often used in Uganda, because of the vast distances involved in
that country (often referred to as AREA SAMPLING). With the use of map references, the
entire area to be sampled is broken down into smaller areas, and a number of these areas are 11
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
selected at random. The sample consists of all the sampling units to be found in these selected
areas.
Advantages: Reduction in cost and increase of speed in carrying out the survey. The method
is especially useful where the size or constitution of the sampling frame is unknown. Nothing
needs to be known in advance about the area selected for sampling, as all the units within it
are sampled; this is very convenient in countries where electoral registers or similar lists do not
exist.
Disadvantages – one disadvantage is that often the units within the sample are
homogeneous, i.e. clusters tend to consist of people with the same characteristics. For
example, a branch of a bank chosen in a wealthy suburb of a town is likely to consist of
customers with high incomes. If all bank branches chosen were in similar suburbs, then the
sample would consist of people from one social group and thus the survey results would be
biased. This can be overcome to some extent by taking a large number of small clusters rather
than a small number of large clusters. Another disadvantage of taking units such as a bank
branch for a cluster is that the variation in size of the cluster may be very large, i.e. a very
busy branch may distort the results of the survey.
Quota sampling is a non-probability sampling technique wherein the assembled sample has
the same proportions of individuals as the entire population with respect to known
characteristics, traits or focused phenomenon. In addition to this, the researcher must make
sure that the composition of the final sample to be- used in the study meets the research's
quota criteria.
12
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Chapter-3
Data classification and tabulation
3.1 INTRODUCTION
Information collected from primary sources is often unorganized and can’t easily be
interpreted. As such, we have to organize and classify it.
3.2 DATA CLASSIFICATION is a process of arranging data into sequences and groups
according to their common characteristics (resemblance) or separating them into different but
related parts. Thus, all CPA paper5 students in my class can be classified according to their
marital status, height, religion etc. If students are classified according to marital status then all
married are put in one group and unmarried in another. If its classification of students heights,
then students who fall in a certain range (group) of heights will be put in one group etc
The bulk of such data must be reduced so that the eye can easily understand them. Here, we
represent any repetitions with a tally instead of re-writing them. The frequency will correspond
to the number of tallies for each particular salary. Thus, a Frequency distribution is a list of
values obtained in the data and the frequency with which these values occur in the data.
For grouped data like ages above, we can generate a frequency distribution as below;
13
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Age Tally Number of
workers
(Check the total:
10-19 //// / 6
6+11+14+10+7=2=50, this
20-29 //// //// / 11
is correct)
30-39 //// //// //// 14
40-49 //// //// 10
50-59 //// // 7
60-69 // 2
Total 50
It can be noted that we have two types of classes which must be taken seriously; inclusive
and exclusive.
Inclusive class intervals: Here the upper limit of the class interval is included in the
given class interval. These class intervals are used when the variable under consideration
happens to be a discrete one (i.e. takes on one figure at a time) e.g. the number of workers.
RELATIVE FREQUENCY
It’s the frequency of the class expressed as a percentage of the total frequency. Relative
frequency = [(frequency / Total Frequency) x 100%]
Class Limits
These are the smallest and largest values or observations that do fall in a given class. The
smallest value is the lower class limit and the largest value is the Upper class limit. e.g. (60 -
62) is the class interval, then 60 is the lower class limit and 62 is the upper class limit.
Case-1. Relative cumulative frequency less than: 68% of the people are below 50 years
of age and so on
Case-2. Relative cumulative frequency more than: …………. of the people are above 40
years of age, and so on. 15
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
3.3 DATA PRESENTATION
After classifying and tabulating our data, our next task is to present it so that it is easily
understood by management.
There are 4 ways of representing statistical data; Text presentation (data presented
in wordy form), Tabular presentation, Diagrammatic presentation, Graphical
presentation
16
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
C. DIAGRAMMATIC PRESENTATION. This involves bar charts, pictograms and pie charts.
Advantages of diagrams
1. Easy and attractive means of representing the dat
2. Facilitate comparisons
3. Save time and labor
4. Effective impressions
5. Have great memorizing value as compared to mere figures
Limitations of diagrams
1. They don’t give accurate results but rather a rough idea
2. Comparisons of diagrams can’t be made if the unit is not common or the phenomenon is not
the same
3. They can be misused easily
4. It’s expensive to draw diagrams.
BAR CHARTS
A SIMPLE BAR CHART
Makerere CPA Pass rates for June 2016 Pass Rate
Sitting
Paper-1 61%
Paper-2 88%
Paper-3 59%
Paper-4 56%
Paper-5 72%
17
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
Paper-1 Paper-2 Paper-3 Paper-4 Paper-5
200%
150%
100%
50%
0%
June.2011 Dec-11 Jun-12 Dec-12 Jun-13
Dec-12
Jun-12
Dec-11
June.2011
PICTOGRAMS use pictures or symbols to represent a number of units of data. The pictures
usually relate to the data shown
A PIE CHART is a circle, which is divided by radial lines into sections or subsections of
different angles so that the area of a particular section is proportional to size of the figure
represented.
Makerere University CPA Intake figures for June Number
2016
Paper-1 66
Paper-2 88
Paper-3 93
Paper-4 102
Paper-5 149
19
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Paper-1, 66
Paper-5, 149
Paper-2, 88
THE HISTOGRAM
It’s a plot of class frequencies against class boundaries. Its body comprises of bars of equal
width whose height represents the class frequency. Unlike in a bar graph, the bars in a
histogram are attached to each other and the width of the rectangle corresponds to the class
width and the height to the class frequencies.
25
20
Frequency
15
10
0
10.-14 15-19 20-24 25-29 30-34 35-39
Test Marks
20
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
We can also construct a histogram for UNEQUAL CLASS WIDTHS by calculating a frequency
density. Frequency density = (frequency / class width)
Example A group of students ran a race. Their times are recorded in the table:
Two class intervals are added to the distribution one at the beginning and the other at the end
of the distribution for purposes of closing the polygon. From the above table, the frequency
polygon is as follows:-
30
25
20
15
10
0
14.-15 15-19 20-24 25-29 30-34 35-39
A LINE GRAPH
This may be used to show changes in some economic variable, say, steel production over time.
In other words, if out of the two variables, one happens to be time (months, years, etc.), we
get a line graph over time or simply time series graph or histogram. A time series
expresses behavior of an economic variable over time.
22
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Example
The table below shows daily temperatures for Kampala City, recorded for 6 days, in degrees
Fahrenheit.
Temperatures In Kampala City
Day Temperature
1 43° F
2 53° F
3 50° F
4 57° F
5 59° F
6 67° F
The data from the table above has been summarized in the line graph below.
Z Chart
Z-charts: A chart often used in industry and constructed by plotting on it three series:
monthly, weekly, or daily data, the moving annual total, and the cumulative total dating from
the beginning of the current year
Z-chart: How to do it
1. Gather data on a regular basis for a given activity. This may be the sales per month, words
written by an author per week or calls handled by a customer response centre per day.
2. Identify the major review period to be considered. When you are gathering data on a
monthly cycle, then this is likely to be 12 months. If your minor review period is one day, then
the major review period is more likely to be something like one month or four weeks.
2. For each minor review period (one month in the examples here), build the longer-term view
by totaling the data for the past major review period (12 months in the examples here).
Note that the rolling total for 12 months ago will require data from 12 months previous to that.
Thus the longer-term view really does look back over a longer period.
3. For the current review (month, in the example here), sum the previous values over the past
major review period (12 months here).
4. Build the Z-chart, as in the diagram above. 23
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
The table below shows how these measurements can all be put into a simple table. You can
use a spreadsheet to automatically build the third and fourth column and also to subsequently
build the final Z-chart.
24
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
HAPPY HOUR: NOW TRY OUT THESE MODEL EXAMINATION NUMBERS
Question-1
According to the 1990 Uganda Statistical Abstract, 9,000,000 people were employed in 1989
by the government of Uganda. Of these 3,000,000 were employed in the education sector,
3,500,000 were employed in the civil service and the rest were in the armed forces. In 1988
the number employees reported were 8,000,000 of which 2,500,000 were in the education
sector, 2,000,000 were in the civil service and the rest were in the armed forces. The same
abstract gave the figure for the total number of government employees in 1987 as 7,500,000
distributed as follows: education sector 3,200,000, civil service 2,000,000 and 1,200,000 in the
armed forces. Present the above information in a tabular format.
Question-2
It has been estimated that the level of late coming of employees at Roofing Limited is affecting
the company’s performance. Given below is a record of the time in minutes of 60 employees
who come late on a particular day.
45 55 20 15 16 17 05 09 10 15
57 12 46 30 32 48 49 59 20 21
26 30 31 19 14 09 03 01 22 23
30 37 38 36 35 42 44 48 47 46
15 18 21 32 31 30 38 39 40 41
17 32 33 36 22 34 35 51 43 31
I. Beginning with a class interval of 0-9, 10-19 construct a frequency distribution of 6 classes
with uniform class width.
II. It the company’s policy is to write a warning letter for those who come late after 30 minutes,
what percentage of employees will be served a warning letter on that day?
III. Construct a histogram and estimate the modal time of late coming
IV. Extract a cumulative frequency curve
Question-3
Weight in kilograms 60 - 65 - 70 - 75 - 80 - 85 -
Number 5 18 42 27 8 4
25
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Chapter-4
Measures of location/ averages / central tendency
4.1 INTRODUCTION
A measure of central tendency is a sample value around which the distribution of data is
centered. Or a single value which can neatly characterize the whole group of data. Includes,
the mean, mode and median
X
f .x
f
Example1: The following set of raw data shows the salaries of CPA students from the
November 2019 class. The facilitator is arguing that on average everybody in the class gets 46
million per year. Is the lecturer correct?
Salaries (Millions) (fi)
0-10 2
10-20 8
20-30 12
30-40 18
40-50 28
50-60 22
60-70 6
70-80 4
Sum = 100
Solution
Salaries (Millions) f Mid-point (x) f .x
0-10 2 ….. ….
10-20 8
20-30 12
30-40 18
40-50 28
50-60 22
60-70 6
70-80 4
f f .x
x
f .x
f 26
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
4.3 ASSUMED/ WORKING MEAN: Is a method that uses an assumed mean in order to
calculate actual mean. This method does cut short the calculations that are more in the direct
method. The formula used for estimating mean of grouped data by using assumed mean
method is given below:
x A
f .d
f
Example: Assume any mean that lies between 40 -44 and calculate the actual mean using this
working mean method
x A
f .d
f
w x i i
xw i 1
n
w i 1
i
Note: when data values vary in importance, the analyst must choose the weight that best
reflects the importance of each data value in the determination of the mean.
Example: The following are 5 purchases of a raw material over the past 3 months.
Purchase Cost per Kilogram Number of Kilograms
($)(x) (w)
1 3.00 1200
2 3.40 500
3 2.80 2750
4 2.90 1000
5 3.25 800
Find the mean cost per kilogram. 27
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Purchase Cost per Number of (WX)
Kilogram ($)(x) Kilograms (w)
1 3.00 1200
2 3.40 500
3 2.80 2750
4 2.90 1000
5 3.25 800
SUM= SUM=
5
w x i i
.............
xw i 1
5
2.96
w
............
i
i 1
Marks x f f .Logx
25 – 29 27 2 2log27=……..
30 – 34 32 4 …….
35 – 39 .. 7 …….
40 – 44 …. 10 ……
45 – 49 …… 8 ……
50 – 54 …… 6 …….
55 – 59 ……. 3 ……..
Sum = fLogx .............
Therefore,
fLogx
Geometric _ Mean _ of _ X AntiLog =………………………….
f
4.6 HARMONIC MEAN (HM): Used when data given consist of a set of rates such as
Shs/kilogram, speed/hour, and output/man U hour etc. It is the reciprocal of the mean of
reciprocals of the item values.
Harmonic _ Mean
f
f
x
28
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Marks x f f
x
30-39 34.5 2
40-49 44.5 3
50-59 .. 11
60-69 .. 20
70-79 .. 32
80-89 .. 25
90-99 .. 7
f 100 f
x 1.4368
100
Harmonic _ Mean 69.60
1.4368
29
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Solution
The location of the median class is obtained by dividing the total number of observations
by 2: Therefore ( N / 2) (50 / 2) 25 th . The median lies in the 25th position. Then where
does the 25th term fall? This can be answered by determining the cumulative frequencies
for the data. When the table is closely examined, the 25th term falls in the fifth -class
interval since there are only 23 values in the first two class intervals. The median value lies
somewhere in the third-class interval (45-49).
Lm=45. And the cumulative frequencies up to but not including the median class interval is
CFb=23, Fm=18, C=5
30
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Therefore;
Median =
The median can also be obtained from the graph (OGIVE) as follows
Space to Insert an O-Give
Quartiles are those observations which divide an ordered data set into quarters (4 equal parts).
Qi Lq 4
n i Cf b
xC
fi
Quartile deviation = [(Q3 – Q1)/2] Also called the semi- inter-quartile range
Inter-quartile range = (Q3 – Q1)
Coefficient of quartile deviation = [(Q3 – Q1)/(Q3+ Q1)]
Likewise, an Nth decile is a value of the item below which N/10 items lie.
Di Ld 10
n i Cf b
xC
fi
Example: From this data, compute the quartile deviation
32
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Practice question
Profits (Shs) 5- 9 10-14 15-19 20-24 25-29 30-34 35-39
Frequency 8 18 27 21 10 28 8
33
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
From the information given below, calculate the;
a) 90th and 10th percentile
b) Quartile deviation, inter-quartile range, & coefficient of quartile deviation =
c) What is the 10th decile in the data given
D1
Mode = Lm xC
D1 D2
Where, Lm is the lower class limit of modal class, D1 is difference between the largest
frequency & frequency immediately preceding it. D2 is the difference between the largest
frequency & frequency immediately following it. C is the modal class width.
12
Mode = 30 x10 36.32
12 7
34
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
We can also obtain the mode from the graph (Histogram)
Space to insert a histogram which we can use to obtain the mode
Histogram
The mode can also be obtained from the O-Give by graphically drawing a perpendicular
to the base from the point on the curve where the curve is steepest/ most nearly vertically.
But, the mode is often estimated from a histogram.
We can also estimate the mode from the formulae. Mode = 3 Median – 2Mean
35
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Chapter-5
Measures of dispersion
5.1 INTRODUCTION
These show the extent to which the values in a distribution are spread around their summary
figures like the Mean, Median and Mode.
Mean
fx 1225 15.3 : Thus MD 5.1Sales
N 80
36
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
5.2.4 COEFFICIENT OF MEAN DEVIATION
= (MEAN DEVIATION / MEAN)
f x x fx fx n fx 2 fx
2 2 2 2
OR s OR: s
2 2 2
s
n 1 n n nn 1
2
OR
2
N N N
Example: Consider a POPULATION of sales from Sadolin Paints in billions
Class X F (F . X) (F.X2)
0-4 2 1 2
5-9 7 14 98 …
10-14 12 23 276
15-19 17 21 357
20-24 22 15 330
25-29 27 6 162 …
Total 80 1225 Total =
x
f .x
f
Then proceed to obtain the variance as follows
fx fx
2 2
Var S 2
.......
N N
Note: If the examiner specifies that the data relates- to a sample; then we use n-1 as the
denominator as follows:
n fx 2 fx
2
s 2
.........
nn 1
37
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Computation of mean deviation, standard deviation and variance using the working
mean method
Disadvantages: It’s complicated to define and calculate, its value can be distorted by
extreme values.
38
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
USES OF MEASURES OF DISPERSION IN BUSINESS
To find out the reliability of an average, to control the variation of data from the central value,
to compare two or more sets of data from their central value, to obtain other statistical
measures for further analysis of data.
Example2: The number of employees and variance of wages per employee for two factories
are given below:
Factory A Factory B
N0. Of employees 50 100
Average wage per employee 120 85
Variance in wages per employee 9 16
I. In which factory is there greater variance in the distribution of wages per employee?
II. Suppose in factory B, the wages of an employee were wrongly written as 120 instead
of 100, what would be the corrected variance for factory B?
39
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
MEASURES OF SKEWNESS AND KURTOSIS
SKEWNESS When a frequency distribution is not symmetrical, it is said to be asymmetrical or
skewed. Thus, skewness is a measure of symmetry, or more precisely, the lack of symmetry.
The direction of variation of the data set (skewness) can be measured by the (Pearson’s
coefficient of Skewness: SK).
SK = [3(Mean - Median)]/ Standard Deviation. Where (-3 < SK < +3)
OR SK = (Mean – Mode)/ Standard Deviation
Interpretation
If SK = 0, then the set of observations is symmetrical and normally distributed. For a
symmetrically distributed population or sample, the (mean = median = mode). Half of all
measurements are greater than the mean, while half are less than the mean.
If SK > 0, then the set of observations is positively skewed (positive skewness): Also
implying that a greater proportion of the observations are less than or equal to (as opposed to
greater than or equal to) the mean. The presence of extreme observations on the right hand
side of a distribution makes it positively skewed. Mean > Median > Mode
If SK < 0, then the set of observations is negatively skewed indicating that more observations
are greater than or equal to the mean. The presence of extreme observations to the left hand
side of a distribution make it negatively skewed and the relationship between mean, median
and mode is: Mean < Median < Mode.
40
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
For moderately skewed distributions, we can get the relationship between the mode, the
median and the mean. In case we have one of the 3, we can compute the other using the
formulae:- Mode ═ 3(median) – 2(mean): An example skewed distributions is the wage
distribution in which we find that most of the wages are skewed to the right.
Advantage of SK: It is independent of the scale. Because (mean-mode) and standard deviation
have same scale and it will be canceled out when taking the ratio. Disadvantage of SK: It
depends on the extreme values.
Example: Comment on the level of skewness using Karl Pearson coefficient of skewness
Advantage: Skq does not depend on extreme values. Disadvantage: Skq does not utilize the
data fully.
Example: The following table shows the distribution of 128 families according to the number
of children.
41
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Find the Bowley’s coefficient of skewness
Solution
42
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
KURTOSIS: This is another measure of the shape of a distribution. Whereas skewness
measures the lack of symmetry of the frequency curve of a distribution, kurtosis is a measure
of the relative peakedness of its frequency curve. Various frequency curves can be divided into
three categories depending upon the shape of their peak. The three shapes are termed as
Leptokurtic, Mesokurtic and Platykurhc as shown in Figure below.
43
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
For normal distribution, k ≈ 0.25 i.e. mesokurtic
A. I f K< 0.25, the distribution is platykurtic
B. I f K> 0.25, the distribution is leptokurtic
Question 2
Weight in kilograms 60- 65- 70- 75- 80- 85-
Number 5 18 42 27 8 4
Question 3
a) A ‘standard deviation is a natural partner to a mean’ discuss the meaning of this statement
whilst highlighting any two properties of an arithmetic mean
b) A machine produces circular bolts and, as a quality control test, 250 bolts were selected
randomly and the diameter of their heads measured as follows:
44
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Diameter of head (cm) Number of components
0.9747 - 0.9749 2
0.9750 - 0.9752 6
0.9753 - 0.9755 8
0.9756 - 0.9758 15
0.9759 - 0.9761 42
0.9762 - 0.9764 68
0.9765 - 0.9767 49
0.9768 - 0.9770 25
0.9771 - 0.9773 18
0.9774 - 0.9776 12
0.9777 - 0.9779 4
0.9780 - 0.9782 1
I. Determine whether the customer is getting reasonable value if the label on the
circular bolt advertises that the average diameter of the head is 0.97642 cm.
II. What is the quartile deviation in diameters of the bolts?
III. Buyers have threatened to return any bolts below 0.9765 cm. What
percentages of bolts were returned to the factory?
Question-4
The following data relates to sales of Besigye enterprises limited
Profits (Shs) 5- 9 10-14 15-19 20-24 25-29 30-34 35-39
Frequency 8 18 27 21 10 28 8
Find; 1) the mode, 2) the median, 3) variance, 4) coefficient of variation, 5) mean deviation,
6) upper quartile and 7) lower quartile, 8) Coefficient of quartile deviation 9) 80% percentile
Question-5
a) Differentiate between the following terms as used in measures of dispersion:
i) Platykurtic versus leptokurtic and mesokurtic.
ii) Coefficient of skewness and coefficient of variation
iii) Quartile deviation and mean deviation
b) The following data relates to sales of Matete enterprises limited
Profits (Shs) 10-19 20-29 30-39 40-49 50-59 60-69 70-79 80-89
Frequency 6 M 4 N 8 P 2 Q
Cumm: Freq A 10 B 18 C 30 D 35
Find;
I) Fill in the missing values
II) Compute the Geometric Mean and 2) Harmonic Mean 3) Weighted Mean 4) Mean
deviation
III) Estimate the semi- interquartile range
IV) Compute Bowleys measure of skewness and interpret the value
V) Construct an O-Give and estimate the median profits
VI) Construct a Histogram and estimate the modal profits
45
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Question-6
The following data relates to sales of Besigye enterprises limited
Profits (Shs) 16-17 18-20 21-24 25-29 30-34 35-44 45-54 55-74
Frequency 4 73 185 104 34 33 22 26
Note the class width are not the same
Find;
I) Construct a histogram and find the mode
II) Compute the average sales
III) Compute the risk (Standard deviation) in sales
IV) Construct a line graph
Question-7
a) As part of a class assignment in Financial Accounting, Kamese has compiled the
following profits made by Mabreezy Ltd and Kyalya Ltd enterprises respectively (In
Millions of Uganda Shillings).
Year 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
Mabreezy Ltd 12 115 6 73 7 19 119 36 84 29
Kyalya Ltd 47 12 76 42 4 51 37 48 13 0
By computing the arithmetic mean and a coefficient of variation for profits of each company,
which of the two companies:- 1) Performed better 2) Was consistent in profitability?
