Ch.1 All & Ch.2 Introduction
Ch.1 All & Ch.2 Introduction
Topic Outline
1. Definition of statistics
2. Application of statistics in business
3. Stages of activities in statistics
4. Classification of data
5. Types of statistics
6. Ethical and reporting considerations in statistics
7. Synopsis
8. Wrap up discussion questions
9. Next session’s assignment
What is statistics?
Give examples of the use of statistics in different functions of business
List and explain types of data.
Define the two types of statistics.
What are the ethical and reporting considerations that must be taken in
statistical investigation?
Reading Text:
Types of Statistics
Statistics must be practiced with integrity and honesty; an independent and principled
point-of-view should be held when analyzing and reporting findings and results. In
statistics, both good and bad results should be presented in a fair, objective and neutral
manner; inappropriate summary measures should not be used to distort facts. The real
contribution of statistics to society is a moral one. For instance, financial analysts need
to provide information that truly reflects a company’s performance so as not to mislead
individual investors; and information regarding product defects that may be harmful to
people must be analyzed and reported with truthfulness.
Synopsis
Topic outline
1. Characteristics of Descriptive Statistics
2. Characteristics of Inferential Statistics
3. Synopsis
4. Wrap up discussion questions
5. Next session’s assignment
What are the similarities and differences of the two types of statistics?
Reading Text:
Statistics started as a purely descriptive science, but it grew into a powerful tool of
decision making as its inferential branch was developed. Modern statistical analysis
refers primarily to inferential statistics. However, descriptive and inferential statistics are
complementary.
Synopsis
Topic outline
1. Tables
2. Graphs
3. Diagrammatic Representation
4. Exploratory Data Analysis (EDA)
5. Synopsis
6. Wrap up discussion questions
7. Next session’s assignment
Tables:
Graphs:
Diagrammatic Representation:
Example 3
20, 18, 25, 68, 23, 25, 16, 22, 29, 37, 35, 49, 42, 65, 37, 42, 63, 65, 49, 42,
53, 48, 65, 72, 69, 57, 48, 39, 58, 67
Required
a. Develop the frequency distribution for the dataset, show the class
boundaries, class mark, relative frequencies, percentage relative frequencies,
less than and greater than cumulative frequencies and their percentages.
b. Draw the ogive
c. Show the stem and leaf display
Solution:
16, 18, 20, 22, 23, 25, 25, 29, 35, 37, 37, 39, 42, 42, 42, 48,48, 49, 49, 53, 57,
58, 63, 65, 65, 65,67, 68, 69, 72
Determine the number of classes (n). Ensure optimum size. You can use models
like n= 1+3.222 Log N ( N=total # of observation) or use 2 n≈N; thus
n=1+3222*log30=5.76≈6 classes
Determine width of Class intervals. Class width (Cw)=(Range/n)=(72-
16)=56/6=9.33≈10
b.
Ogive
40
30
GCF & LCF
20 LCF
10 GCF
0
15-24 25-34 35-44 45-54 55-64 65-74
1 68
2 023559
3 5779
4 2228899
5 378
6 3555789
7 2
Leaf unit=1
Figure 3.2 The Stem and Leaf Display for Example 3.1
2. The following frequency distribution represent the weekly wages of 100 entry-level
workers:
Required: Draw the histogram, frequency polygon, and the frequency curve.
Solution:
1. Assume the following data are taken from a sample salary of 40 employees in
thousands of Birr.
48 35 57 48 52 56 51 44 52 45
40 40 50 31 52 37 51 41 55 41
47 45 46 42 53 43 44 39 46 45
50 50 44 49 45 45 50 42 54 47
12 28 32 24 17 6 34 18 22 42 36 26
4. Construct a stem-and-leaf display and box plot for the following data which is
collected on the monthly rents paid (for public houses) by a sample of 20
households selected from the city of Addis Ababa.
429 540 650 585 578 1020 1070 780 989 930
870 1020 750 660 975 820 550 880 956 950
5. Give your own examples for Contingency table, Pareto-chart; scatter diagram, time
series plot, vertical line graph, pictogram, dot plot, and radar chart.
Synopsis
Descriptive statistics represents data in terms of tables and graphs for easy
understanding using different methods
Tables: Frequency distribution, and Contingency table, etc.
Frequency distributions can be represented in the form of graphs like histogram,
frequency polygon, frequency curve, ogive, and Lorenz curve.
Other graphs may include: bar graphs, pie chart, pareto chart, line graph (time
series plots, vertical line graph), scatter diagram, radar chart, pictogram,
cartogram, dot chart, etc.