YOU MUST BE FULL OF HAPPINESS AFTER UNDERSTANDING THIS TOPIC. NOW TRY OUT
THESE VERY SIMPLE PAST PAPER NUMBERS
EXTRACT THESE NUMBERS FROM THE QUESTION BANK
Sitting Number Sitting Number Sitting Number
June 2016 November 2013 December 2010
November 2015 June 2013 June 2009
August 2015 November 2012 December 2009
June 2015 June 2012 June 2008
November 2014 December 2011 December 2008
June 2014 June 2010
46
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Chapter-6
Introduction to probability
6.1 INTRODUCTION
All future events are uncertain to some degree. Example; That the present NRM government
will still be in power in a year’s time (given that it is not an election year) is likely, but far
from certain; that a conservative party, led by Ken Lukyamuzi will be in power in a year’s time
is highly unlikely, but not totally impossible. Probability theory enables the difference in
the uncertainty of events to be made more precise by measuring their likelihood on a scale.
Definition: Probability is the chance that an event will happen. Probabilities are
expressed as fractions or decimals. They range between 0 – 1 inclusive. Assigning a
probability of 0 means that something can never happen while 1 means that something will
always happen.
A. TYPES OF PROBABILITY
A. CLASSICAL/ A PRIORI PROBABILITY:
Definition1: Either, A type of probability based on the assumption that the outcomes of an
experiment are equally likely. So we obtain the probability of an event taking place by
dividing the number of favorable outcomes by the total number of total outcomes.
47
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Course taught
Grade Business Quantitative Operations
Techniques Management
A 7 8
B 9 10
C 11 12
D 6 9
O 5 8
Note this is a classic example of classical/ aprior probability that is got by dividing the number
of favorable events out of the total events.
NoOfFavourableEvents
That is to say; Probability Pr obability
TotalEvents
15
I. Received grade A : P( A) 0.176
85
38
II. Was in the Business Quantitative Techniques class P( A) 0.447
85
III. Was in the Business Quantitative Techniques class and received grade A
7
P( A) 0.0823
85
Examples:
1. A coin is tossed and the outcome is either a head or a tail.
2. A six sided die is rolled and any figure on the die can come on top but you don’t know
which one will appear on top.
3. A customer enters a shop and either purchases a shirt or does not.
4. Etc.
49
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
So in each of the above random experiments we can specify the outcomes. So the different
possible outcomes are called basic outcomes. The possible outcomes of a random experiment
are called the basic outcomes and the set of all basic outcomes is called the sample space
usually denoted by S.
C. Sample space “S”. It is a set of all possible outcomes of a random experiment; e.g. in
tossing a coin, the sample space is either a tail or head. (H, T). In many cases, we are
interested in some subset of the basic outcomes and not the individual outcomes. For
example if we roll a die we might be interested in whether the outcome is even i.e. 2, 4
or 6. These are evens. So an event is a:-
D. An event. An event “E” is any subset of the basic outcomes from the sample space.
OR: It is one or more of the possible outcomes of doing something, e.g. for tossing a coin, the
event is a head or tail.
Rule1. The probability of an Event E, occurring must be at least or, at most, 1, i.e.
0 P( E ) 1 . That’s to say; P( E ) 0
Rule2. If all possible events are mutually exclusive, the sum of their probabilities is equal to 1.
n
i 1
P( Ei ) 1
Question: Three individuals A, B and C have applied for the position of accounts assistant in
Nile Breweries. Candidate A and B have the same change of being hired. The probability of
candidate C being hired is twice that of candidate A. Find the probability that candidate hired
is: 1) A and 2) B
Answer 50
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Rule3. The general additional rule
If two mutually exclusive events (A&B) occur, the probability that one or the other will
occur equals the sum of their individual probabilities,
P( A B) P( A) P( B)
If they are non-mutually exclusive events then they can occur at the same time. The
probability of their union is: P( A B) P( A) P( B) P( A B) .
The addition law can be extended to cover more events, e.g. for three events A, B and C: P(at
least one of A or B or C occurring) P(AUBUC) and we find that;
P( A B C ) P( A) P( B) P(C ) P( A B) P( A C ) P( B C ) P( A B C )
Unless you like learning formulae, do not bother to remember this as you can always solve
problems involving three events more easily using the Venn diagram.
Example-1: Three Newspapers, Monitor, New Vision and Observer are published in Kampala.
Of the adult population of Kampala, 65% read Monitor, 40% read New Vision, 25% read
Observer, 20% read both Monitor and New Vision, 20% read both Monitor and Observer, 10%
read both New Vision and Observer and 5% read all three. What is the probability that an
adult selected at random reads at least one of the magazines?
Solution
Definition of probabilities
P (M) = 0.65 P (N) = 0.40 P (O) = 0.25
P(M N ) 0.20 P(M O) 0.20 P( N O) 0.10
P( M N O) 0.05
The probability of reading at least one of the newspapers is the Union of all the three sets:
From the above diagram; total people who read at least one of the newspapers is equal to:-
30%+15%+15%+5%+5%+15%, i.e. 85%. So; 85% of the adult population reads at
least one of the newspapers
Example-2: A researcher from the NRM secretariat mentioned that 60% of Ugandans read
New Vision, 40% read Daily Monitor and 30% read the East African. Further, he claimed that
20% read both Daily Monitor and the New Vision, 15% read both the New Vision and the East
African whereas 10% read both Daily Monitor and the East African and 30% read the New
Vision only. Use your knowledge of basic/ introductory probability to answer the following
question. If a person is selected at random, find the chance that:-
51
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
I) That person reads NONE the three Newspapers (ANS 0.1)
II) That person reads at least one of the three papers (ANS 0.9)
III) Estimate the number of people who read at least two papers if the total sample used
in this survey was 500 (ANS 175)
IV) Estimate the chance that if a person reads at least two newspapers, he also reads all
the three papers (ANS 0.1429)
Answer
Example-3: December 2009: QN 2b: In order to establish the reading habits of CPA(U)
students, a sample of 50 students was selected and asked to name the newspapers they read
regularly. The results obtained showed that 25 read The New Vision (N), 16 read The Daily
Monitor (M), 14 read The Observer (O); 5 read both N and M, 4 read both M and O, and 6
read both N and O; and 2 read all the three. Required: Find the probability that a student
selected in this sample reads: (i) at least one of the newspapers. (ii) Only one of the
newspapers. (iii) Only the New Vision
52
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Example-4; June 2010: QN 5a: In order to establish the competitiveness of Yellow
Telecom, Purple Telecom and Red Telecom mobile telephone companies in the Ugandan
market, a survey was conducted on a sample of customers selected from Kampala. The
results generated were as follows: 190 were Yellow Telecom customers, 205 were for Purple
Telecom and 260 for Red Telecom. 15 customers were for all the three networks, 55 were for
Yellow Telecom and Red Telecom, 100 for Yellow Telecom only, 135 were for Red Telecom
only. 180 of the people under study were not on any network. Required: Determine the: (i)
size of the sample. (ii) Probability that a customer is on Purple Telecom only. (iii) Probability
that a customer is on one network only. (iv) Probability that a customer is on two networks
only. (2 marks)
Example 5: In a sample of 50 girls and 50 boys, fair and dark hair occurs as follows.
If a child is selected at random, from the sample, what is the probability of selecting a girl OR
child with fair hair?
Solution
50 55
P(G) ; P( F )
100 100 ;
P(G F ) P(G) P( F ) P(G F ) 50 / 100 55 / 100 33 / 100 72 / 100
Example 6: At Steel Rolling Mills, the production manager has provided the following
chances/ probabilities in regard to the number of defects in monthly shipments to Juba.
a) What is the probability that there will be fewer than 3 defective parts in a shipment?
b) What is the chance that there will be at most 2 defective part in a shipment?
c) What is the chance that there will be at least 2 defective parts in a shipment?
Example: A department store manager at Sadolin paints has monitored the number of
complaints received per week about the poor customer service. The probabilities/ chances for
the number of complaints in a week, established by this review are shown in the following
table. If we let A be the event “There will be at least 1 complaint per week”. And B is the
event that “There will be fewer than 10 complaints per week.”
53
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
6.3 THE MULTIPLICATIVE LAW OF PROBABILITIES
If they are independent events, it means that their respective probabilities are unconnected to
each other. For example, probabilities obtained when a die and coin are tossed together.
Here,
P( AandB) P( A B) P( A) xP( B)
If A, B and C are independent events, then
P( AandBandC) P( A B C ) P( A) xP( B) xP(C )
This continues up to any number of events in the exam question
PB A
pB / A
Similarly, P A provided that P (A) is greater than 0
54
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Example 8: A simple question is given to three CPA QT students; Kapele, Kamese and Kavu.
The chances of them getting right are respectively; ½, 1/3, and ¼. Find the probability that
the problem will be solved by all the three students (ANS = ¾). And the probability that none
of the three students will solve it (ANS = ¼) Note: The number is easily solved using a
probability tree
Example 9: A bag contains 4 white and 6 black balls. Two balls are drawn at random. If the
balls are drawn at random one after the other without replacement, find the probability that
both balls are white (ANS= 2/15). Both balls are black (ANS=1/3). First ball is white and
second ball is black (ANS = 4/15). One ball is white, the other is black (ANS = 8/15). Note:
The number is easily solved using a probability tree
55
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Example 10: After failing in the the presidential elections, Mabreezy applied for a Job in two
different organizations (URA and MTN). The probability that he gets a job in the first
organization (URA) is (2/5). And the probability that he gets a job in the second organization is
(4/5). Required: Assuming that getting both Jobs is independent, find the probability that she
gets:
Probability of getting in URA = (2/5)
Probability of getting in MTN = (4/5)
i) Chance of getting both Jobs = Prob (URA) x Prob (MTN)
P( A B) (2 / 5) x(4 / 5) 8 / 25
Example 11; The just concluded election was a highly contested one with counter
accusations of voter bribery on each side. According to the observers, Rukungiri District was
the epi-center of this voter bribery episode. Records from the Electoral Commission indicate
that 50% of the total registered voters voted for Museveni, 30% voted for Besigye and 20%
voted for John Patrick Amama Mbabazi. The observers maintained that 50% of those who
voted for Mr. Museveni were simply bribed, 82% of those who voted for Besigye were bribed
and 65% of those who voted for Mbabazi were bribed.
Use your knowledge of basic probabilities (Multiplicative law and conditional probabilities) to
answer the following questions
I) If a voter is selected at random, determine the probability that he was bribed
(ANS: 0.626)
II) If a voter is selected at random, and found not to have been bribed, find the
probability that he / she voted for Besigye (ANS: 0.1444)
III) If a voter is selected at random and found to have been bribed, find the
probability that he/ she did not vote for Museveni (ANS: 0.6006)
56
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Example 12: An industry manufactures a particular type of light bulb from three departments
A1, A2 AND A3. 30% are manufactured by A1, 45% by A2, and 25% by A3. It was found that
the bulbs manufactured by A1, 1% were faulty. And 1% were also faulty in A2 and 2% were
faulty in A3. Suppose a bulb is selected and found to be faulty, find the probability that:-
I) It is one of A3’s bulbs (ANS: 0.4)
II) It is one of A1’s bulbs (ANS: 0.36)
III) It is one of A2’s (ANS: 0.24)
Example 13: In a week, a company either has good sales of 200units or bad sales of 100
units. In the first week, the probability of good sales is 0.6. In the second week, the probability
of good sales is 0.8 if the sales in the first week were good or 0.3 if the sales in the first week
were bad. What is the chance of total sales over both weeks being? (I) 400 units (answer;
0.48) (II) 200 units (answer; 0.0.28) (III) 300 units (answer; 0.24)
Example 14: in a factory, a certain type of sweets is packed into boxes of four different
production lines X1, X2, X3 and X4. Records show that a small percentage of boxes are not
packed properly for sale. 1% of X1, 3% from X2, 2.5% from X3 and 2% from X4 are
improperly packed. If the percentages of total output that have come from the production lines
are 35% from X1, 20% from X2, 24% from X3 and 21% from X4, determine the chance that a
box selected at random from the entire output is;
1) Not properly packed (ANS 0.0197)
2) Not properly packed and comes from X1(ANS 35/197))
3) Not properly packed and comes from X2 (ANS 60/197)
4) Not properly packed and comes from X3 (ANS 60/197)
5) Not properly packed and comes from X4 (ANS 42/197)
57
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
THE BAYE’S THEOREM
Previously, we noted that the conditional probability of an event is a probability obtained with
the additional information that some other event has already occurred. We used P(B|A) to
denoted the conditional probability of event B occurring, given that event A has already
occurred. The following formula was provided for finding P (B|A):
In addition to the above formal rule, the textbook also included this "intuitive approach for
finding a conditional probability": The conditional probability of B given A can be found by
assuming that event A has occurred and, working under that assumption, calculating the
probability that event B will occur.
In this part we extend the discussion of conditional probability to include applications of Bayes'
theorem (or Bayes' rule), which we use for revising a probability value based on additional
information that is later obtained. One key to understanding the essence of Bayes' theorem is
to recognize that we are dealing with sequential events, whereby new additional information is
obtained for a subsequent event, and that new information is used to revise the probability of
the initial event. In this context, the terms prior probability and posterior probability are
commonly used.
A prior probability is an initial probability value originally obtained before any additional
information is obtained.
A posterior probability is a probability value that has been revised by using additional
information that is later obtained.
That's a formidable expression, but we will simplify its calculation. See the following example,
which illustrates use of the above expression, but also see the alternative method based on a
more intuitive application of Bayes' theorem.
Example 15: The ministry of Health is conducting a study on cigars smoking. They went to
Mbale and discovered that 51% of the adults are males. This implies that 49% are females.
9.5% of sampled males smoke cigars, whereas 1.7% of females smoke cigars. One adult is
randomly selected. Find the probability that the selected subject is a male
Solution
Let's use the following notation:
58
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Let's now apply Bayes’ theorem by using the preceding formula with M in place of A, and C in
place of B. We get the following result:
Solution
We use the following notation:
A = ELT manufactured by Altigauge
B = ELT manufactured by Bryant
C = ELT manufactured by Chartair
D = ELT is defective
D = ELT is not defective (or it is good)
59
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
a. If an ELT is randomly selected from the general population of all ELTs, the
probability that it was made by Altigauge is 0.8 (because Altigauge manufactures 80% of
them).
b. If we now have the additional information that the ELT was tested and was found
to be defective, we want to revise the probability from part (a) so that the new
information can be used. We want to find the value of P(A|D), which is the probability
that the ELT was made by the Altigauge company given that it is defective. Based on the
given information, we know these probabilities:
Example 17: Suppose a day’s production schedule calls for 9000 items. Three machines each
with a daily capacity of 4000 items are available and the probability that an item is defective is
1, 2 and 4 percent of the machines A, B and C respectively. On a given day, 4000 items were
produced on A, 4000 on B, and 1000 on C. One item is selected at random and found to be
defective. What is the probability that it was produced on machine A?
Solution
A: Item was produced on machine A
B: Item was produced on machine B
C: Item was produced on machine B
D: Item is defective
Obviously, we want to solve P (A/D)
It is given that P (A) = P(B) = P(C) = 4000/9000 = 4/9
Also; P (D/A) = 1%, Also; P (D/B) = 2%, Also; P (D/C) = 4%
Thus, applying BAYES theorem,
ANS= 1/4
60
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Example 18: Up to now, Mr. Amama Mbabazi’s camp still maintains that all figures from tally
sheets were wrongly entered at Central tallying center in Nambole. They claim that Electro
commission hired only three data entry specialists to enter all information from all tally sheets
and declaration forms. According to this camp, Data entry specialist 1 processed 30% of all
declaration forms; specialist 2 processed 45% of all declaration forms while specialist 3
processed only 25%. The proportions of incorrectly entered declaration forms by data entry
specialist 1, 2 and 3 are 0.3, 0.5 and 0.2 respectively. Assuming that a random declaration
form is found to have been incorrectly entered, what is the probability that it was processed by
data entry specialist 1?
Example 19: If a machine is correctly set up, it will produce 90% acceptable items. If it
incorrectly set up, it will produce only 40% acceptable items. Experience shows that the 70%
of the setups are correctly done. Find the probability that a setup has been correctly done if
refer the set up; 1) the first item produced is accepted (ANS = 21/25), 2) the first two items
produced are accepted (ANS = 189/205), 3) the first three items produced are acceptable,
acceptable, and unacceptable respectively (ANS = 63/95)
61
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Example 20: There are three plants in a factory producing the same product independently.
Experience shows that 2%, 4% and 5% of the items produced by the respective plants are
defective. The daily outputs of the respective plants are 2000, 3000 and 5000 units
respectively. If an item is selected at random from the outputs of the three plants, and is
found to be defective, find the probability that it was produced by the second plant. (ANS =
12/41)
Example 21: Multiple Industries has four plants scattered around the city producing the same
homogeneous item at all plants. The first plant produces 30 per cent of the total production,
second plant 25 per cent, third plant 35 per cent and the fourth plant 10 per cent. The firm
has a single warehouse in the city for storing the finished product of all the plants without any
distinction. From the past performance records on the proportion of defectives, it has been
found that 5 per cent, 10 per cent, 15 per cent and 20 per cent from the items produced at
plants 1, 2, 3, and 4 respectively are defective. Before the shipment of the items to a dealer,
one unit is selected and found defective. What is the probability that the item was produced in
plant 3?
Example 22: When visiting a friend, John may go by road, air or rail. The probabilities of
using a road, air or rail are 0.3, 0.8 or 0.6 respectively. The corresponding probabilities of
arriving on agreed time are 0.2, 0.8 & 0.1 respectively. Find the probability of having used the
road given that he arrived on time. (ANS = 0.0789)
62
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Chapter-7
Decision-making
7.1 INTRODUCTION
Decision making is a process of selecting from a set of alternative courses of action, the best
option that is considered to meet the objectives of a decision problem more satisfactorily than
as judged by the decision maker. A good decision is always based on logic, uses data, applies
techniques of quantitative analysis and considers all possible alternatives.
As the three states of nature are exhaustive and mutually exclusive, these three probabilities
must sum to one (i.e. it is a certainty that the pound will be relatively weak, stable or strong).
These probabilities enable us to calculate the expected monetary values of each decision
alternative.
Expected Monetary Value (EMV): Given a set of n possible monetary values (say, x1, x2, x3,
…..xn) and n probabilities associated with each monetary value (say, (say, p1, p2, p3, …..pn),
the expected monetary value (EMV) is calculated as the sum of the products of the monetary
values and their associated probabilities (i.e.
EMV ( p1 x1 p2 x2 .....pn xn )
. For i 1,2,3......n 63
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
With regard to decision-making problems, the x's are the payoffs resulting from each decision
alternative for a given state of nature and the p's are the state of nature probabilities.
I.e. It’s the difference between the payoff under certainty and the payoff under risk
Where EVwPI is the expected value with perfect information?
Example: JPAM after bitterly losing an election wants to construct a milk plant in his home
village of Kinkizi. Below are the pay offs and associated chances of getting such pay offs
Multistage problems involve a sequence of several decisions and outcomes. It is possible for a
decision to be immediately followed by another decision. Decision trees are best for showing
the sequential arrangement. Suppose they will first decide whether to pay $4000 to conduct a
market survey. Survey results will be imperfect. Then they will decide whether to build a large
plant, small plant, or no plant. Then they will find out what the outcome and payoff are
Conduct the survey. If the survey results are positive, then build the large plant (EMV =
$141,840). If the survey results are negative, then build the small plant (EMV = $16,540)
Example; Mr. Ogga has won a jackpot of UGX 2 billion. He has to decide where to invest his
money in one of the following real estate projects: Building apartments, Building offices or
Building a warehouse. The annual payoffs are in UGX’ millions are presented in the table
below.
65
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
State of the Economy
Good Economy Bad Economy
Apartments 50 30
Decision Offices 100 -40
Alternatives Ware House 30 10
Assuming there is a 60% chance that the economy would be good in the above scenario, use
your knowledge of decision theory (especially decision trees) to advise Mr. Ogga where to
invest
Example 2; A Japanese automobile manufacture currently produces its bestselling U.S model
in Japan, but the relative strength of the Japanese Yen versus the U.S dollar has been making
the car very expensive for the U.S market. To ensure a lower and more stable price, the
company is considering the possibility of manufacturing cars for U.S consumers at one of its
American plants. Its payoff table, in millions of dollars, is given below:
66
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Example 3; A farmer intends to drill a well for his herd. In the past, only 80% of the wells
drilled in this area were successful at 12 meters of depth. On finding no water at 12M, some
farmers drilled further up to 18 meters but only 30% struck water at 18M. The prevailing cost
of drilling the first 10M is Sh500, 000 and drilling an extra meter costs Sh60,000. The farmer
knows that if he does not get his own well, he will have to pay one million shillings to get
water from a neighbor’s well. Draw an appropriate decision tree and determine the farmer’s
strategy under the Expected Value (EMV) approach.
Example 4; A company is considering a short manufacturing run for either of two new
products, a temperature sensor or a pressure sensor. The market for each product is known if
the products can be successfully developed. However, there is some chance that it will not be
possible to successfully develop them. Revenue of $1,000,000 would be realized from selling
the temperature sensor and revenue of $400,000 would be realized from selling the pressure
sensor. Both of these amounts are net of production cost but do not include development cost.
If development is unsuccessful for a product, then there will be no sales, and the development
cost will be totally lost. Development cost would be $100,000 for the temperature sensor and
$10,000 for the pressure sensor.
Example 5; A company has a wonder product of which it expects great things. At the
moment the company has two courses of action open to it: test-market the product or
abandon it. If they test it, it will cost Sh10 million and the response could be positive or
negative with respective probabilities of 0.6 and 0.4. If the response is positive the company
could either abandon the product or market it full scale. If it markets the product full scale the
outcome might be low, medium or high demand, and the net pay-offs are: a loss of Sh20
million, profit of Sh20 million and Sh100 million respectively. The outcomes have
corresponding probabilities of 0.2, 0.5 and 0.3. If the result of the test marketing is negative
and the company goes ahead and markets the product, estimated losses would be Sh60
million. If, at any point, the company abandons the product, there would be a net gain of Sh5
million shillings from the sale of scrap. You are required to draw an appropriate decision tree
and determine the company’s strategy under the Expected Monetary Value (EMV) approach.
Example 6: A trust officer, Mugisha for a major banking institution is planning the investment
of a $ 1 million family trust for the coming year. The trust officer has identified a portfolio of
stocks and another group of bonds that might be selected for investment. The family trust can
be invested in stocks or bonds exclusively, or a mix of the two. This trust officer prefers to
divide the funds in increments of 10 percent; that is, the family trust may be split 100 percent
stocks /0 percent bonds, 90 percent stocks /10 percent bonds, 80 percent stocks /20 percent
bonds, and so on. The trust officer has evaluated the relationship between the yields on the
different investments and general economic conditions. Her judgment is as follows:
1) If the next year is characterized by solid growth in the economy, bonds will yield 12
percent and stocks 20 percent.
2) If the next year is characterized by inflation, bonds will yield 18 percent and stocks 10
percent.
3) If the next year is characterized by stagnation, bonds will yield 12 percent and stocks 8
percent.
I) Formulate a payoff table where payoffs represent the annual yield, in dollars,
associated with the different investment strategies and the occurrence of various
economic conditions
II) Suppose that a leading economic forecasting firm projects P (solid growth) = 0.4, P
(inflation) =0.25, and P (stagnation) = 0.35. Use the expected value criterion to select
67
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
the appropriate strategy.
III) What is the expected value with perfect information?
Example 7: Define the term ‘expected monetary value’ and explain its usefulness in
decision making
Mukwano Industries is considering its strategy for the next five years. It has three possible
courses of action, 1, 2 and 3, which are described below:-
1. Expand the business by re-locating to a larger site. This is estimated to have a 70% chance
of success, which will increase the company’s profits by UGX 40 million, but there is a 30%
chance of failure, which will decrease the company’s profits by UGX 60 million.
2. Take over a small rival company. The probability of successfully taking over the rival
company is estimated to be 0.5, and the rival company’s financial statements suggest that
there would then be a 60% chance of increasing profits by UGX 50 million and a 40% chance
of increasing profits by just UGX 10 million. However, if the takeover attempt fails (probability
= 0.5), profits are estimated to fall by UGX 15 million.
3. Make no new investments. In this case, it is estimated that there is a 50% chance that
profits will remain unchanged, a 20% chance that profits will increase by UGX 10 million and a
30% chance that profits will fall by UGX1 million. Using a decision tree and calculating the
Expected Monetary Values, advise the company on the best strategy to follow.
c) One of the company’s directors argues that the company should adopt strategy3. He points
out that strategy 3 will at worst lead to a UGX 1 million fall in profits, while adopting strategy 1
could result in a UGX 60 million fall in profits and adopting strategy 2 could result in a UGX 15
million fall in profits. Comment on the director’s choice of strategy.
Solution
The expected monetary value of a business decision is the average return that can be
expected, taking into account probabilities. For any business strategy, the EMV is calculated by
multiplying the estimated value of the possible outcomes by their associated probabilities and
then summing. It is a useful measure in business as it allows decision-makers to compare
alternative decisions - the highest EMV being one criterion employed by decision-makers to
choose among alternative strategies.
a)
68
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Re-location: EMVI = (0.7 x 40) + (0.3 x (-60)) = 10
Takeover: EMVII = 0.5 [(0.6 x 50) + (0.4 x 10)] + (0.5 x (-15)) = 9.5
No new investment: EMVIII = (0.5 x 0) + (0.2 x 10) + (0.3 x (-1)) = 1.7
Based on the EMV criterion, the best strategy is Strategy I, i.e. to expand by re-location.
b) The director is clearly risk-averse and is adopting the ‘maximin’ criterion, i.e. choosing
Strategy III with the highest minimum outcome. This may be sensible in this example, as
making no new investment avoids the risk (present in the ‘relocation’ and ‘takeover’
strategies) of experiencing substantial falls in profit which could place the company in a very
difficult financial position. There may also be non-pecuniary factors which the directors
should take into account, such as the preferred location of the company and the level of
employment that may be created by each strategy. The environmental effects (externalities)
of the three strategies might also be important.
Example: Quality Chemicals Ltd is introducing a new HIV vaccine which requires testing at a
cost of Shs.20 million. The diagram below illustrates the decisions taken:
69
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Required:
Study the decision tree and evaluate the expected values of each unit course and give your
conclusion.
Solution
A. Drug Testing positive and marketing it
Expected
EMV= (0.8x0.2 x(600 20) 0.8x0.5x(300 20) 0.8x0.3x(400 20) 104 M
B. Drug testing positive and abandoning it
Expected EMV= 0.8 x(80 20) 48M
C. Drug testing negative and marketing it
Expected EMV= 0.2 x(400 20) 84 M
70
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Chapter-8
Permutations & Combinations
8.1 INTRODUCTION
What is a permutation? A permutation of n objects taken r at a time is a selection of r
objects from a total of n objects where (r ≤ n), where ORDER MATTERS. The word
permutation means a particular sequence or order of arrangement of things. For example, BAC
and CBA are both permutations of the first three letters of the alphabet. Permutation problems
are concerned with the number of possible sequences into which things can be arranged.
There is a basic principle governing such problems: "If one operation can be done in m ways,
and if a second operation can be done in n ways, then the two operations can be done in
succession in m times n different ways."
Example1. There are three different colored buses (red, yellow and green) which run
between two places. If I want to use a different colored bus for each direction, in how many
different ways can I make the double journey?
Applying the basic principle, we see that the first part of the trip can be done in three ways
(red, yellow and green), while the second part of the trip can be done in only two ways
(excluding the color already used). Thus the total number of different possible ways is 3x2= 6
AB BA CA DA
AC BC CB DB
AD BD CD DC
Here 4P2 = 12. This can be proved by the box method. Assume we have 2 boxes to fill- one for
the first position, and one for the second. The first box can be filled in 4 ways (A, B, C, D),
71
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
leaving 3 items over. The second box can now only be filled in 3 ways. As any item in the first
box can be associated with any item in the second box 4P2 = 4 x 3 = 12 ways
8.3 COMBINATIONS
When dealing with permutations, we are concerned principally with the order or sequence in
which things occur. Other problems occur in which we need to calculate the number of groups
of a certain size irrespective of their sequence. These are called combinations
The number of ways (combinations) of n things taken r at a time without following a particular
order is;
n!
n
Cr
(n r )! r!
72
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Example3 of selecting combinations of n things taken r at a time
What is the probability of getting 3 white balls in a draw of 3 balls from a box containing 5
white and 4 red balls?
Solution
Chapter-9 73
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Discrete probability distribution
9.1 INTRODUCTION
In the previous chapters, we began our development of probability to represent situations with
uncertainty outcomes. In this chapter, we use those ideas to develop probability models with
an emphasis on discrete random variables. Probability models have extensive application to a
number of business problems and some of these applications are covered here.
Discrete random variable: It is a random variable which can take on one countable number
at a time.
Example:
Number of computers in a store at any one moment
Number of defective products in a sample of 20 items from a large shipment to Juba
74
3 0.3
Required
(i) Find the: mean, Variance and Standard Deviation of the number of weekly breakdow
N0 0 1 2 3 4
P(x) 0.10 0.26 0.42 0.16 0.06 Total
xP (x) 0 0.26 0.84 0.48 0.24 1.82
x2 0 1 4 9 16
x 2
. px 0 0.26 1.68 1.44 0.96 4.34
75
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
EMV E ( x) x. p x 1.82
V ( x) x 2 . px E x (4.34 1.82 2 ) 1.03
2
N0 0 1 2 3 4
Costs (x) 0 1,500 3,000 4,500 6,000
P(x) 0.10 0.26 0.42 0.16 0.06 Total
xP (x) 0 390 1260 720 360 2730
x2 0 2250000 9000000 20250000 36000000
x 2 . px 0 585000 3780000 3240000 2160000 9,765,000
EMV E ( x) x. p x 2,730
V ( x) x 2 . px E x (9765000 2730 2 ) 2,312,100
2
1 k
forX 1,2,3,....6
P ( X x) kx
0....Elsewhere
Find:
I) The value of K
II) The expectation of K
III) The variance in the distribution
IV) The probability P( X 3) /( X 4)
76
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
c) A discrete random variable X is represented by the probability distribution
Example 2: Ruhama constituency was a very tricky voting battle ground for all candidates in
the previous presidential election. Being a very wise presidential candidate, Mr. Mabreezy had
complied a table of possible votes and the respective chances of getting those votes in that
district as follows:-
Compute the mathematical expected number of votes. Help Mr. Mabreezy compute the
Variance and Standard deviation in votes obtained
77
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
9.4 THE BINOMIAL PROBABILITY DISTRIBUTION
The binomial probability distribution is applicable in situations where an experiment, consisting
of a certain number of trials satisfies all of four conditions. These are:
1. The number of trials must be fixed and finite. This number is usually denoted by n.
2. Every trial must result in one or other of only two mutually exclusive possible
outcomes, which, for convenience, we usually label "success" or "failure". Examples
are:
(a) When we roll a die we get a six or we do not get a six.
(b) A product is either defective or not defective.
(c) A tossed coin comes down heads or tails.
(d) A child is either a boy or a girl.
We must, of course, define which event is the success before the term is used.
3. The probability of a success or failure at each trial must remain constant throughout
the experiment. The probability of a success is usually denoted by p, and that of a
failure by q.
4. The outcome of every trial must be independent of the outcome of every other trial.
For example, if a coin is unbiased, then the probability of obtaining a head on the tenth
time it is tossed remains 1/2, even if the previous nine tosses have all resulted in
heads.
Provided these four conditions all hold, the binomial distribution enables us to work out the
probability of any given number of successes or failures in a specified number of trials.
78
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
79
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
80
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
81
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
82
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
83
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
84
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
85
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
86
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
NOTE: WE CAN ALSO READ THESE FIGURES DIRECTLY FROM BINOMIAL TABLES
GIVEN IN AN EXAM
PART OF THE BINOMIAL TABLE
There is another very common use for the Poisson distribution, and that is as an approximation
to the binomial distribution. If n is large, the calculations for the binomial distribution are often
very tedious, even with a calculator. Fortunately, if we put m = np, the Poisson terms can, in
certain circumstances, be used instead of the binomial terms. The conditions for using the
Poisson as an approximation to the binomial are that:
1) n should be large;
2) p should be small, or q = (1-p) should be small.
Question 6
Question 7
a) In a family with 10 children, if the probability of having a male child is the same as that of
having a female child, what is the probability that: (a) 6 of the children will be boys, (b)
none will be a girl, (c) at most, 2 will be boys?
b) Calculate the probability that, for 6 telephone lines: (a) at least 1 of the lines is engaged
and (b) all 6 lines are engaged. When the probability of 1 line being engaged is 1/4.
Question 8
88
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Chapter-10
Continuous probability distribution
10.1 INTRODUCTION
In the previous subsection, we developed discrete random variables and their respective
probability distributions. Here we extend the probability concepts to continuous random
variables and probability distributions. Many economic and business measures such as sales,
investment, consumption, costs, and revenues are represented by continuous random
variables. In addition, measures of time, temperature, distance and weight also fall in this
category. Probability statements for continuous random variables are specified over
ranges. Example: The probability that sales of UTL will be between 140 and 196 millions for
March is a typical example.
If we take a continuous random variable, say the height of students, weights of people,
length of items cut by a machine, and plot the probability density function, one fact emerges
that most of these density curves resemble in shape. This common shape is bell shaped,
uni-modal with most of the density concentrated near the mode and very little
density lying towards the lower and upper extremes.
This relative frequency curve includes the relative frequency of all the values of the variable x,
from ∞ to +∞ and the area between the curve and the x-axis is one unit and so it is a
theoretical relative frequency distribution or a probability distribution. This uni-modal
symmetrical bell-shaped curve is of great theoretical and practical importance in statistical
work. It is called the normal distribution or sometimes the Gaussian distribution after the
scientist who developed its use for examining random errors of observation in experimental
work.
89
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
When we consider the relative frequency curves of continuous variables, we discover a similar
pattern in the measurements of a great many natural phenomena. For example, the frequency
curve obtained from the set of heights of 80 employees, used earlier in the course, is uni
modal with small frequencies in the tails. Later, we calculated the mean, standard deviation
and median of this set of data. Since we were dealing with a comparatively small sample, the
values of these measures were empirical, and it is reasonable to assume that the theoretical
relative frequency distribution (the probability distribution) deduced from the data would be
symmetrical and normal. The same type of frequency distribution is found in the populations of
dimensions of items from a factory production line and errors of observation in scientific
experiments. This wide range of application of the normal distribution accounts for its
importance in statistical inference.
Since σ = 1, the ordinates drawn in Figure above are 1, 2 and 2.5 standard deviations on each
side of the mean. You can see that values of z more than 3 standard deviations from the mean
are very unlikely to occur, and that 50% of the values of z lie below zero (the mean), and 50%
above zero. It can be calculated that:
I. About 68% of the distribution lies within 1 standard deviation of the mean.
II. About 95% of the distribution lies within 2 standard deviations of the mean.
III. About 99% of the distribution lies within 2.5 standard deviations of the mean. 90
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Since z is a continuous variable we cannot find the probability that z takes any exact value but
only the probability that it lies in a given range of values. In other words, the probability that
the value of z lies between any two given values is equal to the area under the curve between
the ordinates at these two values.
Example: Salaries of MPs are normally distributed with mean m = 23.56 Million and standard
deviation, σ = 4.55 Million. What proportion of MPs would have salaries of 14.45 Million or
less?
Example 1: Carrots entering a processing factory have an average length of 15.3 cm, and
standard deviation of 5.4 cm. If the lengths are approximately normally distributed, what is the
maximum length of the lowest 5% of the load? (I.e., what value cuts off the lowest 5 %?)
91
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Example 2: X is a normally normally distributed variable with mean μ = 30 and standard
deviation σ = 4. Find a) P(x < 40), b) P(x > 21), c) P (30 < x < 35)
Solution
Note: What is meant here by area is the area under the standard normal curve.
Example 3: A radar unit is used to measure speeds of cars on a motorway. The speeds are
normally distributed with a mean of 90 km/hr and a standard deviation of 10 km/hr. What is
the probability that a car picked at random is travelling at more than 100 km/hr?
Solution
Let x be the random variable that represents the speed of cars. x has μ = 90 and σ = 10.
92
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Illustration on the probability on the normal curve
We have to find the probability that x is higher than 100 or P(x > 100)
For x = 100 , z = (100 - 90) / 10 = 1
P(x > 90) = P(z >, 1) = [total area] - [area to the left of z = 1]
= 1 - 0.8413 = 0.1587
The probability that a car selected at a random has a speed greater than 100 km/hr is
equal to 0.1587
Example 4: The time taken to assemble a car in a certain plant is a random variable having a
normal distribution of 20 hours and a standard deviation of 2 hours. What is the probability
that a car can be assembled at this plant in a period of time a) less than 19.5 hours? b)
Between 20 and 22 hours?
Solution
Illustration on the probability on the normal curve
Example 5: The annual salaries of employees in a large company are approximately normally
distributed with a mean of $50,000 and a standard deviation of $20,000.
a) What percent of people earn less than $40,000?
b) What percent of people earn between $45,000 and $65,000?
c) What percent of people earn more than $70,000?
Solution
Illustration of the probability on the normal curve
93
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
a) For x = 40000, z = -0.5
Area to the left (less than) of z = -0.5 is equal to 0.3085 = 30.85% earn less than
$40,000.
c)For x = 70000, z = 1
Area to the right (higher) of z = 1 is equal to 0.1586 = 15.86% earn more than
$70,000.
Question 2: The time taken to assemble a car in a certain plant is a random variable having a
normal distribution of 20 hours and a standard deviation of 2 hours. What is the probability
that a car can be assembled at this plant in a period of time a) less than 19.5 hours? b)
between 20 and 22 hours?
Question 3: A professor has determined that the final averages in his statistics course are
normally distributed with a mean of 72 and a standard deviation of 5. He decides to assign his
grades for his current course such that the top 15% of the students receive an A. What is the
lowest average a student must receive to earn an A?
Question 4: The amount of tip the waiters in an exclusive restaurant receive per shift is
normally distributed with a mean of $80 and a standard deviation of $10. A waiter feels he
has provided poor service if his total tip for the shift is less than $65. Based on his theory,
what is the probability that he has provided poor service?
Question 5:
a) With illustrations, what do you understand by the following terms
I. Standard error of the mean
II. Two tailed test
a) Daily power consumption at Nile Breweries is normally distributed with mean 10,000
kilowatts and a standard deviation of 2,000 kilowatts. What is the chance that the
consumption of electricity on a given day is:
I) Greater than 13,000 kilowatts
II) Less than 8,000 kilowatts
III) Between 7,500 and 14,000 kilowatts
b) A total population of 700 students sat an examination for which the pass mark was 50.
Their marks are normally distributed. 28 students scored below 40 marks while 35
scored above 60 marks. 1) Compute the mean and standard deviation in marks. 2) By
using the computed mean and standard deviation, what is the chance that any student
passed the examination? 3) If the pass mark is reduced to 48, find the increase in the
number of students who passed 94
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
c) To investigate the average amount spent each year on home improvements in Uganda,
a random sample of 900 homeowners is selected. The sample mean is Uganda shillings
420 Millions and the standard deviation is estimated to be Uganda shillings 240 Millions.
I) Calculate 95% confidence intervals for the population mean.
II) What sample size would be required to estimate the population mean to within
Uganda shillings 10 Million with 95% confidence?
Question 6
a) List 4 properties of the normal distribution curve
b) Explain the difference between:
I. Type I and Type II error
II. Confidence interval and confidence limits
III. Null and alternative hypothesis
c) Kamese Jimmy, an employee of “Tuvawala enterprises” has estimated that the daily
demand for his products is approximately normally distributed with average sales of
twelve per day and standard deviation of four products. How many products must
Kamese stock in the morning so that not more than one chance in five of running out of
stock during the day?
d) His boss is not amused about his dwindling daily sales and has instructed him to write a
report about sales performance. After careful analysis, he has discovered that twenty
three percent of total daily sales are greater than sixty but only twenty one percent of
total daily sales are below forty. What average and standard deviation in the distribution
of sales should Kamese include in her report?
Question-7: The CPA registrar; has summarized examination results as shown below
The registrar knows that a candidate fails if he/she obtains less than 40 marks (out of 100),
while he must obtain at least 75 marks in order to pass with a distinction. Determine the mean
and standard deviation in the distribution of marks, assuming this to be normal
95
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
CUMMULATIVE NORMAL DISTRIBUTION TABLE (CHOICE- A)
96
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
NORMAL DISTRIBUTION TABLE (CHOICE B)
97
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
STUDENT’S T- DISTRIBUTION TABLE
98
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Chapter-11
Estimation and hypothesis testing
If a frequency distribution of these k sample means is now constructed, and the mean and
standard deviation of this distribution of the k sample means is found, the following features
will emerge.
I. The sample means are themselves random variables since the values are likely to vary
from sample to sample.
II. The means of these k sample means will be found to be equal to the true population
mean; µx. Therefore, it is possible to say that the average of the sample means equals the
population mean
III. The standard deviation of these k-sample means will be equal to; x x
/ n
. This is
called the Standard Error (SE) of the sample means. The SE assumes that the
population from which the sample is drawn is large and that the population standard
deviation is known.
IV. If the distribution of these k sample means will have the shape of the normal
distribution, this distribution of the sample means is called the sampling distribution of the
sample means (x bar).
This process is known as interval estimation. BUT how wide is this interval? is usually a big
problem. To solve this problem we specify a confidence interval or level of confidence.
This refers to the degree of confidence we have. That is to say; the probability that the actual
population parameter lies within a specified range about the sample statistics. The population
parameter is what we are trying to estimate (µ in this case) and the sample statistic is what we
know by measurement (x‾) in this case. It can also be noted that apart from the interval
estimate, a point estimate also exists in estimating values of the true population. A point
99
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
estimate is when a single sample statistic is taken as the estimate of the unknown population
parameter.
Confidence interval for a single POPULATION MEAN µ [Large sample & population
standard deviation is known]
We shall use a simple formula below;
xz [x z xz ]
n Which expands into; n n
The one which has the negative sign is the lower confidence limit and one with a plus sign is
the upper confidence limit.
There are three key assumptions made;
A. The populati
B. The sample analyzed is large, and
C. The sample is a random sample.
The new Kampala City Authority is trying to establish satellite centers in the remote outskirts of
Kampala to reduce traffic congestion. A researcher from the authority has taken a sample of
100 taxi commuters from Gayaza and established that the sample mean time to reach Kampala
daily was is approximately 35.8minutes. UTODA already knows that the population standard
deviation in commuting time is 20minutes. Help the authority researcher set a 95% & 90%
confidence limits for the time taken to reach Kampala.