Stem and leaf display and box plots are graphic (diagrammatic) displays used for
(EDA) exploratory data analysis (preliminary analyses of data in order to gain
insights about patterns and relationships)
Topic outline
Reading Text:
The arithmetic mean is the most widely used and widely reported measure of
central tendency.
Population mean µ=∑X/N; Sample Mean =∑X/n;
for grouped data: =∑f*cmi/n; f=frequency, cmi=class mark (midpoint), n=total
observation
The Median is the positional measure of average; it is the value (the average of
the two values) at the middle of the dataset sorted in ascending or descending
c n
order. For grouped data, Median x l
~ c. f ;
f 2
Where: l = LCB of the median class, c= class interval of the median class,
n
f=frequency of the median class, c.f=cumulative frequency just less than
2
The mode is the most frequently occurring value; for grouped data:
f1 f 0
Mode xˆ l c
1 f f 0 f1 f
2
or
1
Mode xˆ l
c
1 2
The Pth percentile is the value for which P percent are less than that value and
(100–P)% are greater than that value.
f (m ) 2
( mf ) 2
2 (Population) =
N
or m f N
2
N
Where; m = midpoint of a class; µ or = population or sample mean; F = Class
frequency; N (n) = total number of population (sample) size, N.B. for a sample
denominator is n-1
Standard deviation (σ or s) is the positive square root of the variance for
population or sample.
Coefficient of variation (V) measures relative dispersion: (population)V=σ/µ,
(sample) V=s/
3. Shapes of distribution refer to its symmetry or lack of it (skewness); and its
peakedness (kurtosis). It is the pattern of distribution of values from the lowest to
the highest value. Two or more distribution may have the same mean and equal
standard deviation but may differ in shape.
a. Skewness is the measure or degree of asymmetry; it influences the
relative position of the mean, median and mode; there are different
measures of skweness, one of which is Pearson’s coefficient of skewness
i.e. (Skp)=3(Mean-Median)/S.dev. Skp is 0 if distribution is symmetric. A
dataset can be:
o Symmetric (Mean=Median=Mode), left and right halves are equal;
o Left (negatively )skewed (Mean<Median< Mode); left tail is
elongated, few small extremes drag the mean to the left of the
median, E.g. consider dataset of exam completion time, normally,
few students finish their exams early, while the majority stays until
about the end of the exam
o Right (positively) skewed (Mean>Median>Mode); few large
extremes drag the mean to the right of the median, the right tail is
elongated. E.g. consider salary of employees in a company,
majority of workers salary is low, while the salary of only few
executives is high.
Example 4
A large department store collects data on sales made by each of its salespeople. The
sample of number of sales made on a given day by each of 20 salespeople is shown
below.
9, 6, 12, 10, 13, 15, 16, 14, 14, 16, 17, 16, 24, 21, 22, 18, 19, 18, 20, 17
X X- (X- )2 X2 (X- )4
6 -9.85 97.0225 36 9413.366
(∑X)2= 100489
(∑X)2 /n=5024.45
Exercise 4
1. A sample of 15 college seniors showed the following credit hours taken during
the final term of the senior year.
15 21 18 16 18 21 19 15
14 18 17 20 18 15 16
a. Find the sample mean, median and mode number of credit hours taken during
the final term
b. Find the following fractiles: P60, P90; Q1, Q3, and D3,
c. Find the range, inter-quartile range, variance, standard deviation, and coefficient
of variation for the number of credit hours taken during the final term
d. Find the Pearson’s coefficient of skewness and coefficient of kurtosis for the
distribution and describe its shape.
2. The time taken to serve each of a sample of 100 customers at a bank was
observed. The following table gives the frequency distribution of service times for
these 100 customers.
Service time (minutes) Number of customers
0-2 18
2-4 30
4-6 24
6-8 16
8-10 8
10-12 4
Required: Provide the following sample descriptive statistics and interpret results
List the basic types of summary measures, their methods, and purpose
How is the shape of a distribution measured?
Explain the relative position of the mean, median, and mode in symmetric and
skewed distributions
Topic outline
1. Concepts of probability
2. Approaches to assign probabilities
3. Synopsis
4. Wrap up discussion questions
5. Next session’s assignment
Descriptive statistics focuses with summarizing something that has already happened. It
is inferential statistics that is involved in making predictions, generalizations &
conclusions based on the sample taken from the populations. Since the probability of
error exists in statistical inference, estimates or tests of a population characteristic are
given together with the chance or probability of being wrong. Probability theory
therefore, forms the basis for inferential statistics (statistical inference), and other fields
that require assessment of occurrences. Much of the decision making environment
involves uncertainties, for instance consider the following:
What is the chance that the new product will be welcome in market?