Solution
Interpretation
The probability is 0.95 that the actual time taken to commute from Gayaza to Kampala daily
lies between 31.88 minutes and 39.73 minutes
After several complaints of under pay far below the legal Ugandan minimum wage, the
workers MP along with a team of researchers from several labor unions undertook a study on
sample of 100 employees and found that average pay was just UGX35,600 and population 100
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
standard deviation in pay of UGX12,500. If the salaries are assumed to be normally distributed,
help the MP obtain a 95% and 99% confidence limits for the true population mean pay and
interpret the results.
Solution
Walk to work campaigns have been ignited by the escalating cost of living yet salaries are
stagnant says DP president; Norbert Mao. To prove this, a Makerere Don from the Institute of
Social Research has carried out a study on a random sample of 1,600 households Arua town.
He has established a sample mean salary of UGX25, 000 with a standard deviation of UGX12,
000. (a) Calculate a 95 per cent confidence interval for the mean salaries. AND (b) what
sample size would be required to estimate the mean to within ± UGX 400 with a 95 per cent
confidence interval?
Solution
Confidence interval for a single POPULATION MEAN µ [Large samples & population
standard deviation is UNKNOWN]
Previously, we have been assuming that the sample standard deviation is known. This
assumption helped to compute the standard error of the sample mean. However, in most
cases, the population standard deviation is unknown and needs to be estimated from the
sample standard deviation. When this is the case, probabilities for the sampling distribution of
the sample mean are derived, not from the z distribution, but from a distribution called the
student’s t distribution. Definition: this is a similar to the z distribution, bell shaped and
symmetrical about the mean of zero but doesn’t have a constant standard deviation (which the
z has). Its standard deviation is based on sample standard deviation and varies inversely with
the size of the sample used. Thus, when the population standard deviation is UNKNOWN and is
estimated by the sample standard deviation, the relationship between the sampling distribution
of the sample mean and its standardized values is given by the student’s t-distribution
translation formulae.
x x s s
t( n1) [ x t( n1) x t( n1) ]
( s x / n ) AND the confidence Interval is; n n
Now repeat this problem assuming that only 16 consumers were surveyed in Kampala and 12
liked the drink. And in Jinja only 10 were surveyed and 8 liked the drink. Solution
Now repeat the same question assuming that only 15british and 10 German tourists had been
interviewed. Interpret your findings Solution
Hypothesis tests will be conducted on the following 4 population parameters which are usually
measures of location
( x )
A. Hypothesis tests on a single population mean
B. Hypothesis tests on a single population proportion ( )
C. Hypothesis tests on a difference between two population means (1 2 )
D. Hypothesis tests on a difference between two population proportions (1 2 )
Process / steps in hypothesis tests. To adequately carry out the tests, we follow a
standard five step procedure.
1. Formulate the null & alternative hypotheses
2. Determine the area of acceptance/ rejection (ie chose the level of
significance)
3. Compute the corresponding sample statistic
4. Compare the sample statistic to the area of acceptance
5. Conclude; i.e. accept or reject the null hypothesis
Steps in detail
The null hypothesis, H0: the sample belongs to the population.
The alternative hypothesis, H1: the sample does not belong to the population.
Definitions: A research hypothesis is a prediction of the outcome of a study. The
prediction may be based on an educated guess or a formal theory.
The null hypothesis states that the true population parameter value is equivalent to the
hypothesized population parameter (Stated as H0). The alternative hypothesis states that
the true population parameter is not equivalent to the hypothesized population parameter
(Stated as H1). It is always expressed in a manner which negates the null hypothesis.
The null hypothesis can be stated in one of the three ways depending on the nature of
hypothesis tested.
104
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
2. A ONE-SIDED hypothesis test [UPPER TAILED hypothesis test]
When a claim is made that the population parameter IS LESS THAN OR EQUAL to a
specified value and the test is to verify the specified value only, then the null hypothesis is
formulated as a one-sided (UPPER TAILED) test.
H0: Population parameter specified value (like 10.9km)
H1: Population parameter specified value (like 10.9km)
The null hypothesis will be rejected in favor of the alternative hypothesis if evidence points
towards the true population parameter being only significantly greater than the null
hypothesized value.
For a two tailed test, the rejection area is both below and above the null hypothesized
value of the population parameter.
105
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
For a upper tailed test, the rejection area is above the null hypothesized value of the
population parameter.
For a lower tailed test, the rejection area is below the null hypothesized value of the
population parameter.
Types of errors: Because the sample measures are variables, we run a risk of making the
wrong decision in two ways. We may reject Ho when it is true, making what is called a Type I
error, or we may accept Ho when it is false, making what is called a Type II error.
Accept H0 Reject H0
H0 true Correct decision Making a type I Error
H0 false Making a type II Error Correct decision
Thus; a TYPE I ERROR is the probability of rejecting the null hypothesis when in fact it’s true
(it is represented by the area under the sampling distribution over the area of rejection). A
TYPE II ERROR is the probability of accepting the null hypothesis when in fact it is false. The
type I error is called the level of significance and is represented by the symbol (alpha).
The level of significance is the area (probability) in the “tails” of a sampling distribution and
used to find cut off points that separate the area of acceptance from the area of rejection. This
level of significance in fact defines the likelihood of rejecting the null hypothesis when it is in
fact true (that’s to say; type I error).
106
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
The figure below illustrates the relationship between the level of significance and the area of
acceptance for any sampling distribution for a two tailed test
The level of significance (α - alpha) appears only in the upper tail of a sampling distribution for
a one sided (upper tailed) hypothesis test. Conversely, for a one sided (lower tailed)
hypothesis, the level of significance is contained only in the lower tail of the sampling
distribution. The shaded areas represented above show the level of significance.
There is no direct algebraic connection between the values of the risks of making the two
errors, but as one risk is decreased the other is increased. We use the Type I error in
hypothesis tests and the Type II error to find the power of the test, which is outside the scope
of this course. Having formulated the hypotheses, you must decide on the size of the risk of
making a Type I error that you are prepared to accept. For example, you may decide that a
5% risk is acceptable (making the wrong decision once in 20 times) or you may require only a
1% risk (wrong once in a 100 times). You then divide all the possible values of the sample
measure into two sets, putting (say) 5% of them in the critical or rejection region and 95% in
the acceptance region. The percentage chosen is called the level of significance. Next calculate
the value of the measure for the sample you are using, see which region it belongs to and
state your decision.
Example1 on hypothesis test for as SINGLE POPULATION MEAN [Large samples &
population standard deviation is known]
Dr. Tanga Odoi, the chairman Makerere University Staff Association claims that on average a
Ugandan university lecturer receives UGX2.85millions per month which is too little to
counteract the rising cost of living. Ministry of education officials wish to test this claim and
have conducted a study on a sample of 250 lectures from all the government universities
throughout the country. They have discovered that sample mean monthly pay is
UGX3.03millions. Assuming that the population standard deviation of monthly pay is
UGX1.00million, test the claim at 5% significance level that the average monthly pay is indeed
UGX2.85millions
Solution
Step1. Stating the hypotheses
H0: Mean salary = 2.85 millions
H1: Mean salary 2.85 millions
107
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Step2. Determine the acceptance and rejection regions (using the level of significance)
This is a two tailed test; 5% . This is a TWO TAILED TEST.
Illustration
Example2 on hypothesis test for a SINGLE POPULATION MEAN [Large samples &
population standard deviation is known]
Test the above claim at 5% that the monthly pay is more than UGX2.85millions
Solution
NOTE: This is a one sided test (UPPER TAILED TEST) because the claim that the monthly
wage is strictly greater than UGX2.85millions.
108
Step3. Compute the corresponding sample statistic
Step4. Compare the sample statistic to the area of acceptance
Test the above claim at 5% that the monthly pay is LESS THAN UGX2.85millions
Solution
Example3 on hypothesis test for as SINGLE POPULATION MEAN [Large samples &
population standard deviation is known]
Question1. Explain the difference between the following terms:
a. Interval estimate and a point estimate
b. A confidence interval and a random sample
c. A z and t-test
d. Small and large sample
Question2. A sample of 100 fluorescent light tubes from the Short Life Tube Company has a
mean life of 20.5 hours and a standard deviation of 1.6 hours. Test:
(a) At the 1% level whether the sample comes from a population with mean 23.2 hours.
(b) At the 5% level whether it comes from a population with mean 20.8 hours.
(c) At the 5% level whether it comes from a population with mean less than 20.8 hours.
Solution
Example4 on hypothesis test for a SINGLE POPULATION MEAN [Large samples &
population standard deviation is known]
I. Explain any 5 properties of a normal distribution
II. Explain the difference between the following terms
a. Type I and type II errors
b. Null and Alternative hypotheses
c. One tailed and two tailed hypothesis tests
d. Lower tail and upper tail tests.
III. A sample of 150 students had an average IQ of 112 with a standard deviation of 9.
(a) At what level of significance would you accept that this sample is taken from a student
population with average IQ of 110?
(b) At the 5% level would you accept that the student population had an average IQ greater
than 113?
Solution
Solution
H0: 1 2 0
H1: 1 2 0
109
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Note the z statistic to be used is given as;
x1 x 2
z 1.73
1 2
2 2
n1 n2
Assume that Alice wished to test whether POSTA’s mean delivery time is greater than EMS
delivery time. Test this claim at 5% level of significance.
110
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
p1 p2
z 3.175
ˆ (1 ˆ )
1 1 x x x x
n n ˆ 1 2 p1 1 ; p2 2
1 2
BUT;
n1 n2 ; n1 n2
If we are dealing with SMALL SAMPLES (n1+n2 is less than 30); we use the respective
statistics.
Example1 on small samples; the mean weekly sales of NOMI detergent was 146.3 Sackets
per store. After an extensive advertisement campaign, the mean weekly sales in 22 stores for
a particular week increased to 153.7 and showed a standard deviation of 17.2. Is the evidence
conclusive that the advertising campaign was successful? Test this at 5% significance level.
Solution
x x
t 1.92
x
s / n . Mind about the (n-1) degrees of freedom
n1 n2 2
12.1 INTRODUCTION
Previously, we learned how to test hypotheses using data from one or two samples. We used
one-sample tests to determine if a mean was significantly different from a hypothesized value
and two-sample tests to determine if the difference between two means was significant. These
tests are known as parametric tests, because they involve testing the parameters of a
population, such as the mean and proportions. They use the parametric statistics of
samples that come from the population being tested. To formulate these tests, we
make assumptions about the population, for example, that the population is normally
distributed. There are certain kinds of data that cannot be tested in this way such
as: data which was not collected in a random sample and therefore does not have a
normal distribution; ordinal data; ranked data; and data from more than two
populations. In business, we often encounter data of this type, such as:
For these types of data, it is necessary to use tests which do not make restrictive assumptions
about the shape of population distributions. These are known as non-parametric tests. Non-
parametric tests have certain advantages over parametric tests:
It is not necessary to assume that the population is distributed in the shape of a normal
curve or in any other specific shape
They can be used with data measured in ordinal or nominal scales, which is often the
type of data obtained from business surveys, particularly where self-completion
questionnaires are used
Generally, they are quicker to do and easier to understand; sometimes, formal ranking
or ordering is not necessary.
But non-parametric methods are often not as precise as parametric tests, because they use
less information. In this study unit, we are going to consider one of the most commonly used
non-parametric tests, called the chi-squared test (pronounced "ki-squared"). Critical values
in the chi-squared test are denoted by the symbol χ2. The chi-squared test is principally used
to determine:
112
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
12.2 THE CHI – SQUARED GOODNESS OF FIT
The chi-squared test can also be used to determine if there is a similarity – that is, a good fit
– Between the distribution of observed data and a theoretical probability distribution, such as
the normal distribution or the binomial or Poisson distributions. We first perform the chi-
squared test to establish whether there is a significant difference between our observed
distribution and the theoretical distribution we have chosen; this information then enables us
to decide whether the observed data is a sample from our hypothesized theoretical
distribution. The chi-squared goodness of fit test is therefore a useful tool for managers, who
often need to make decisions on the basis of statistical information. For example, a
maintenance manager at a factory may use information about the frequency of breakdowns to
decide how many engineers to deploy on each shift.
We will now consider each of these in more detail, taking as an example data collected on the
amount of money consumers spend on chocolate each week.
Example 1: In one of the random surveys, IPSOS consultancy services asked clients one
question ‘In general, how do you rate the quality of services provided by Ugandan businesses?’
The distribution of responses to this question was as follows:
Excellent 8%
Pretty good 47%
Only fair 34%
Poor 11%
UMEME Management wants to establish whether the results of this consumer survey apply to
their company. They randomly interviewed 207 consumers as leave UMEME offices in various
parts of the city. The key question asked was ‘How do you rate the quality of customer service
at our branches’. The results obtained were
Use the chi –squared goodness test to establish whether the observed frequencies of
responses from the company survey are the same as the frequencies that-would-be expected
based on the IPSOS national survey. Use a 5% level of significance
Solution
(a) Formulating the null and alternative hypotheses
113
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
H0: The observed distribution is the same as the expected distribution
HA: The observed distribution is NOT the same as the expected distribution
Response f0 fe f fe 2
f
e
Excellent 21 16.56 (
Pretty good 109 97.29
Only fair 62 70.38
Poor 15 22.77
Total 207 207 6.25
f 0 fe
2
2 6.25
fe
Illustration of the Chi Squared distribution, the computed and critical values
(e) Ascertaining whether the sample statistic falls inside the acceptance region
Because the computed X2 value of 6.25 falls in the acceptance region, we do not reject the null
hypothesis. Therefore, the data gathered on 207 customers indicates that the distribution of
114
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
responses of customers is not significantly different from the distribution of responses to the
IPOS Survey.
Example 2: Dairy farmers in Mbarara want to know whether sales of their milk are uniformly
distributed over a full year, so that they may plan for milk production and storage. A uniform
distribution means that the observed frequencies are the same in all categories. The farmers
gather data from one large supermarket and record it as follows
Solution
(a) Formulating the null and alternative hypotheses
Month f0 fe f fe 2
f
e
January 1610 1537.25
February 1585 1537.25
March 1649 1537.25
April 1590 1537.25
May 1540 1537.25
June 1397 1537.25
July 1410 1537.25
August 1350 1537.25
September 1495 1537.25
October 1564 1537.25
November 1602 1537.25
December 1655 1537.25
Total 18,447 18,447 74.37
f 0 fe 2
2 74.37
fe
115
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
(d) Determining the appropriate number of degrees of freedom
With four categories, K = 4: Thus, the degrees of freedom are V = K-1; V= 12-1 = 11
Therefore; ,v 2 0.01,11 24.725
2
Illustration of the Chi Squared distribution, the computed and critical values
(e) Ascertaining whether the sample statistic falls inside the acceptance region
Because the computed X2 value = 74.37 is greater than the critical Chi Squared value, there is
enough evidence to reject the null and accept the alternative that the distribution of milk sales
is not uniform
Example 3: Kifefe, a teller within DFCU acacia branch believes that the distribution of random
arrivals at his branch are Poisson distributed. He attempts to test this hypothesis by recording
the information below relating to observed frequencies of arrivals on a one minute interval
within the bank.
Number of arrivals Observed frequencies
0 7
1 18
2 25
3 17
4 12
5 5
Solution
116
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Number of arrivals Observed frequencies Arrival x Observed
frequencies
0 7 0
1 18 18
2 25
3 17
4 12
5 5
84 192
192
The mean = 2.3
84
With the value of lambda, and the Poisson distribution table / computed Poisson
probabilities, the supervisor can determine the probabilities of the number of arrivals in each
category. The expected probabilities can also be computed manually (Seen this already for X =
0, 1, 2, 3, and 4). Using these probabilities and the total of 84 from the observed data, she
computes the expected frequencies by multiplying each expected probability by the total 84.
Number of f0 fe f fe 2
arrivals
f
e
0 7 8.42 0.24
1 18
2 25
3 17
4 12
5 5 7.03 0.59
84 84 1.74
f 0 fe 2
2
1.74
fe
Illustration of the Chi Squared distribution, the computed and critical values
(e) Ascertaining whether the sample statistic falls inside the acceptance
region
Because the computed X2 value = 1.74 is not greater than the critical Chi Squared value of
9.488, so the supervisor’s decision is not to reject the null hypothesis. That is, she fails to
reject the hypothesis that the distribution of the bank arrivals is actually Poisson.
Example 5: Suppose that a researcher has developed an overall attitude scale to determine
how employees feel towards their company. In theory, the scores can vary from 0 to 50. The
researcher pretests the measurement instrument on a randomly selected group of 100
employees. He tallies the scores and summarizes them into the six categories shown below.
Solution
(a) Formulating the null and alternative hypotheses
H0: The attitude scores are normally distributed
HA: The attitude scores are NOT normally distributed
118
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
(c) Calculating the chi-squared statistic of the sample
To compute the calculated Chi squared value, the researcher must determine the expected
frequencies. For a normal distribution, he computes the expected frequencies by using the Z
table, the mean and the standard deviation. Thus, we need to first compute the mean and the
standard deviation
X
f .x .. 24.9
f ...
fx fx
2 2
2 7.194 ( SD)
N N
Therefore, we use these two statistics to compute the expected probabilities for each category
of score. Use the normal probability distribution Z scores as indicated below
X
Z
SD
The six probabilities do not have a sum of 1.000. Even though observed frequencies were
obtained only for the six categories, getting a score less than 10 or greater than 40 was also
possible. Because 0.500 of the probabilities lie in each half of a normal probability distribution
and utilizing the sum of the expected probabilities on each side of the mean, 24.9, the
researcher can obtain the probability of <10 category as -.5- (0.0646+0.1645+0.2517)=
0.0192. Similarly, he can obtain the probability of the >40 category; 0.5 –
(0.004+0.2571+0.1581+0.0629) = 0.0179
119
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
He obtains the expected frequencies by multiplying each excepted probability by the total
frequency (100) as shown below
As the <10 and >40 categories have expected values of less than 5, each must be collapsed
(combined) with the adjacent category. As a result, the <10 category becomes part of the 10-
15 category and the >40 category becomes part of the 35-40 category.
Category f0 fe f fe 2
f
e
10- under 15 11 8.38 0.82
15- under 20 14 16.45 0.36
20- under 25 24 25.57
25- under 30 28 25.71
30- under 35 13 15.81
35- under 40 10 8.08 0.46
2.44
f 0 fe 2
2 2.44
fe
(e) Ascertaining whether the sample statistic falls inside the acceptance
region
Because the computed X2 value = 2.44 is not greater than the critical Chi Squared value of
7.815, so the researcher does not reject the null hypothesis that these categories are normally
distributed
The Chi square test of independence can be used to analyze any level data measurement, but
it is particularly useful in analyzing nominal data.
Type of gasoline
Income Regular Premium Extra premium Totals
Less than 85 16 6 107
USD30,000
30,000-49,999 102 27 13 142
50,000-99,999 36 22 15 73
More than 15 23 25 63
100,000
Total 238 88 59 385
Using a level of significance of 10%, use the chi square test of impendence to ascertain
whether the type of gasoline preferred is independent of income level.
121
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Solution
(a) Formulating the null and alternative hypotheses
f0 fe f fe 2
f
e
85 66.15 5.37
16 24.46
6 ..
102
27
13
36
22
15
15
23
25 24.42
Total 70.78
f 0 fe 2
2 70.78
fe
Illustration of the Chi Squared distribution, the computed and critical values
(e) Ascertaining whether the sample statistic falls inside the acceptance
region
Because the computed X2 value = 70.78 is greater than the critical Chi Squared value of
16.812, the researcher rejects the null hypothesis. That’s to say, the type of gasoline preferred
is not independent of income.
Question 3: DFCU Bank undertook a study on staff preferences for a car leasing scheme in
three locations of Mbarara, Jinja and Kampala.
Question 4: It is thought that the ages of the club’s members may be related to their
preferred sports. To investigate this, the following table was constructed:
Preferred sports
Age Swimming Badminton Squash
Under 20 20 20 10
20 and under 40 10 30 20
Over 40 10 10 30
Calculate the statistic to test the hypothesis that age and preferred sport are independent.
2
Question 5: Is the transportation mode used to ship goods independent of the type of
industry? Suppose that data shown represent frequency counts of types of transportation used
by the publishing and computer hardware industries. Analyze the data by using the chi- square
test of independence to determine whether the type of industry is independent of the
transportation mode at 5% level of significance.
Transportation mode
Air Train Truck
Industry Publishing 32 12 41
Computer hardware 5 6 24 123
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Below is the chi-squared distribution table that appears in your examinations
124
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Chapter-13
Control charts
13.1 INTRODUCTION
It’s essential that all products produced in a factory meet the requirements of those who use
them. Every product possesses a number of elements that jointly describe its fitness for use.
These elements are called quality characteristics.
In most companies, it’s difficult to produce products that have no defects. If a product is to
meet the customers’ fitness for use criteria, it should be produced by a process that’s stable
(repeatable). The production process should operate in little variability around the target
dimensions of the products’ quality characteristics. Statistical Process Control is a powerful
collection of problem solving tools useful in achieving process stability and improving the
capability through reduction of variability.
The 7 major tools that are commonly used to reduce variability in production processes and
products include; the control chart, Histogram, Check sheet, Pareto chart, Cause and effect
diagram, Defect concentration diagram and the Scatter diagram. For this particular paper, we
shall concentrate more on the control charts.
Other kinds of variability due to; improperly adjusted machines, operator errors, defective raw
materials may occasionally be present in products. These are usually large and easily
noticeable. We refer to these sources of variability that are not part of the chance caused
pattern as assignable causes. A production process operating in the presence of assignable
causes of variation as operating out of control
The major objective of statistical quality control is to quickly detect the occurrence of
assignable causes of process shifts so that investigations of the process and collective action
may be undertaken before any non-conforming units are manufactured. Here we use control
charts to do the investigation.
Example.
Each control chart has a centerline, an upper control limit and a lower control limit. Data are
recorded on the control chart, and the control chart is examined for disturbing patterns or for
data points that indicate a process is out of control. Once it is determined that a process is out
of control, measures can be taken to correct the problem causing the deviation.
It is a graph of sample means computed over a series of small random samples over a period
of time. These means are average measurements of some product characteristic. For example,
the measurement could be the volume of fluid in a liter of alcohol, the thickness of a metal
sheet etc. The sample means are plotted on a graph that contains a centerline and upper and
lower control limits (UCL & LCL).
Therefore;
Upper Control Limit (UCL) X A2 R
Center Line (X Bar Bar)
X=
X
k
Lower Control Limit (LCL) X A2 R
Alternatively, you can still use control limits by computing standard deviations for each sample
Upper Control Limit (UCL) X A3 S
Center Line (X Bar Bar)
S=
S
k
Lower Control Limit (LCL) X A3 S
127
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Example 1: Given the data below: Is the process in control?
Number of samples
Observations 1 2 3 4 5 6 7 8 9 10
1 5 0 -15 -15 -5 10 10 -5 5 20
2 25 10 -10 0 5 -15 15 -5 5 -10
3 -25 0 0 -25 5 -10 5 5 -5 0
4 0 5 -10 -5 0 -10 -10 -15 -5 -5
5 15 0 0 5 -10 5 -15 10 10 5
The data given below are in 0.0001 units. Plot both the X bar and Range charts
Solution
First, compute the mean and ranges of all the 10 samples given as follows
Number of samples
Observations 1 2 3 4 5 6 7 8 9 10
Sample mean (X Bar) 4 3 -7 -8 -1 -4 1 -2 2 2
Sample range (R.) 50 10 15 30 15 25 30 25 15 30
Sample standard
deviation (S)
Note: A2 value corresponding to a sample of size 5 is obtained directly from the tables:
When using Standard deviation approach, A3 is also obtained directly from tables
128
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
The Range (R.) Chart
Upper Control Limit (UCL) =D R 3
Note: D3 and D4 corresponding to sample size are obtained directly from the table
From the previous number,
Upper Control Limit (UCL) = D R (0 x 24.5) 0
3
129
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Example; Given the following data. Construct a proportion chart (P)
Sample 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
N 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50 50
N0 out of 4 3 1 0 5 2 3 1 4 2 2 6 0 2 1 6 2 3 1 5
compliance
Proportion (P) 0.08 0.06 0.10
Solution
Some of the more commonly encountered functions in the business world include:
Linear, Polynomials, Multivariate, Logarithmic, Discrete vs continuous, step function and
exponential functions.
Note: The examiner is testing whether you can apply this knowledge of formulating two
simple equations and then solving them.
Example1; the sum of two numbers is 10. Their difference is 6. What are the two numbers?
x + y = 10 Write a system of equations to model the situation.
x – y = 6 x = one number
y = the other number 132
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
x+y= Add the equations to eliminate the y-term and then
10 solve for x.
+ x–y=
6
2x = 16
x =8
x + y = 10 Substitute the value for x into one of the original
8 + y = 10 equations to find y.
y =2
Example2; A theater sold 800 tickets for Friday night’s performance. One child ticket costs
$4.50 and one adult ticket costs $6.00.The total amount collected was $4,500. How many of
each type of ticket were sold?
Example3; When Kapele’s wife priced the pens at 110/-, 30 pens were bought. But when she
reduced the price to 50/-, the demand increased by 20 units. Required: If you assume that
the relationship between the price of these pens and quantity demanded is linear of the form
P= a + bQ, determine the equation for this relationship and compute the price when 22 pens
are bought
133
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Solution
Desirable linear equation:P= a + bQ
PriceQuantity
110/-30 Units
50/-50 Units
The two linear equations are110 = a + 30b
50 = a + 50b
On solving simultaneously; a=200 & b = - 3
Final linear equation= P= 200 - 3Q
When 22 pens are bought; the price is; P= 200 – 3 (22)= 134/-
Example4; A shirt is on sale for $15.00 and has been marked down 35%. How much was the
shirt being sold for before the sale?
Solution
Let’s start with defining p to be the price of the shirt before the sale. It has been marked
down by 35%. This means that 0.35p has been subtracted off from the original price.
Therefore, the equation (and solution) is,
So, with rounding it looks like the shirt was originally sold
for $23.08.
Example1: Gashumba has a small shop at Garden city shopping mall those sales ladies
leggings (q). He has cost his latest leggings from China as follows:
5 leggings cost Shs. 13,525
2 leggings cost Shs. 12,880 135
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
8 leggings cost Shs. 14,280
Assuming Gashumba’s cost of the leggings is a quadratic function in terms of q, obtain that
particular quadratic function and also establish the cost of 12 pants based on the same cost
equation
Solution
Desirable quadratic equation: Cost a Q 2 bQ C
Quantity Cost
5 Units 13,525/-
2 Units 12,880/-
8 Units 14,280/-
The three quadratic equations are 13525 25a 5b c.........i
12880 4a 2b c..........ii
14280 64a 8b c............iii
When equations (i) and (ii) are 645 21a 3b...........iv
solved simultaneously, we get:-
When equations (ii) and (iii) are 1400 60a 6b..............v
solved simultaneously, we get:-
When equations (iv) and (v) are a 6.11
solved simultaneously, we get:- b 172.22
When the values of (a) and (b) 13525 25(6.11) 5(172.22) c.
obtained above are substituted in c 12,511 .2
equation (i) above; we get (c) as:-
Cost 6.11 Q 2 172 .22Q 12511 .2
Thus; the cost of 12 leggings is Cost 6.11(12 2 ) 172.22(12) 12511 .2 15457 .68 /
Example 2: Solve x2 + 5x + 6 = 0.
This equation is already in the form "(quadratic) equals (zero)" but, unlike the previous
example, this isn't yet factored. The quadratic must first be factored, because it is only
when you MULTIPLY and get zero that you can say anything about the factors and
solutions. You can't conclude anything about the individual terms of the unfactored
quadratic (like the 5x or the 6), because you can add lots of stuff that totals zero.
So the first thing I have to do is factor:
x2 + 5x + 6 = (x + 2)(x + 3)
Set this equal to zero:
(x + 2)(x + 3) = 0
Solve each factor:
x + 2 = 0 or x + 3 = 0
x = –2 or x = – 3
The solution to x2 + 5x + 6 = 0 is x = –3, –2
137
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
HAPPY HOUR: NOW TRY OUT THESE NUMBERS
Solve these quadratic equations
1 x 2 2x 8 2 3 7x 2x 2 0 3 22t t 2 21
4 t 2 7 6t 5 5x 2 2 9x 6 6 p 2 5 17 p
7 y2 2y 4 8 2r 2 6r 3 9 3q 2q 2 7
10 20 3r 2 5r 11 5 2x x 2 12 3t 2 14t 5
f’(x) y’
dy d
[ f ( x)]
dx dx
RULES OF DIFFERENTIATION
1. The Constant Function Rule
The derivative of a constant function f (x) = k, where k is a constant, is zero.
Given f (x) = k, f (x) = 0
138
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Example
2. The Linear Function Rule
If y = a + bx
dy
b
dx
dy
e.g. y = 10 + 6x then dx 6
139
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Example
d du dv
(uv) v u
Or; for f (x) = (u v). f (x) = dx dx dx
Example
Find the derivative of y = (6x3-x) (10-20x)
Solution
Example
Find the derivative of 140
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Solution
dy dy du
x
dx du dx
Example
Find the derivative of the function: f (x) = (4x3 +7)5
Solution:
DIFFERENTIATION IN ECONOMICS
Application I
• Total Costs = TC = FC + VC
• Total Revenue = TR = P * Q
• = Profit = TR – TC
• Break-even: = 0, or TR = TC
• Profit Maximisation: MR = MC
142
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Example 3: A company manufactures and sells X units of a product per month. The monthly
cost and demand equations are given respectively as below
x
C 72,000 6 x and P 200 respectively and 0 x 6,000
30
Required
Find;
(i) The sales revenue function
Solution
Given that price function= x
P 200
30
It should be noted that sales revenue = (Price x Quantity)
Thus, the sales revenue x x2
P (200 ) x Re venu 200 x
30 30
(ii) The production level that will maximize the company’s profits and state the resulting
profit.
Solution
Sales Revenue= x2
Re venu 200 x
30
Cost function = C 72,000 6 x
Profit function= Total Revenue – Costs
Profit function= x2
200 x (72,000 6 x)
30
Profit function= x2
194 x 72,000
30
OR; Multiplying through by 30; Pr x 2 5820 x 2,160,000
At maximum profit d Pr
0; Thus;2 x 5820 0; x 2,910
dX
On the second derivative, d 2 Pr
2 0; Thus _ Maximum
dX 2
Hence; 2,910 Units maximize the company’s profits
Maximum Profit= Pr (2,910 ) 2 5820 (2,910) 2,160,000
= 6,308,100/- 143
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
HAPPY HOUR: NOW TRY OUT THESE SIMPLE NUMBERS
Question1: The daily profit, P, of an oil refinery is given by P = 8x − 0.02x2, where x is the
number of barrels of oil refined. How many barrels will give maximum profit and what is the
maximum profit?
Question 2: Given that a firm has a total revenue function of TR= 3300q - 26q2 and the
firms total costs are represented by q3-2q2+420q+750. Assuming that q > 0; find the
profit maximising function and also compute the maximum profit.
Question4.
If a company has a demand function; p = x2 – 24x+ 117. Establish the number of units of
production and the price at which total revenue will be maximized and hence find the total
maximum revenue.
YOU MUST BE FULL OF HAPPINESS AFTER UNDERSTANDING THIS TOPIC. NOW
TRY OUT THESE VERY SIMPLE PAST PAPER NUMBERS
EXTRACT THESE NUMBERS FROM THE QUESTION BANK
Sitting Number Sitting Number Sitting Number
June 2016 November 2013 December 2010
November 2015 June 2013 June 2009
August 2015 November 2012 December 2009
June 2015 June 2012 June 2008
November 2014
June 2014
December 2011
June 2010
December 2008
144
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Chapter-15
Linear programming
15.1 INTRODUCTION
Linear Programming: Problem Formulation
Linear Programming is a mathematical technique for optimum allocation of limited or scarce
resources, such as labor, material, machine, money, energy and so on to several competing
activities such as products, services, jobs and so on, on the basis of a given criteria of
optimality.
The term ‘Linear’ is used to describe the proportionate relationship of two or more variables
in a model. The given change in one variable will always cause a resulting proportional change
in another variable. The word ‘programming’ is used to specify a sort of planning that
involves the economic allocation of limited resources by adopting a particular course of action
or strategy among various alternatives strategies to achieve the desired objective. Hence,
Linear Programming is a mathematical technique for optimum allocation of limited or scarce
resources, such as labor, material, machine, money energy etc among competing products or
activities.
Optimize (Maximize or Minimize) Z = c1x1+ c2x2+ ……. cnxn. Where Z is the measure of
performance variable. x1, x2,…xn are the decision variables. And c1, c2, ……. cn are the
parameters that give contribution to decision variables. The constraints are the set of linear
inequalities and/or equalities which impose restriction of the limited resources.
MAXIMIZATION PROBLEMS
Example1. MAXIMIZATION Problem (Generating a linear programming problem &
solving it using the GRAPHICAL METHOD)
Compute the values of X & Y that maximize the above Linear Programming Problem (LPP)
Solution
The three inequalities in the curly braces are the constraints. The area of the plane that
they mark off will be the feasibility region. The formula "z = 3x + 4y" is the
optimization equation (Objective function). We need to find the (x, y) corner points of
the feasibility region that return the largest values of z.
The step is to solve each inequality for the more-easily graphed equivalent forms:
It's easy to graph the system. Obtain coordinates for each constraint indicated above
147
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
To find the corner points -- which aren't always clear from the graph – we pair the lines
(thus forming a system of linear equations) and solve:
y = –( 1/2 )x + 7 y = –( 1/2 )x + 7 y = 3x
y = 3x y=x–2 y=x–2
–( 1/2 )x + 7 = 3x –( 1/2 )x + 7 = x – 2
3x = x – 2
–x + 14 = 6x –x + 14 = 2x – 4
2x = –2
14 = 7x 18 = 3x
x = –1
2=x 6=x
y = 3(–1) = –3
y = 3(2) = 6 y = (6) – 2 = 4
corner point at (2, 6) corner point at (6, 4) corner pt. at (–1, –3)
So the corner points are (2, 6), (6, 4), and (–1, –3).
Somebody really smart proved that, for linear systems like this, the maximum and
minimum values of the optimization equation will always be on the corners of the
feasibility region. So, to find the solution to this exercise, I only need to plug these
three points into "z = 3x + 4y".
Solution
The question asks for the optimal number of calculators, so my variables will stand for that:
x: number of scientific calculators produced
y: number of graphing calculators produced
Since they can't produce negative numbers of calculators, I have the two constraints, x > 0
and y > 0. But in this case, I can ignore these constraints, because I already have that x > 100
and y > 80. The exercise also gives maximums: x < 200 and y < 170. The minimum shipping
requirement gives me x + y > 200; in other words, y > –x + 200. The profit relation will be my
optimization equation: P = –2x + 5y.
148
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
So the entire Linear Programming Problem is:
Maximize P = –2x + 5y, subject to:
100 < x < 200
80 < y < 170
y > –x + 200
When you test the corner points at (100, 170), (200, 170), (200, 80), (120, 80), and (100,
100), you should obtain the maximum value of P = 650 at (x, y) = (100, 170). That is, the
solution is "100 scientific calculators and 170 graphing calculators".
MINIMIZATION PROBLEMS
Example1. MINIMIZATION Problem (Generating a linear programming problem &
solving it using the GRAPHICAL METHOD)
In order to ensure optimal health (and thus accurate test results), a lab technician needs to
feed the rabbits a daily diet containing a minimum of 24 grams (g) of fat, 36 g of
carbohydrates, and 4 g of protein. But the rabbits should be fed no more than five ounces of
food a day. Rather than order rabbit food that is custom-blended, it is cheaper to order Food X
and Food Y, and blend them for an optimal mix. Food X contains 8 g of fat, 12 g of
carbohydrates, and 2 g of protein per ounce, and costs $0.20 per ounce. Food Y contains 12 g
of fat, 12 g of carbohydrates, and 1 g of protein per ounce, at a cost of $0.30 per ounce. What
is the optimal blend?
Solution
Since the exercise is asking for the number of ounces of each food required for the optimal
daily blend, my variables will stand for the number of ounces of each:
x: number of ounces of Food X
y: number of ounces of Food Y
Since I can't use negative amounts of either food, the first two constrains are the usual ones:
x > 0 and y > 0. The other constraints come from the grams of fat, carbohydrates, and protein
per ounce:
149
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Fat: 8x + 12y > 24
Carbs: 12x + 12y > 36
Protein: 2x + 1y > 4
Also, the maximum weight of the food is five ounces, so:
x+y<5
The optimization equation will be the cost relation C = 0.2x + 0.3y, but this time I'll be finding
the minimum value, not the maximum.
After rearranging the inequalities, the system graphs as:
(Note: One of the lines above is irrelevant to the system. Can you tell which one?)
When you test the corners at (0, 4), (3, 0), and (1, 2), you should get a minimum cost of sixty
cents per daily serving, using three ounces of Food X only.
MINIMIZATION PROBLEMS
Question1. MINIMIZATION Problem (Generating a linear programming problem &
solving it using the GRAPHICAL METHOD)
At the Uganda Wildlife Education center, the game warden wants all his animals to get a
minimum of 36 milligrams (mg) of iodine, 84mgs of iron, and 16mg of zinc each day. One feed
of y1 provides 3mgs of iodine, 6mg of iron and 1mg of zinc. A second feed y2 provides 2 mg of
iodine, 6mgs of iron and 4mgs of zinc. The first type of feed costs $20, the second costs $15.
Using the graphical method, advise the game warden on the least cost combination of feeds
that guarantees daily requirements.
The agricultural research institute suggested the farmer to spread out at least 4800 kg of
special phosphate fertilizer and not less than 7200 kg of a special nitrogen fertilizer to raise the
productivity of crops in his fields. There are two sources for obtaining these – mixtures A and
mixtures B. Both of these are available in bags weighing 100kg each and they cost Rs.40 and
Rs.24 respectively. Mixture A contains phosphate and nitrogen equivalent of 20kg and 80 kg
respectively, while mixture B contains these ingredients equivalent of 50 kg each. Write this as
an LPP and determine how many bags of each type the farmer should buy in order to obtain
the required fertilizer at minimum cost.
4. The most negative entry in the bottom row identifies a pivot column.
5. Calculate the quotients. The smallest quotient identifies a pivot row. The
element in the intersection of the column identified in step 4 and the row
identified in this step is identified as the pivot element. The quotients are computed
by dividing the far right column by the identified column in step 4. A quotient that is a zero,
or a negative number, or that has a zero in the denominator, is ignored.
6. Perform pivoting to make all other entries in this column zero apart from the
pivot element
7. When there are no more negative entries in the bottom row, we are finished;
otherwise, we start again from step 4.
8. Read off your answers. Get the variables using the columns with 1 and 0s. All other
variables are zero. The maximum value you are looking for appears in the bottom right
hand corner.
Solution: In solving this problem, we will follow the algorithm listed above.
1. Set up the problem. That is, write the objective function and the constraints.
Since the simplex method is used for problems that consist of many variables, it is not practical
to use the variables x, y, z etc. We use the symbols x1, x2, x3, and so on.
Let x1 = The number of hours per week Niki will work at Job I. 152
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
and x2 = The number of hours per week Niki will work at Job II.
It is customary to choose the variable that is to be maximized as Z.
The problem is formulated the same way as we did in the last chapter.
Maximize Z = 40x1 + 30x2
Subject to: x1 + x2 ≤ 12
2x1 + x2 ≤ 16
x1 ≥ 0; x2 ≥ 0
2. Convert the inequalities into equations. This is done by adding one slack variable
for each inequality. For example to convert the inequality x1 + x2 ≤ 12 into an equation, we
add a non-negative variable y1, and we get; x1 + x2 + s1 = 12. Here the variable s1 picks
up the slack, and it represents the amount by which x1 + x2 falls short of 12. In this
problem, if Niki works fewer that 12 hours, say 10, then s1 is 2. Later when we read off the
final solution from the simplex table, the values of the slack variables will identify the unused
amounts. We can even rewrite the objective function Z = 40x1 + 30x2 AS – 40x1 – 30x2 + Z =
0.
After adding the slack variables, our problem reads
3. Construct the initial simplex tableau. Write the objective function as the bottom
row.
Now that the inequalities are converted into equations, we can represent the problem into an
augmented matrix called the initial simplex tableau as follows;
x1 x2 s1 s2 Z C
1 1 1 0 0 12
2 1 0 1 0 16
–40 –30 0 0 1 0
Here the vertical line separates the left hand side of the equations from the right side. The
horizontal line separates the constraints from the objective function. The right side of the
equation is represented by the column C (which are the solution quantities). The reader needs
to observe that the last four columns of this matrix look like the final matrix for the solution of
a system of equations. If we arbitrarily choose x1 = 0 and x2 = 0, we get;
S1 S2 Z C
1 0 0 12
0 1 0 16
0 0 1 0
Which reads; s1 = 12, s2 = 16 and Z = 0
The solution obtained by arbitrarily assigning values to some variables and then solving for the
remaining variables is called the basic solution associated with the tableau. So the above
solution is the basic solution associated with the initial simplex tableau. We can label the basic
solution variable in the right of the last column as shown in the table below.
x1 x2 s1 s2 Z
1 1 1 0 0 12 s1
2 1
–40 –30 0
0 1
0
0
1
16
0
s2
Z
153
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
The most negative entry in the bottom row identifies a column.
The most negative entry in the bottom row is –40, therefore the column 1 is identified.
x1 x2 s1 s2 Z
1 1 1 0 0 12 s1
2 1 0 1 0 16 s2
–40 –30 0 0 1 0 Z
Question. Why do we choose the most negative entry in the bottom row?
Answer. The most negative entry in the bottom row represents the largest coefficient in the
objective function – the coefficient whose entry will increase the value of the objective function
the quickest.
The simplex method begins at a corner point where all the main variables, the variables that
have symbols such as x1, x2, x3 etc., are zero. It then moves from a corner point to the
adjacent corner point always increasing the value of the objective function. In the case of the
objective function: Z = 40x1+ 30x2, it will make more sense to increase the value of x1 rather
than x2. The variable x1 represents the number of hours per week Niki works at Job I. Since
Job I pays: Shs. 40 per hour as opposed to Job II which pays only Shs. 30, the variable x1 will
increase the objective function by Shs. 40 for a unit of increase in the variable x1.
5. Calculate the quotients. The smallest quotient identifies a row. The element
in the intersection of the column identified in step 4 and the row identified in this step is
identified as the pivot element.
As mentioned in the algorithm, in order to calculate the quotient, we divide the entries in the
far right column by the entries in column 1, excluding the entry in the bottom row.
x1 x2 s1 s2 Z
1 1 1 0 0 12 s1 12 ÷ 1 = 12
2 1 0 1 0 16 s2
–40 –30 0 0 1 0 Z
The smallest: of the two quotients, 12 and 8, is 8. Therefore row 2 is identified. The
intersection of column 1 and row 2 is the entry 2, which has been highlighted. This is our
pivot element.
Question: Why do we find quotients, and why does the smallest quotient identify a row?
Answer When we choose the most negative entry in the bottom row, we are trying to
increase the value of the objective function by bringing in the variable x1. But we cannot
choose any value for x1. Can we let: x1 = 100? Definitely not! That is because Niki never
wants to work for more than 12 hours at both jobs combined. In other words: x 1 + x2 ≤ 12.
Now can we let x1 = 12? Again, the answer is no because the preparation time for Job I is two
times the time spent on the job. Since Niki never wants to spend more than 16 hours for
154
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
preparation, the maximum time she can work is 16 ÷ 2 = 8. Now you see the purpose of
computing the quotients.
x1 x2 s1 s2 Z
1 1 1 0 0 12
1 1/2 0 1/2 0 8
–40 –30 0 0 1 0
To obtain a zero in the entry first above the pivot element, we multiply the second row by –1
and add it to row 1. We get;
x1 x2 s1 s2 Z
0 1/2 1 –1/2 0 4
1 1/2 0 1/2 0 8
–40 –30 0 0 1 0
To obtain a zero in the element below the pivot, we multiply the second row by 40 and add it
to the last row.
x1 x2 s1 s2 Z
0 1/2 1 –1/2 0 4 s1
1 1/2 0 1/2 0 8 x1
0 –10 0 20 1 320 Z
We now determine the basic solution associated with this tableau. By arbitrarily choosing x2
= 0 and s2 = 0, we obtain: x1 = 8, s1 = 4, and z = 320. If we write the augmented matrix,
whose left side is a matrix with columns that have one 1 and all other entries zeros, we get the
following matrix stating the same thing.
155
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
x1 s2 Z C
0 1 0 4
1 0 0 8
0 0 1 320
7. When there are no more negative entries in the bottom row, we are finished;
otherwise, we start again from step 4.
Since there is still a negative entry, –10, in the bottom row, we need to begin, again, from
step 4. This time we will not repeat the details of every step, instead, we will identify the
column and row that give us the pivot element, and highlight the pivot element. The result is
as follows.
x1 x2 s1 s2 Z
0 1/2 1 –1/2 0 4 y1
x1 x2 s1 s2 Z
0 1 2 –1 0 8
1 1/2 0 1/2 0 8
0 –10 0 20 1 320
Now to make all other entries as zeros in this column, we first multiply row 1 by –1/2 and add
it to row 2, and then multiply row 1 by 10 and add it to the bottom row.
x1 x2 s1 s2 Z
0 1 2 –1 0 8 x2
1 0 –1 1 0 4 x1
0 0 20 10 1 400
Z
We no longer have negative entries in the bottom row, therefore we are finished.
Question. Why are we finished when there are no negative entries in the bottom row?
Answer. The answer lies in the bottom row. The bottom row corresponds to the following
equation. 156
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
0x1 + 0x2+ 20s1 + 10s2 + Z = 400 or
Z = 400 – 20s1 – 10s2
Since all variables are non-negative, the highest value Z can ever achieve is 400, and that will
happen only when s1 and s2 are zero.
Thus; the company’s profits would increase by (4/3) Shs for a unit increase in the constant of
constraint 2. Likewise, profits would also increase by (1/2) Shs for a unit increase in the
constant of constraint3. And by 0 for a 1-unit increase in the constant of constraint 1(OR-The
addition of still another unit would add nothing to the profit function-marginal value is 0).
H3 MINIMIZATION PROBLEMS
If the simplex algorithm is used for minimization problems, the negative values generated by
the surplus variables present a special problem. The first basic solution will consist of totally
negative numbers and so will not be feasible. Consequently, still other variables, called artificial
variables must be introduced to generate an initial feasible solution. It is frequently easier to
solve minimization problems by using the DUAL.
1 2 40
1 1 30
12 16 0
Observe that this table looks like an initial simplex tableau without the slack variables. Next,
we write a matrix whose columns are the rows of this matrix, and the rows are the columns.
Such a matrix is called a transpose of the original matrix.
We get;
1 1 12
2 1 16
40 30 0
The following maximization problem associated with the above matrix is called its dual.
Maximize Z = 40y1 + 30y2
Subject to: y1 + y2 ≤ 12;
2y1 + y2 ≤ 16;
y1 ≥ 0; y2 ≥ 0
We have chosen the variables as y's, instead of x's, to distinguish the two problems.
158
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
We have plotted the graph, shaded the feasibility region, and labeled the corner points. The
corner point (20, 10) gives the lowest value for the objective function and that value is 400.
Again, we have plotted the graph, shaded the feasibility region, and labeled the corner points.
The corner point (4, 8) gives the highest value for the objective function, with a value of 400.
We observe that the minimum value of the minimization problem is the same as the maximum
value of the maximization problem; they are both 400. This is not a coincident. We state the
duality principle. The Duality Principle: The objective function of the minimization problem
reaches its minimum if and only if the objective function of its dual reaches its maximum. And
when they do, they are equal. Our next goal is to extract the solution for our minimization
problem from the corresponding dual. To do this, we solve the dual by the Simplex method.
Example13. Use the simplex method to solve the minimization problem below by solving its
dual. Minimize Z = 12x1 + 16x2
Subject to: x1 + 2x2 ≥ 40; x1 + x2 ≥ 30; x1 ≥ 0; x2 ≥ 0
c. Explain very well the assumptions made when solving linear programming problems.
d. What factors should be put in consideration when constructing a linear programming
problem?
e. Explain very well business areas in which linear programming has found enormous use
160
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Chapter-16
Correlation & regression analysis
A INTRODUCTION
CORRELATION ANALYSIS. When two variables are related in such a way that the change in
one variable is accompanied by a change in another then the two variables are said to be
correlated. Thus; Correlation Analysis is a statistical measure of the strength and
direction of the relationship between two or more variables. The strength of the relationship
is given by the correlation coefficient that ranges between [(– 1 and +1 that’s to say (- 1 ≤
rxy ≤ +1)].
B SCATTER DIAGRAM.
This is another visual display of data. It shows the association between two variables acting
continuously on the same item. The scatter diagram illustrates the strength of the correlation
between the variables through the slope of a line.
A. Scatter diagrams make the relationship between two continuous variables stand out
visually on the page in a way that the raw data cannot.
B. Scatter diagrams may be used in examining a cause-and-effect relationship between
variable data (continuous measurement data).
C. They can also show relationships between two effects to see if they might stem from
a common cause or serve as surrogates for each other. They can also be used to
examine the relationship between two causes
Student A B C D E F G
N0 of Sessions 3 5 6 7 3 2 1
QT Marks 44% 68% 81% 88% 67% 44% 29%
Scatter diagram for QT marks against number of revision sessions would appear as below.
100
90
80
70
QT Marks in %
60
50
40
30
20
10
0
0 1 2 3 4 5 6 7 8
Number of QT Revision Sessions attended
161
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Note: Since most of the scatter points are lying on the estimated regression line, we can use
this scatter diagram to estimate say; QT marks when a student attends 20 revision sessions
Scatter diagrams show relationships, but do not prove that one variable causes the other!
rxy
x x y y 3
x x 2 y y 2
The strength and the direction of the relationship between two variables on the basis of
correlation coefficient are summarized in the table below.
Example1. Using any of the formulae above, compute the Pearson or product-moment
correlation coefficient to establish the strength & direction of the relationship between
advertising expenditure and company sales: Interpret your results. Also, test whether the
computed correlation is significant at 5%:
The Pearson’s product moment correlation coefficient can be calculated as follows. 162
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
-Let X and Y represent the advertising costs and sales made respectively.
-We can use the table.
Month X Y XY X2 Y2
Jan 15 28 420 225 784
Feb 30 37 1110 900 1369
March 31 42 1302 961 1764
April 40 45 1800 1600 2025
May 17 25 425 289 625
June 39 50 1950 1521 2500
July 24 32 768 576 1024
August 25 35 875 625 1225
Total 221 294 8650 6697 11316
The above result shows that there is a very strong positive relationship between
the advertising expenditure and the sales revenues got for the last 8 months.
This is how the scatter diagram appears to display the relationship between advertising
expenses and sales revenues.
6 d 2
rs 1 2
n n 1 WHERE 1 rs 1
Solution
X Y Rank of X Rank of Y d = (rank X – rank Y) d2 = (rank X – rank Y)2
8 56 4.5 4 0.5 0.25
5 44 2.5 2 0.5 0.25
11 79 7 8 -
1.0 1.00
13 72 8 7 1.0 1.00
10 70 6 6 0.0 0.00
5 54 2.5 3 -
0.5 0.25
18 94 10 10 0.0 0.00
15 85 9 9 0.0 0.00
2 33 1 1 0.0 0.00
8 65 4.5 5 -
0.5 0.25
∑d2 = 3
6 d 2
rs 1 2
6 x3
1 0.98
n n 1 10(10 2 1)
Solution (Interpretation of findings is very important)
I REGRESSION ANALYSIS
The correlation coefficient computed earlier tells us the extent to which the two variables x & y
are linearly related BUT it doesn’t tell us how to find a particular straight line that represents
the relationship! The problem of establishing which particular line best fits the points of the
scatter diagram comes under the heading Linear Regression Analysis. Thus, Regression is
the statistical method, which measures the average relationship between two or more 164
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
variables and then help us to estimate or predict the unknown value of one variable from the
known value(s) of related variable(s).E.g., the marketing manager can find the average
relationship between money spent on adverts and sales revenues made and then be able
to estimate the level of sales at a given advertising budget. The actual estimation is done by
using a rule, which specifies the estimate of one variable when given the value of another
variable.
We can find the relationship between the harvests and the amount of fertilizers applied as
follows:- 165
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
-We assume that the harvests depend on the amount of fertilizers applied and therefore let X
be the amount of fertilizers applied and Y represent the output respectively.
-The required relationship can be summarized in the form of the general equation.
Y = a + bx
-The values of the constants are obtained from the method of least squares formulae above
with the help of the table below.
X 10 20 30 20 40
Y 98 102 115 212 200
Question
a) When do we use a spear man correlation coefficient
b) Five candidates who applied for a post of chief cook at Sheraton Kampala Hotel were
interviewed by two panelists and the following are the points awarded by each panelist.
Candidate A B C D E
1st Panelist 5.7 5.8 5.9 5.6 5.6
2 Panelist
nd
5.6 5.7 6.0 5.5 5.8
Rank the marks-compute and interpret the Spearman’s coefficient of rank correlation
Question
a) Correlation is not necessarily causation. Discuss
b) List four (4) differences between correlation and regression
c) The following data gives actual sales of Hima cement factory in each of the 8 regions of
Uganda together with the forecast of sales by two different methods.
Region A B C D E F G H
Actual Sales 15 19 30 12 58 10 23 17 167
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Forecast -1 13 25 23 26 48 15 28 10
Forecast -2 16 19 26 14 65 19 27 22
Required
Calculate the spearman’s rank correlation coefficient between;
1) Actual sales and forecast-1
2) Actual sales and forecast-2
3) Which forecasting method would you recommend next year and why?
d) Extract a simple linear regression line (Y = a + bX) between actual sales (Y) and Sales
forecast-1 (X)
Question
The following data relates to costs derived by a cost accountant (Anita) at Java Café.
Week 1 2 3 4 5 6 7 8 9 10
Output (X) ‘000 20 2 4 23 18 14 10 8 13 8
Total costs (Y) 60 25 26 66 49 48 35 18 40 33
a) Find the least squares regression line for the costs on output
b) What are the fixed costs of the factory?
c) In a given week, Café Javas wants to produce 25,000 units, what would be the total
cost to be incurred?
d) Compute the Pearson correlation coefficient and interpret your results
Question
Uchumi and Tusky’s have gone fwhaaa-Bankrupt! You have been hired as a takeover
consultant by the regional supermarket Giant Nakumatti. As part of your preliminary research
work, you have discovered and recorded the following data relating to turnover, number of
stores and regions within the east African region
UCHUMI
Region A B C D E F G H I
Stores (X) 4 6 7 2 9 11 15 5 8
Turnover (Y)’ Billions 45 61 89 12 48 76 67 76 80
TUSKY’S
Region A B C D E F G H I
Stores (X) 7 8 4 7 8 14 18 14 15
Turnover (Y)’ Billions 65 67 44 80 67 60 80 90 95
a) Find the least squares regression lines for each supermarket store
b) Assume the investors want to open up 50 stores in each region, predict the turnover for
each supermarket and comment on the likely accuracy of the answer
c) For each supermarket, what is the amount of turnover that is not affected by the
number of stores?
d) Predict the number of stores an investor needs to open up for each supermarket if he
wants to realize a total turnover of 200 billion in each case.
e) By using the Pearson’s correlation coefficient, which supermarket would you invest in
and why? 168
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Question
a) List four (4) differences between correlation and regression
b) The Electronics World Chain of stores selling equipment wanted to study the
relationship between store sales volume y (measured in millions of shillings) and the
number of households x, in the store’s area in one month. The data for 15 stores is as
follows:-
Store 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
N0 of 161 99 135 120 164 221 179 214 204 101 231 206 248 107 205
Households
(X)
Sales 15.7 9.3 13.7 12.4 15.4 24.2 20.2 23 20.7 13.5 22.5 19.5 24.3 11.5 19.8
Volume
(Y)
Question
a) Correlation doesn’t necessarily mean causation. Discuss this statement and highlight the
importance of correlation and regression analysis in modern business management.
b) A lecturer wanted to wanted to compare the two methods of teaching (A) and (B). She
selected a random sample of 24 students, which she grouped into 12pairs so that the
students in pair have approximately equal scores in an intelligent test. In each pair, one
student was taught by method A and the other by method B and then examined at the
end of the semester. The marks obtained were recorded as follows.
Pair 1 2 3 4 5 6 7 8 9 10 11 12
A 58 68 48 38 70 48 64 70 50 66 32 54
B 84 80 42 62 56 64 48 50 42 32 52 60
Find the rank correlation coefficient and comment on the result obtained
c) To investigate the relationship between annual food and transport expenditure and
annual household disposable income, a sample of 100 households was taken and the
following sums were calculated from the sample data: y 752 ,
y 2 6,220 ,
x 2,486 , x 2
63,622 , and xy 19,600 . Where y represents annual
expenditure on food and transport (in ‘000 UG. Shillings) and x represents annual
household disposable income (in ‘000 UG. Shillings).
Taking annual expenditure on food and transport as the dependent variable and
household disposable income as the independent variable, calculate and interpret the
equation of the least-squares regression line
169
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Chapter-17
Forecasting: Time series analysis
A INTRODUCTION
A TIME SERIES is a collection of figures or values about some event belonging to different
time periods. OR, a series of recorded observations arranged in a chronological order. These
time periods may be in years, months, days or even hours.
5. Exponential smoothing
This is the best and desirable method. It has in built mechanisms for assigning weights in a
desirable way. It involves assigning small weights to older data and vice versa. The problem is
that the initial weight assigned is done subjectively.
6. Historical analogy
In this method, when past data on a particular item is not available, analysis is made on items
similar to that under forecast and then such analyses are used to establish the life cycle and
expected patterns of the item under consideration/ forecast. However, a lot of care is needed
in using analogies which relate to similar items in different time periods.
The seasonal variations may be attributed to two major causes: -Natural factors. This is
climatic changes in different seasons of the year bring about fluctuations in some events.eg
sale of soft drinks will be higher in hot seasons and lower in cool seasons. -Man made
conventions. This is where habits, fashions, customs and other conventions of people in
society cause variations in the time series data during a year e.g. Sales of Christmas cards
are high in December etc.
c) Long term cyclic variations: Such variations in business data can be thought of as due
to underlying economic causes outside the scope of the immediate business environment.
Examples include standard trade cycles or minor recessions. Cyclic patterns usually reflect
periods of boom, depression or recovery. Because of the multiplicity of factors that bring
about these cycles, it is very difficult to predict such characteristics/ patterns. For example,
a depression may occur because of unprecedented increase in petrol prices or stock price
crush. This characteristic also does have a fixed period pattern
172
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Year (s) Year1 Year2 Year3
Quarter 1 2 3 4 1 2 3 4 1 2 3 4
Passengers
(Millions) 220 140 108 210 122 242 246 160 260 ? ? ?
Solution
The trend line equation is given by: t = a + bx. Where,
n xy x y
b
n x 2 x
2
AND a Y bX
We put the various variables in a table
X Y (X.Y) X2
Year1 1 220 220 1
2 140 280 4
3 108 324 9
4 210 840 16
Year2 5 122 610 25
6 242 452 36
7 246 1722 49
8 160 1280 64
Year3 9 260 2340 81
Total 45 1708 8068 285
On substitution,
n xy x y
b 7.869 AND a Y b X 229.1
n x 2 x
2
Therefore, the trend equation is given by: t = 229.1 + 7.869x. We can then forecast the
passenger numbers in year3 second quarter as follows: t = 229.1 + 7.869 (10) = 150.41
Approximately 150 passengers would go through the airport in the second quarter of year3
Advantages: The best technique so far for estimating a trend. Simple to use and it is
applicable to changing conditions Disadvantages. No trend values are obtained for the
beginning and end points of a series. The method is difficult to use for predicting or
forecasting future values which is the main objective of time series analysis.
173
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
4-Moving 4-Moving Centered Moving
Year Qtr Data Total Average Average (Trend)
Year1 1 2.2
2 5
3 7.9 18.3 4.575
4 3.2 4.75 4.66
Year2 1 2.9 19
2 5.2
EXPONENTIAL SMOOTHING
For a moving average (MA), each value in the average was given an equal weight. In a three
point moving average, each value is given a weight of 1/3. Exponential smoothing is a way of
constructing an ‘average’ which gives more weight to recent values of the variable. The
smoothed series is given by the equation:
New smoothed value = (1 – α) [Previous smoothed value] + α [most recent actual value]
Ie. St = (1 – α) St-1 + αxt.
Where α is between 0 and 1. The value of α (alpha) is chosen by the forecaster. The larger its
value, the heavier the weighting being given to the recent values. It’s value may be selected
after testing out several values and measuring which is the best. In practice α is usually in the
174
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
range 0.1 to 0.4le
Thus, to make reasonably accurate forecasts, it is important for us to make adjustments to the
business data so that provisions for the four components are taken care of.
The decomposition of time series data implies ensuring that the time series forecasts
made are accurately computed by making adjustments for trend, cyclical, and random
variations.
175
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
F METHODS USED IN DECOMPOSING TIME SERIES DATA
There are two methods used in decomposing time series data; additive and multiplicative
models.
A. ADDITIVE MODEL
The model assumes that the effects of the various factors/ components of the time series are
additive in nature. Sales = ( T+S+C+R)
B. MULTIPLICATIVE MODEL
The model assumes that the effects of the various factors/ components of the time series are
multiplicative in nature. Sales = (T x S x C x R). The factors are interdependent on each other
and inter related
When business data is compiled, it is imperative to separate the seasonal variations from such
business data (De- Seasonalize). A seasonal index will show us whether there is a fall/ rise in
data and it is expressed in percentages.
The seasonal variations are measured using the additive model where as the
seasonal indices are measured using the multiplicative model.
To accurately measure the variations and seasonal indices, the time series data is split into
quarters and hence we obtain quarterly averages for each quarter.
Example:
The table below indicates profits made by Sejusa Corporation.
Year Quarter-1 Quarter-2 Quarter-3 Quarter-4
2011 97 106 96 100
2012 101 104 80 118
2013 116 111 93 114
2014 120 101 101 106
Solution
Step2: Obtain the average of each quarter
Step3: The seasonal variations for each quarter is the deviation of each quarterly average from
the average of averages
II) Seasonal indices are obtained as percentage variations of quarterly averages above using
the multiplicative model below
Example:
The table below indicates profits made by Sejusa Corporation.
X Y (X.Y) X2
2011 1 20
2 32
3 62
4 29
2012 5 21
6 42
7 75
8 31
2013 9 23
10 39
11 77
12 48
2014 13 27
14 39
15 92
16 53
Total 136 710 6,661 1,496
Y Trend Seasonal
X estimates variations
2011 1 20 30.58 65
2 32 32.42 99
3 62 34.26 181
4 29 36.10 80
2012 5 21 37.94 55
6 42 39.78 106
7 75 41.62 180
8 31 43.46 71
2013 9 23 45.30 51
10 39 47.14 83
11 77 48.98 157
12 48 50.82 94
2014 13 27 52.66 51
14 39 54.50 72
15 92 56.34 163
16 53 58.18 91
5. Sales forecast
The above average seasonal averages are used to forecast the yearly quarterly sales as shown
in the table below using the formulae:
Example 2
The sales of a leasing company have a trend line that is represented by the equation
(Y = 55,000 + 1250t), where t =1 in the first quarter of 2006. The four quarterly seasonal
indices of sales are 107, 100, 82 and 111
Required:
1. Approximate how sales have been increasing in each quarter up to the year ended
2008 180
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
2. Forecast the quarterly sales for the year ended 2009
Solution
Part-1
Trend Seasonal Seasonally
T estimates Indices adjusted
quarters % forecasts
2006 1 56,250 107 60,188
2 57,500 100 57,500
3 58,750 82 48,175
4 60,000 111 66,000
Part-2
Trend Seasonal Seasonally adjusted forecasts
T estimates indices
quarters %
2009 13 71,250 107 76,238
14 72,500 100 72,500
15 73,750 82 60,475
16 75,000 111 83,250
Trend estimates are obtained by substituting quarterly values in the trend equation
Required
I.
II.
Using the additive model, compute the seasonal variations for each of the 4 quarters
Obtain the seasonally adjusted outward passenger movements.
181
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
III. Forecast/ project passenger movements for all quarters of 2011
IV. Plot both the trend values and seasonally adjusted values on the same graph.
Question2
Question3
182
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Chapter-18
Net work scheduling & analysis
18.1 INTRODUCTION
NETWORK SCHEDULING is a task of breaking down a project into its constituent activities
and deciding on their time relationships. The technique is a method of minimizing the trouble
spots such as production, delays and interruptions by determining critical factors and
coordinating various parts of the overall job.
There are two basic planning & control techniques that do utilize a network to complete a
predetermined project or schedule.
The planning phase. This involves setting the objectives of the project and the assumptions
to be made. It also involves the listing of tasks or jobs that must be performed to effectively
complete a project under consideration. In this phase, men, materials, machines, estimated
duration and costs required for the project are also determined.
The scheduling phase. This consists of laying the activities according to the precedence order
and determining:-
I. The start and finish times for each activity
II. The critical path on which the activities require special attention and
III. The slack and float for the non critical paths
The controlling phase. This comes after planning and scheduling phases.
It involves;
I. Making periodical progress reports
II. Reviewing the progress
III. Analyzing the status of the project
IV. Management decisions regarding updating, crashing and resource allocation
183
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
B. An Activity. Some action, which is time consuming effort necessary to complete a
particular part of a project. As such, each activity has a beginning and ending point. It
is represented by an arrow. The illustration below is an activity on arc.
It should be noted that the above activity A has been placed on an arrow/ arc. Thus, it
is termed as activity on arc. But, it is also possible to draw activities G and H on nodes
as illustrated below.
G H
G
5 6
Tail event Head event
D. Dummy activity. Certain activities which neither consume time nor resources but are
used simply to represent a connection or link between the events. Shown with a dotted
line.
-
1 2
Example2.
A manager at Spear Motors has compiled the following Jobs whose precedence relationships
are given below.
Activities/ Jobs 1-2 1-3 2-3 2-5 3-4 3-6 4-5 4-6 5- 6-
6 7
Duration 15 15 3 5 8 12 1 14 3 14
I. Draw the activity network to show the relationship between activities and events 185
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
II. List all the possible routes/ paths of the network and their associated durations
Solution
187
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Illustration of the network
B. Dangling
This implies disconnecting an activity before the completion of all the activities in a network
diagram.
C. Redundancy
This occurs if a dummy activity is the only activity emanating from an event and which can
easily be eliminated.
PRACTICE EXERCISES
Question1
Construct a network for the project whose activities and their precedence relationships are
given below.
188
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Activities A B C D E F G H I
Immediate - A A - D B,C,E F D G,H
predecessor
Question2
Construct a network for the project whose activities and their precedence relationships are
given below.
Activities A B C D E F G H I J K
Immediate - - - A B B C D E H,I F,G
Qu
predecessor
est
ion3
Construct a network for the project whose activities and their precedence relationships are
given below.
Activities A B C D E F G H I J K
Immediate - - A A I,J,K B,D B,D F A G,H F
predecessor
Question4
Construct a network for the project whose activities and their precedence relationships are
given below.
Activities A B C D E F G H I
Immediate - - A,B B B A,B F,D F,D C,G
predecessor
Example6.
Given the following network diagram,
I. Compute the activity durations for each route/ path
II. What is the critical path and total project duration?
189
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
A list of all the possible routes/ paths
Routes / paths Duration (Days)
Route 1 ABE 6
Route 2 ADF 12 (This is the critical path)
Route 3 ACZF 7
It should be noted that a dummy activity (Z) consumes no time in terms of activity duration.
Thus a 0 is appended to any dummy activity.
18.7.2 THE EARLIEST START TIMES (EST) AND LATEST START TIMES (LST) OF AN
ACTIVITY
A critical path can also be obtained by calculating Earliest Start Times (EST) and Latest Start
Times (LST) of an activity.
Assume the following network has been drawn and the activity times estimated in days.
190
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
The EST's can be inserted as follows:-
Note that for simple networks the critical path can be found by inspection, ie looking for the
longest route, but the above procedure is necessary for larger projects and must be
understood. The procedure is similar to that used by most computer programs dealing with
network analysis.
We use the FORWARD PASS to identify earliest times (EST and EFT)
EST Rule: All immediate predecessors must be done before an activity can begin
– If only 1 immediate predecessor, then: EST = EFT of predecessor
– If >1 immediate predecessors, then EST = Max {all predecessor EFT’s}
192
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Node Notation
Example:-
General Foundry Inc has 16 weeks to install a complex air filter system on its smokestack. It
may be forced to close if not completed within 16 weeks due to environmental regulations.
They have identified 8 activities as listed below.
SOLUTION
Part-1: ACTIVITY ON NODE DIAGRAM
195
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
TOTAL SLACK TIME VS. FREE SLACK TIME
• Total slack time is shared by more than 1 activity
Example: A 1 week delay in activity B will leave 0 slack for activity D
• Free slack time is associated with only 1 activity
Example: Activity F has 6 week of free slack time
There are three types of float, Total Float, Free Float and Independent Float. To
illustrate these types of float, part of a network will be used together with a bar diagram of the
timings, thus:
196
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
(a) Total Float. This is the amount of time a path of activities could be delayed without
affecting the overall project duration. (For simplicity the path in this example consists
of one activity only, ie Activity K).
Total Float = Latest Head time - Earliest Tail time - Activity Duration
Total Float = 50 - 10 - 10
= 30 days
(b) Free Float. This is the amount of time an activity can be delayed without affecting the
commencement of a subsequent activity at its earliest start time, but may affect float of
a previous activity.
Free Float = Earliest Head Time - Earliest Tail time - Activity Duration
Free Float = 40 - 10 - 10
= 20 days
(c) Independent float. This is the amount of time an activity can be delayed when all
preceding activities are completed as late as possible and all succeeding activities
completed as early as possible. Independent float therefore does not affect float of
either preceding or subsequent activities.
Independent float = Earliest Head time - Latest Tail time - Activity Duration
Independent float = 40 - 20 – 10 = 10 days
PRACTICE EXERCISES
a) A project consists of a series of tasks labeled A, B ………..H, I with the following
constraints. A D, E ; B, D F ; C G; B H F , G I . Note that W X , Y implies that
X and Y can not start until W is completed.
197
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Activity A B C D E F G H I
Preceding activity - - - A A B, D C B F, G
Normal time (Days) 23 8 20 16 24 18 19 4 10
I. Construct the network using the information above
II. Compute all the Earliest Event Time and Latest Event Time for each event
III. Compute the Total and Free floats for each activity and hence the critical path
Solution
Total Float = Latest Head time - Earliest Tail time - Activity Duration
198
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Task TOTAL FLOAT
A 1-2 0
B 1-3 8
C 1-4 18
D 2-3 0
E 2-6 20
F 3-5 0
G 3-6 24
H 4-5 18
I 5-6 0
It should be mentioned that some ways of reducing project duration are cheaper than others
and as such the analysis of costs in networks is of paramount importance. It helps in
identifying the cheapest way of reducing a project overall duration.
When a project is completed early, it often attracts a bonus. Again, when a project is
completed late, it often attracts a penalty. Therefore, any one handling extensive projects
needs to analyze costs to find out whether it is worthwhile to be late and pay extra costs or
rush a project and get a bonus.
The network diagram is used to identify activities whose duration should be shortened so that
the completion time of a project can be shortened.
A. Normal cost
Normal cost is equal to the absolute minimum of direct cost required to perform an activity. It
is the cost associated with a normal time estimate for a particular activity. Has an associated
normal time.
199
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
B. Normal time
This is a time estimate for a particular activity when resources (that is to say; men, machines
etc) are used in the most efficient manner possible.
C. Crash cost
It is the minimum direct cost required to achieve the crash performance time. Such crash costs
are usually higher than the normal cost because of extra wages, overtime premiums and extra
facility among other factors
D. Crash time
This represents the minimum activity duration time that is possible and any attempts to further
crash would only raise the activity cost without reducing the time.
E. Cost slope
This is the average cost of shortening an activity by one time unit (day, week, month etc). The
cost slope is generally assumed to be linear and calculated from;
Step4. Update all activity times. If the desired due date has been reached, stop. If not, return
to Step 2
Note – When there are multiple options all with the same crash cost, look at
whichever is least complicated and would affect the fewest number of people
200
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Example7.
The table below provides cost and estimates for a seven activity project
NORMAL CRASH
Activity Preceding Duration Cost Duration Cost
activity (weeks) $ (weeks) $
A 1-2 - 2 10 1 15
B 1-3 - 8 15 5 21
C 2-4 A 4 20 3 24
D 3-4 B 1 7 1 7
E 3-5 B 2 8 1 15
F 4-6 C,D 5 10 3 16
G 5-6 E 6 12 2 36
Find;
I. Draw the network from the above table and identify the critical path
II. Find the total normal cost and duration of the entire project
III. Crash the project activities so that the overall project completion time is reduced to 9
weeks
Solution
Part-I-
Part-II-
The total normal cost is obtained by summing all the normal costs of the project. And as such,
the total normal cost of the project is $82,000
The duration of the critical path is the total normal duration of the entire project. It is 16
Weeks
Part-III-
Note. Only activities on the critical path can be crashed. Thus their total duration is called the
crash duration.
Step1. Compute the crash cost per time period OR cost slope and also indicate the available
time to crash each activity (difference between the crash time and normal time)
201
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Activity OR Crash Available Narration
Cost Per Crashing
Week $’000 Time
= (CC - NC) / (NT - CT) (NT - CT)
A 1-2 5 1
B 1-3 2 3 To be crashed first
C 2-4 4 1
D 3-4 0 0
E 3-5 7 1
F 4-6 3 2 To be crashed third
by 2 weeks
G 5-6 6 4 To be crashed
second by only 2
weeks
Step2. Using current activity durations, draw the network, find the critical path and identify
the critical activities
The network has been drawn on the previous page and the critical path is B, E and G OR (1 –
3 – 5 - 6). These are the activities to be crashed
Note: we are supposed to reduce the total project duration from 16 to 9 weeks
1st Crashing: Since activity (1- 3; B) has the least crash cost per week, we crash it by all the
3 weeks at a rate of $2 (‘000) per week. Thus, total new project duration and associated new
total costs are given below.
The table below is necessary in identifying the possible routes and their associated new
durations
1-3-5-6 BEG 16 13 11 9
1-3-4-6 BDF 14 11 11 9
1-2-4-6 ACF 11 11 11 9
2nd Crashing:
After the first crash, we still maintain a critical path of B, E and G (1 – 3 – 5 – 6) (duration is
13 weeks which is the longest). We are done with activity B (1-3). As such, we only have E, G 202
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
(3-5-6) on the critical path. Therefore, we crash activity G (5-6) since it has the lowest crash
cost per week. We reduce G by only 2 weeks to create a total duration of 11 weeks which is
parallel to other routes. If we reduce it by more than 2 weeks, we would have created other
two new critical paths before completing the G crash.
3rd Crashing:
Note: Duration must be reduced to 9weeks. We have only managed to reduce it to 11 weeks.
At the third crashing attempt, we can see that all possible routes are now critical (all have the
same maximum durations of 11weeks)
Step3 clearly highlights that; if there is more than one critical path;
• Select one activity from each critical path such that;
– Each selected activity can still be crashed
– The total crash cost of all selected activities is the smallest and a note that the
same activity may be common to more than one critical path.
Possible routes OR
1-3-5-6 BEG G (5-6) has the lowest cost slope. We crash it by
2weeks at $6 per week creating an extra cost of
($6 x2) = $12
1-3-4-6 BDF F (4-6) is common to both routes and has the
1-2-4-6 ACF lowest cost slope, we crash it by 2 weeks at $3
per week creating an extra cost of ($3 x2) = $6
Thus total new costs =$12 + $6 = $18
Conclusion
The project duration can be reduced up to 9 weeks only but the total project costs has shot up
to a total of $118,000.
We can use the new project durations and the new project costs to plot a time cost trade
off curve as illustrated below.
203
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Example8. Incorporating over head costs
NORMAL CRASH
Activity Duration Cost $ Duration Cost $
(weeks) (‘000) (weeks) (‘000)
1-2 8 100 6 200
1-3 4 150 2 350
2-4 2 50 1 90
2-5 10 100 5 400
3-4 5 100 1 200
4-5 3 80 1 100
Given that indirect costs for the following project are $70 (‘000), crash the project and
determine the optimal duration and associated costs for completing this project.
Solution
Step1. Compute the crash cost per time period OR cost slope and also indicate the available
time to crash each activity (difference between the crash time and normal time)
Step2. Using current activity durations, draw the network, find the critical path and identify
the critical activities 204
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
The critical path is (1 – 2 – 5). These are the only activities to be crashed
Original project duration = 18 Weeks
Original project direct costs = $580
Original project indirect costs = $1,260 (18 x $70)
Total original project costs = $1,840 ($580 + $1,260)
1st Crashing: Since activity (1- 2) on the critical path has the least crash cost per week,
we crash it by all the 2 weeks at a rate of $50 (‘000) per week. Thus, total new project
duration and associated new total costs are given below.
The table below is necessary in identifying the possible routes and their associated new
durations
Possible Before 1st 2nd 3rd
routes Crashing Crashing Crashing Crashing
1-2-5 18 16 12
1–2–4- 13 11 11
5
1–3–4- 12 12 12
5
2nd Crashing:
After the first crash, we still maintain a critical path of (1 – 2 – 5) (duration is 16 weeks which
is the longest). We are done with activity (1-2). As such, we only have (2-5) on the critical
path. Therefore, we crash activity (2-5) since it has the lowest crash cost per week on this
205
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
critical path. We reduce (2-5) by only 4 weeks to create a total duration of 12 weeks which is
parallel to other routes (like1-3-4-5).
3rd Crashing:
Note: Duration must be reduced as much as possible. We have only managed to reduce it to
12 weeks.
At the third crashing attempt, we can see that two new paths are now critical (1-2-5) and (1-3-
4-5).
Step3 clearly highlights that; if there is more than one critical path;
• Select one activity from each critical path such that;
– Each selected activity can still be crashed
– The total crash cost of all selected activities is the smallest and a note that the
same activity may be common to more than one critical path.
Possible routes
1-2-5 (2-5) has the lowest cost slope and still
has balance duration of only 1week. We
crash it by 1weeks at $60 per week
creating an extra cost of ($60 x1) = $60
1–3–4-5 (4-5) has the lowest cost slope of $10.
Thus, we crash it by 1weeks at $10 per
week creating an extra cost of ($10 x1) =
$10
Total extra costs $70 (60 + 10)
resulting from
simultaneous crash
Conclusion
The project duration can be reduced up to 11 weeks only but the total project costs has shot
up to a total of $1,760.
Example8. Dealing with bonuses for early and penalties for late completion of
projects
For national and regional security, the Government of Uganda has signed a contract with a
Chinese arms manufacturer to produce weapons that will supplement the current space based
defense system. The contract stipulates that the weaponry be ready within a year (50 working
weeks). There is a penalty of $100,000 per week for late delivery; there is also a bonus of
$50,000 per week for early delivery.
206
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Several tasks are necessary to complete the project. First, the implications of the contract have
to be detailed by personnel, materials and systems departments. A joint team will carry out
this task, which is expected to take two weeks. Once this is complete the next three tasks can
be undertaken concurrently: ordering raw materials (three weeks), allocating and hiring
personnel (14 weeks), planning the modification and adaptation of equipment (12 weeks).
After the raw materials have been ordered it will take a further 20 weeks for them to be
delivered. Once the modifications have been planned it will take a further 18 weeks for them
to be carried out. When materials, equipment and personnel are available it will take two
weeks to train the quality control personnel and three weeks to train the production personnel.
These activities can take place simultaneously. When completed the production phase can start
and this will last for 18 weeks, followed by a further five weeks of testing. The final
preparations stage then takes three weeks. A few of the activities can be crashed, as shown in
the following table:
Required
1. How long will the project take to complete, without crashing any activities? What are
the critical activities?
2. The time for the project can be reduced by crashing activities but this will affect bonus
or penalty payments specified in the contract. What schedule of activities is in the best
financial interests of the company?
Solution
207
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Task How long? What other task does it
depend on?
Choose which bag you want to take 5 minutes none
Find the bag 5 minutes Depends on choosing which
bag I want to take
Find the books you need for the day 5 minutes none
Get someone to make you a packed 10 minutes none
lunch
Get a letter from a parent/carer giving 20 minutes none
you permission to take part in a school
trip
Put the books and lunch in your bag 2 minutes Depends on finding the
books, and getting someone
to make the lunch
Close up the bag 1 minute Depends on putting the books
and lunch in the bag
Put the permission letter in your pocket 1 minute Depends on getting the letter
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
Tasks
Choose bag
Find bag
Find books
Get letter
Close bag
letter in pocket
Some of the early DOSbased estimating and planning packages were particularly bad for this.
The only way to learn to use them was to be shown how by an expert. The package itself and
the supporting documentation did not contain the amount of information required for a new
user to learn how to use the system. Modern packages are more approachable, but large
project-planning packages can still appear to be daunting to the uninitiated. Ideally, even the
most complex packages should be readily approachable and should offer support and
assistance to new users. Staff will respond much more readily to such software than they
would to unapproachable systems.
• The systems should use familiar displays. The most successful complex packages produce
displays and outputs that are compatible with what users expect. Most people expect a cost
report to have a certain type of appearance. It should typically show the budget limit for a
particular work package, how much has been spent to date and what level of progress has
been made in incurring these costs. The most effective displays and reports are those that
present this standard information as clearly and as succinctly as possible.
• The system should be CMS compatible. Large and complex projects often use a configuration
management system where all information is centralized and distributed to all members of the
project team through remove PCs. The amount of information that each person can obtain and
the level of access that is permissible is restricted by individual authority clearance levels. The
higher the clearance the greater the amount of information the person has access to within the
DMS. A centralized version of the project master schedule (PMS) is particularly important since
most members of the project team have a considerable interest in it.
• The system should be extendable. Complex software increasingly incorporates a degree of
over design. A common complaint about Microsoft Word is that it incorporates a great many
functions and applications that nobody ever uses. This overdesign results in the use of large
amounts of ROM so that the microchip makers have to keep developing new and more
powerful processors. The end result is that PCs become obsolete very quickly as only the latest
chips can run the latest version of the software. This end result is not intentional on the part of
Microsoft. They are simply designing systems that have built in capacity for extension. Many of
the functions that are not used are in fact designed to lead in to the next version of the
software. The potential for extension is an important aspect of software design. The software
should not be limited to what people want now. It should also be forward looking and try to
incorporate next generation ideas so that innovation supply stays ahead of demand.
PRACTICE EXERCISES
Question1.
The normal cost/duration and other relevant information for a project is given below.
There is a bonus of $100 per day for every day that is saved below the contract period of
15days and a penalty of $200 for each day after the 15days.
Required
a) Calculate the normal duration and the normal cost of the project
b) Calculate the minimum cost of completing the project in 15days
c) State the optimum plan for the company to attempt
d) Revert to the normal programme and normal costs and state what action you would
recommend to ensure completion by the original date, if after the tenth day, the actual
situation was as follows;
I. Activities completed at normal cost; 1-2; 2-3; 3-6; 2-4; 2-5
II. Activities not yet started 4-6; 5-6; 6-7
What is the revised cost of the project in these circumstances?
Question2
Delco is a Swedish small engineering company that has been contracted by BATA shoe
company to design and produce a machine that will help increase mass production of BATA
shoes. The activities required in the design and manufacture of the machine are listed below
together with duration and costs.
In addition to the above cost figures, overheads of $250 per week will be incurred for the
duration of the project.
Required
I. What is the critical path and duration of the project?
II. What would be the effect of a strike at the factory supplying the internal machinery
thereby delaying its delivery for 4 weeks?
III. What would be the effect if the test on the machinery had been done incorrectly and
212
had to be redone taking another 4 weeks?
Required
a. Draw a network for the scheme of activities set out above. indicate full information on
earliest and latest event times and indicate the critical path
b. Draw a bar chart for the scheme and indicate the total float of each activity
c. If each activity requires one member of office staff of the association, so that the
activities may be completed in the estimated times, what is the minimum number of
staff that should be allocated to the scheme?
d. What would be the effect on the total time if one of the allocated staff was taken ill for
the duration of the scheme and not replaced?
Example4
The following time- cost table (time in weeks and cost in shillings) applies to a project. Use it
to arrive at the network associated with completing the project in minimum time with minimum
cost.
Activity Normal Crash
Time Cost Time Cost
1-2 2 800 1 1400
1-3 5 1000 2 2000
1-4 5 1000 3 1800
2-4 1 500 1 500
2-5 5 1500 3 2100
3-4 4 2000 3 3000
3-5 6 1200 4 1600
4-5 5 900 3 1600
213
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Example5
Activity Required Normal Time Normal Cost Crash Time Crash Cost
Predecessor
A (None) 8 $4000 6 $6000
B (None) 5 1500 4 2000
C (None) 6 2500 4 3000
D A 4 1800 3 2000
E A, B 6 1000 5 1200
F C 7 2000 5 3000
G A 5 3000 3 6000
H D, E, F 8 4500 5 9000
I C 9 6000 4 10000
J D, E, F 6 6000 4 8000
K G, H 4 2000 3 2600
L D, E, F 6 3000 3 9000
M I, J 4 8000 2 12000
Required
The scheduled completion time is 25 weeks. You must pay a $1,500 penalty for every week
you are late. You get a $1,000 bonus for every week you are early.
1. Show your network diagram.
2. Find the normal completion time and the critical path.
3. Determine the schedule that minimizes your total cost for this project, including any penalty
or bonus.
NOW THAT YOU ARE HAPPY, TRY OUT THESE PAST PAPER EXAM NUMBERS
214
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Chapter-19
Index numbers
19.1 INTRODUCTION
Bank of Uganda, consumer groups, families, Trade unions, several company directors, business
leaders and other economists are very much interested about changes in prices. Such changes
in prices are usually explained by a consumer price index (CPI), a producer price index etc.
The cost of living index is the most common type of index. It comprises the cost of so many
types of goods: like foods, fuel, transport, etc. instead of using the miscellaneous and often
confusing prices of all these purchases, we use an index number which summarizes them for
us. If the index number for 1996 is 182.1 compared with 165.3 in 1995, we can calculate that
the cost of living has risen by:-
182 .1 165 .3
165 .3 x100 10.2% This is rather easier than having to cope and deal with the range
of individual price increases involved.
Or
An index number shows the average percentage change of a variable from one point of time to
another. Here, they reduce a group of numbers to a mere usable and easily understood
figures which can easily be compared. Even where degree of confidentiality is required, an
index number may be used, e.g. a company may not be willing to disclose to the public its
level of profits so it opts for index numbers.
Example; In 2001, the average hourly wage was Ug. Shs. 20,000 and in 2006 the average
hourly wage is Ug.Shs. 35,000. What is the index for hourly earnings in Uganda for 2006
basing on 2001?
Solution:
Hourly earnings in 2006 x 100% 35,000 shs x100 175
20,000 shs
Hourly earnings in 2001
Interpretation:
In Uganda, the hourly earnings did increase by 75% (175-100) between 2001 and 2006.
Example; The Uganda Bureau of Statistics reported that annual Banana production dropped
from 30,529,000 in 2002 to 5,100,000 in 2006. What’s the index for banana production during
the period?
The right hand column shows Jan set to the standard figure of 100 and all the profit values set
to percentages of this standard. The figures in the right hand column are called index numbers
of profits and in this case, Jan is called the base month. Against which all others are compared.
So the index number illustrates changes by expressing the items in time series as percentages
of item at a chosen base period.
216
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
a) SIMPLE PRICE INDEX OR RELATIVE
This is a ratio of the (value) of one variable of a single commodity (price x quantity) in a given
year to its value in another year commonly referred to as base year.
A simple index number can be;
I. Price relatives
II. Quantity relatives
III. Value relatives
Case1. Price relative: This is the ratio of the price of a commodity in one period to the price in
another period.
pn
Price relative = …expressed as a ratio
p0
pn
Price relative = x100 …expressed as a percentage
p0
Case2. Quantity/ volume relatives: Index number expressed as a ratio of the quantity of a
single commodity in a given period to its quantity in another period
qn
Price relative = …expressed as a ratio
q0
qn
Price relative = x100 …expressed as a percentage
q0
Case3. Value relatives: Index numbers expressed as the ratio of the value of a given single
commodity (That’s to say; price x quantity) in a given period to the value in another period
called the base year
Pn Qn
I x100
P0Q0
Qn
I x100
Q0
Pn Qn
I x100
P0Q0
Compute
a) Simple aggregate price index (121%)
b) Simple aggregate volume/ quantity index number (131%)
c) Simple aggregate value index numbers (123%)
218
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Qn
I Q0
x100
n
Pn Qn 1
I x
P0Q0 n
n is the number of commodities
The advantage here is that prices of all the commodities don’t have to be in the same units,
although the prices of each individual item must be in the same units. This is a useful measure
when we are dealing with results from different countries.
The major disadvantage of this index is that the relative importance of the items included is
not considered, i.e. items are not weighted in accordance with their relative importance.
Compute
a) Price average relative (1.05)
b) Volume average relative (1.58)
c) Value average relative (0.37)
Note: A simple aggregate index is easy to construct and easy to understand but it does not
meet a criterion of adequacy called the UNITS TEST. The test requires that index yield the
same results regardless of the units for which the prices of items are quoted. We can avoid
this problem by using the RATIO OF PRICES of a given item rather than the actual prices
them- selves. See previous example
or
Example: Find index numbers for the following data taking 1980 as base year.
Years
Price
Solution
Link Relatives
Year Price Chain Indices
220
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Example: Illustration of computing index numbers
a) A weighted index
Example: The table below represents the weights and indexes of five items
Index
WI
W
Solution
Item Weight (W) Index (I) WI
Food 304 124 …
Tobacco & alcohol 129 126 …
Housing 331 127 …
Transport 120 119 ….
Goods & services 116 128 …
Total 125,115
IndexNo
WI ...
W ...
Example: Given the following information: construct a cost of living index
Solution
CostofLivi ng
GroupIndexXweight ... 125.96
W ...
222
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
b) WEIGHTED AVERAGE PRICE INDEX
This is computed using weights of any typical year and not necessarily the base year or current
year.
Pn wi
Weighted average price index = I x100
P0 wi
Where; wi is the weights of the ith commodity
Compute
The weighted average price index (148.2%)
Steps involved.
Step 1: Assign a weight to the price of each item by multiplying the price by the weight
selected.
Step 2: Add these products to get the weighted aggregates.
Step 3: Divide each of these totals by the weighted aggregates of the year selected as base.
Use the formula:
PnQ0
I x100 Where Qo is the quantity consumed in the base period and other
P0Q0
variables as defined before.
Note: This kind of index reflects changes in prices alone – the same quantities of goods are
priced at two different terms.
This is called the Paasche’s price Index. OR current weighted aggregative index.
223
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
PnQn
I x100
P0Qn
Compute
a) The Laspeyres price index (81.01%)
b) Paashe’s price index (79.77%)
It can be noted that the Laspeyres method results in a more logical answer as compared to the
simple index. But it should be noted that this method uses the amounts consumed in the base
period as weights. Therefore, it assumes that the eating habits of a typical consumer didn’t
change from 2011 to 2012. Thus only price fluctuated casing a decrease in the index from 100
in the base period to 81.01% in the 2012. Another major Disadvantage of the Laspeyres index
is that the weights become out of date as the pattern of demand changes. This index tends to
overstate the changes in prices as it takes no account of the fall in consumption when prices
rise.
Interpretation:
A market basket of food selected in 2012 costs 20.23% less than the same basket of foods at
2011 prices.
This method has an advantage of using current consumption figures as weights; that is the
consumption pattern is always up to date. However, this method poses a practical problem as
it needs to be revised every year using new present year weights. This consumes a lot of time
and involves expensive survey work. It tends to understate the changes in prices as most
people tend to buy less of those commodities which have gone up in prices. Because of the
above short comings, the Laspeyres method is most commonly used when assessing relative
changes in price levels/ computation of index numbers (with some modifications).
Pn (Qn Qn )
Edeworth- marshall’s price index = I x100
P0 (Qn Qn )
Compute
The Edeworth- marshall’s price index
Pn (Qn Qn ) 152900
I x100 83.92%
P0 (Qn Qn ) 182200
Pn Q0 Pn Qn 1
x 81x79.77 x
1
DORBISH- BOWLEY’S INDEX= I
P0Q0 P0Qn 2 2
However, one has to note that in inflationary situations, real income/ wage has to be deflated
in order to make an allowance for the effect of changing price levels. In deflating real income, 225
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
we shall obtain the deflated income and this shall be taken as the new real income for index
number computation purposes.
Pn w
Cost of living index= I x100
P0 w
These index numbers are generally constructed for each week. Where the average of the
weekly index numbers is taken as the index number for the month and the average of the
monthly index numbers gives the cost of living for the whole year.
d) Converting data into indexes makes it easier to assess the trend in a series composed
of exceptionally large numbers.
e) An index number is the most convenient way of expressing a change in a
heterogeneous / different group of items. The consumer price index for example
encompasses over 400 items. Only the conversion of these many diverse goods to one
index number enables the government and other concerned authorities informed about
the overall movement in prices.
f) They are used in deflating nominal values to obtain real ones. E.g. if today’s dollar will
buy 100/130 of 77% of what it could buy in the base year. If the average weekly
earnings were $115 in the base year and are now $125 in current period, then today’s
real earnings per week is 125 x 77% = $96.25, which is $18.75 less than in base year.
226
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
19.7 IMPORTANCE OF COST OF LIVING INDEX NUMBERS
a) They are purposely used in determining remunerations so as to maintain the same
standard of living in other years as that of the base year
b) Their reciprocal is used in measuring the purchasing power of money
c) They are also used to find real wages by the process of deflation
d) They give guidelines to the government on its decisions about the income and general
economic policy, wage policy and price policy among others
c) The choice of time periods that are to be compared. Here, we choose a relatively
recent period of good economic activity, e.g. if we are to compare prices of goods from
1982 to 1992: 1982 acts as our base year and 1992 acts as our given period.
d) The choice of appropriate weights. The choice of the relative importance of each item
is very difficult. Here we weigh the prices of items in such a way as to account for their
relative importance in practice.
e) The choice of a sustainable formula. There are various ways in which relative changes
in prices of quantities can be described. However, choice on which is best is not easy.
d) Uganda National Bureau of Statistics recorded the prices of the following commodities in
rural collection centres as:
I) Using 2001 as the base year, the Laspeyre’s Price index number for 2005 is 199%. Find
the value of x to the nearest kilogram.
II) Calculate Paasche’s Price index number and comment on the answer
III) Compute the Laspeyers Quantity index number and interpret your answer.
Question 2
a) Define an index number and explain the factors that must be considered when one is
constructing index numbers.
b) How does a price relative differ from a value relative?
c) Given the following information,
I) If the Laspeyres price index number was computed as 81.01%, compute the Price of
commodity C in the new year (missing value X)
II) Compute a Simple average of value relatives
III) Compute a Paashe’s quantity index
Question 3
a) Write short notes on the following:
I) Cost of living index number
II) Weighted index numbers
III) Fixed base index numbers
IV) The chain based method
228
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
I) If the ratio between Laspeyre’s and Paasches’ index numbers is 28:27, find the
missing figure on the table.
II) Now, find the Pasche Quantity Index and Interpret your results
III) Compute the value relative Index number
Question 4
a) Distinguish between fixed-base and chain-base index numbers in time series.
b) The table below shows the average monthly wage (Shs ‘000’) for workers in each grade in
2000 and 2002 for a certain factory:
Required:
I) Obtain the values of x, y and z.
II) Obtain an index number for the monthly wage of the whole factory in 2002.
c) A recent survey carried out by a researcher on the prices of various goods and services
and the spending of an average person in Mbarara Town revealed the data indicated in
the table below:
Required:
I) Calculate the cost of living index in Mbarara Town, giving the answer to 1 decimal place.
II) Comment on the answer in (d) (i) above.
Question 5
a) List any four (4) basic problems encountered when constructing Index numbers
b) Explain the distinction between the following terms:
I. A price Index number and Quantity Index Number
II. A weighted and Un-weighted Index number
III. A fixed base and chain based index number
Compute:-
I. The Laspeyres price index
II. The Laspeyres Quantity index
III. Paashe’s quantity index
IV. Fishers ideal price Index
Question 6
a) Explain the importance of “weighting” in index number construction.
b) The following table shows the quantities of materials (steel, cement and timber) used
in the construction industry in 1995 and 2000, together with the proportions of total
value in 1995.
I. Calculate the simple (Unweighted) aggregate quantity index for these materials in
2000, using 1995 as the base year.
II. Calculate an (Unweighted) arithmetic mean of the relatives, quantity index for 2000
using 1995 as the base year.
III. Using the proportions of the total value as weights, calculate the weighted arithmetic
mean of relatives’ quantity index for 2000 with 1995 as the base year.
230
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
Question 7
a) The following table is a summary of consumption patterns of a number of residents
from Bugolobi suburb for two periods.
Note:
Quantity is a ratio of expenditure in a given year to price per unit in the same year. Construct
a suitable index number of prices from the above information concerning consumption patterns
for the above residents from 2003 to 2004.
NOW THAT YOU ARE HAPPY, TRY OUT THESE PAST PAPER EXAM NUMBERS
231
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
SYLLABUS FROM THE CPA GUIDE
A. INTRODUCTION o-give, Lorenz and Z charts)
1. Meaning of statistics (viii) Basic principles of drawing each type of graph;
(a) Functions and limitations of statistics interpretation and
(b) Types of statistics: descriptive and inference statistics limitations of graphs
B. STATISTICAL DATA, PRESENTATION AND MEASURES 2. Measures of location / central tendency
1. Statistical data (a) Measures of averages: arithmetic mean,
(a) Primary and secondary weighted mean, mode, median,
(i) Sample and population harmonic and geometric means
(ii) Discrete and continuous (b) Characteristics, merits and demerits of each type
(iii) Quantitative and qualitative of average
(iv) Variable and attribute (c) Computation of the mean using actual values
(b) Sources of data and /or working mean
(c) Methods of data collection: interview, questionnaire, (d) Determination of the mode and median using
direct observation, census both computational and graphical
and sample survey methods
(d) Merits and demerits of each method of data (e) Interpretation of calculated values of each
collection measure
(e) Response errors (f) Computation of weighted mean, harmonic mean
(f) Sample frame and geometric mean
(g) Sampling techniques: random sampling (simple (g) Relationship between mean, mode and median
random and stratified), quasi Note: Knowledge of logarithms either from
random sampling (systematic and multistage sampling), mathematical tables or calculator
non- random sampling is assumed.
(cluster and quota sampling) 3. Measures of dispersion
(h) Advantages and disadvantages of each technique (a) Meaning of dispersion and variation
of sampling (b) Measures of dispersion: range, quartile deviation
(i) Methods of data presentation: (interquartile range), semi quartile
(i) Tables (row and column, two-way, frequency and deviation (semi-interquartile range), decile and
percentage distribution percentile range, mean
tables) deviation, quartile coefficient of dispersion, standard
(ii) Basic principles of table design deviation and variance
(iii) Interpretation of data presented in each type of (c) Characteristics, merits and demerits of each
table measure of dispersion
(iv) Limitations of each type of table (d) Determination of quartile, decile and percentile
(v) Pictograms, charts (simple and compound bar by computation and graphically
charts, and pie charts) (e) Computation of quartile coefficient of dispersion
(vi) Methods of construction of each type of chart; type (f) Computation of mean deviation, standard
of data suitable for deviation and variance using the
each type of chart; limitations of each type of chart working mean method
(vii) Graphs (line graph, histogram, frequency polygon (g) Properties of standard deviation and coefficient
and curves, including of variation
(h) Interpretation of calculated values of each
measure
4. Measures of skewness
(a) Positive, symmetric and negative skewness
(b) Relationship among mean, mode and median in
a skewed distribution: 3
(mean – median) = (mean – mode)
(c) Computation of measures of skewness using Karl
Pearson’s and Bowley’s
formulae
(d) Graphical illustration of the concept of skewness
and kurtosis
(e) Comparison of skewness with measures of central
tendency and dispersion
(f) Interpretation of skewness and degree of
skewness
232
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
C. PROBABILITY AND DISTRIBUTIONS critical (acceptance and rejection) regions, and level of
1. Probability theory significance
(a) Basic concepts of probability: event, outcome, sample, sample (b) Methods of testing (normal Z score, Student’s t and chi-
space and equiprobable square)
(b) Classical definition of probability (c) Reading values from tables (one tailed and two tailed)
(c) Properties of probability theory: (probability limits, total (d) Computation of the statistic and determination of the level
probability and complimentary probability); types of events of significance at a
(independent, dependent, and particular percentage
mutually exclusive events) (e) Testing of population mean and population proportion
(d) Rules used in probability theory that is addition and (f) Testing goodness of fit, use of contingency tables, and
multiplication computation of expected
(e) Conditional probabilities (prior and posterior, including use of values for chi-square test
probability trees and Bayes’ theorem) as methods of computation 3. Control charts
(f) Probability frequency distribution and the computation of (a) Definition of the concept ‘control chart’
expected values (expectation) (b) Advantages and disadvantages of control charts
Note: Application of the knowledge of set theory in probability is (c) Drawing control charts
assumed. (d) Interpretation and making conclusions from control charts
2. Permutations and combinations E. INDEX NUMBERS
(a) The factorial notation 1. Meaning, uses, importance and limitations of index numbers
(b) Application of permutations and combinations in probability 2. Price and quantity relatives
3. Probability distribution 3. Considerations in the construction of index numbers
(a) The concept of probability distribution 4. Types of index numbers: simple and weighted (price and
(b) Normal distribution – properties and standardization quantity)
(c) Computation of probabilities using normal distribution tables 5. Computation, comparison and interpretation of Laspeyres
(d) Binomial distribution probability properties and use of binomial and Paasche price and
formula quantity indices
(e) Computation of the mean, variance and standard deviation of 6. Time series relatives: fixed base and chain relatives
a binomial distribution 7. Published indices: cost of living, consumer, stock and retail
(f) Use of binomial distribution tables to find probabilities indices
(g) Approximating a binomial distribution to a normal distribution F. REGRESSION AND CORRELATION
(h) Properties and computation of probabilities using poisson 1. Correlation
distribution (a) Definition and importance of correlation analysis
D. ESTIMATION AND HYPOTHESIS TESTING (b) Types of correlation: linear or non-linear, positive, negative,
1. Estimation zero and no
(a) Definition of concepts: estimation, estimator (point estimator correlation
and interval (c) Scatter diagrams
estimator), standard error of the mean, and population and (d) Methods of calculating coefficient of correlation: product
sample mean moment (Karl
(b) Confidence limits and intervals Pearson’s) and rank correlation (Spearman’s formula).
(c) Estimation of the population mean from a large sample using (e) Interpretation of correlation co-efficient
normal distribution (f) Merits and demerits of each method
(d) Estimation of the mean from a small sample using Student’s t (g) Link between correlation and causation
distribution 2. Regression
(e) Use of contingency tables for chi-square distribution (a) Definition of terms: regression, regression equation,
(f) Estimation of the population proportion from a large sample regression coefficients
2. Hypothesis and significance tests (b) Distinction between correlation and regression
(a) Definition of basic concepts used in hypothesis testing: null (Ho) (c) Uses of regression line
and the alternative (Ha) hypotheses, errors in hypothesis testing (d) Definition of dependent and independent variables
(type (I) and type (II) errors), (e) Regression equation: y on x and x on y
(f) Calculation of regression equation using least squares
method
(g) Interpretation of regression line and prediction of other
values using the line
Note: This excludes non-linear regression and non-linear
correlation.
233
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
G. TIME SERIES AND FORECASTING (f) Definition of terms: cost and marginal cost, revenue and
1. Time series marginal revenue, and
(a) Definition of terms: forecasting, time series profit functions
(b) Examples of time series and their uses / (g) Application of differentiation in maximization and
importance minimization of revenue, cost
(c) Components of time series: secular trend, and profit functions in production functions including
seasonal variation, cyclic variation and irregular determination of their
variation values at any given level
(d) Seasonal adjustment and deflating time Note: This excludes logarithmic differentiation, exponential
series differentiation,
2. Methods of computing trend derivatives of parameters, implicit functions and partial
(a) Moving average method and exponential differentiation.
smoothing I. DECISION THEORY
(b) Computation of trend by least squares 1. Types of decision making
method 2. Decision rules and decision trees
(c) Graphical representation of data 3. Computation of expectation or expected values
(d) Estimation of production or output using a 4. Construction and interpretation of results from decision trees
trend line 5. Advantages and disadvantages of decision trees
(e) Limitations and merits of moving averages 6. Determination of redundancy
and least squares methods J. LINEAR PROGRAMMING
3. Forecasting 1. Definition of terms: linear programming problem, objective
(a) Steps involved in forecasting function, constraints,
(b) Methods of forecasting: quantitative and feasible solution and optimum solution
qualitative, regression analysis 2. Assumptions applied in linear programming
and depersonalization method, exponential 3. Advantages and limitations of linear programming
smoothing, interpolation and 4. Formulation of linear programming model
extrapolation and their interpretation 5. Graphical and simplex methods of solving linear
(c) Construction and uses of the Z chart programming problems
H. LINEAR ALGEBRA AND CALCULUS 6. Advantages and disadvantages of graphical and simplex
1. Algebra methods
(a) Linear, quadratic and simultaneous equations 7. Optimization (maximization and minimization) in linear
in 2 or 3 variables programming
(b) Methods of solving quadratic equations by 8. Terms associated with simplex method: slack, surplus,
factorization, completing the square shadow prices, primal, dual
and use of the quadratic formula and tableau
(c) Methods of solving simultaneous equations by 9. Solving primal and dual linear programming problems
elimination, substitution and 10. Interpretation of primal and dual solutions
matrix methods 11. Advantages of the dual method
(d) Formulating equations involving quadratic Note:
and simultaneous equations in 2 (i) This excludes non- linear programming, transportation,
variables or 3 variables assignment and nonlinear
2. Functions and graphs dynamic programming.
(a) Explicit functions in one variable of (ii) Basic knowledge of solving simple linear inequalities and
polynomial nature up to the third degree (i.e. interpretation of matrix
f(x) = a + bx + cx2+ dx3 +……….) representation is assumed.
(b) Graphical representation of functions K. NETWORK ANALYSIS
3. Calculus 1. Definition of terms: network, activity, event, dummy variable, critical
(a) Geometric interpretation of differentiation path, float, cost
slopes, dangler and lead time
(b) Basic rule of differentiation: y = xn ,
2. Drawing of networks and Gantt charts and their interpretation
Dx/dy = nx n-1 3. Advantages and limitations of network analysis
(c) Differentiation methods: sum, difference, 4. Determination of earliest start time (EST), latest start time (LST), and
chain rule (substitution), product total cost
and quotient rule of the functions (including 5. Techniques of network analysis: PERT (program evaluation and
differentiation of fractional and review technique) and
negative indices) CPM (critical path method)
(d) Finding a second derivative 6. Crashing of projects, cost analysis
7. Resource and cost scheduling
(e) Application of differentiation (or derivatives)
8. Use of computers in network analysis
in determining maxima and minima
values of functions
234
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023
END AND
GOOD LUCK
235
HASSAN MBAZIIRA (0703-859 320). ICPAU- PAPER 2 (QT). 2022-2023