What is the chance that the new machinery will increase product?
How likely will the project be completed on time?
Thus, knowledge of probability is required so that calculated risk would be taken.
Probability is a numerical measure of the relative likelihood that a particular event will
occur. It is the science of uncertainty. It is the mathematical means of studying
uncertainty & variability. Probability theory provides mechanism for measuring &
analyzing uncertainties associated with future events. Values of probability for an event
ranges between 0 and 1(100%) inclusive; the former represents (non-occurrence or
impossible event), and the latter indicates sure event.
Sample space is the set that consists of all the possible outcomes of a random
experiment. E.g. in tossing 3 coins the sample space: {T T T, HT T, THT, HHT, THH,
HTH, HHT, HHH}; this sample space has 8 (sample points) or elements.
Assigning Probabilities:
Probability for a single event can be assigned objectively (based on information) using
the classical, and empirical approach; or subjectively (based on a person’s belief or
estimate of an event’s likelihood).
This is based on relative frequencies. E.g. 10 students scored A out of 100 registered
students for the course Managerial statistics last semester, based on this observation,
the probability that a student will score A =10/100=0.1
Exercise 5
1. A newly established company is planning to recruit trainees for four jobs in the
marketing department. The marketing manager contacted an employment
agency. The agency has selected four candidates and sends them to the
Company. The company will hire those who fulfill the requirement of the job.
Assume that a candidate’s chance to pass/fail the final evaluation is equally
likely. Use P=pass and Pc=fail
a. List all the sample space outcomes for the experiment
b. Identify the sample outcomes corresponding to the following events
i. None of them will qualify (pass)
ii. At least three of them will qualify (pass)
Synopsis
Define probability, random experiment, event, sample space, and sample point.
Why should probability be studied in statistics?
Explain the objective and subjective approaches of assigning probability.
Topic outline
Reading Text:
If A is any event, then the complement of A, denoted by A or Ac, is the event that A
does not occur. For instance, if A is an event that refers to Head, the complement of A
will be Tail in a single toss. The sum of all complementary events probabilities amount
to 1:
P(A) + P(Ac) = 1.
P(Ac) = 1 - P(A).
Thus, if we know the probability of an event, we can find the probability of its
complementary event by subtracting the given probability from 1.0.
Example 6.1: In a given football match the chance of winning (W) is 25% and the
chance of tying (T) is 40%, what is the chance of losing (L)?
P(L)=1-P(W)+P(T)=1-(0.25+0.4)=1-0.65=0.35
Rule of Addition
A. For mutually exclusive events (events that do not occur at the same time or have
common point):
Probability of at least one, either, P(A or B)=P(AuB)= P(A)+P(B)
Example 6.2: Ethiopian Airlines tests female candidates for hiring hostesses, if the
chance of being underweight (U), overweight (O) and satisfactory(S) is 0.025, 0.075
and 0.9 respectively; what is the probability that a randomly selected female candidate
can be satisfactory or overweight?
P(SuO)=0.075+0.9=0.975
B. For non-mutually exclusive events (events that can occur at the same time or
have common point
P(AuB)=P(A or B)=P(A)+P(B)- P(A B); Note P(A B)=P(A and B)
Example 6.3: A student is sitting for an entrance exam on English and Mathematics;
the probability that he passes in English is 0.65 and the chance that he passes in math
is 50%. Also the chance that he passes both is 0.25;
P(EUM)=0.65+0.5-0.25=0.9
Example 6.4: There are 16 eggs in a container 6 of which are rotten (E), 2 eggs are
successively selected randomly without replacement. Find the (joint) probability that
both eggs being spoiled.
B. For independent events (the occurrence of A is not connected in any way to the
occurrence of B)
Example 6.5: Two coins are tossed. What is the probability that both will land tail (T)
up? Prove by listing.
Exercise 6
Synopsis
Probability rules:
o Complementary events rule
P(A) + P(Ac) = 1.
P(Ac) = 1 - P(A).
o Addition rule
(Non-mutually exclusive events) :P(AuB)=P(A or B)=P(A)+P(B)-
P(A B)
(Mutually exclusive events): P(AuB)=P(A or B)= P(A)+P(B)
o Multiplication rule (Joint probability)
(Dependent events): P(A B)=P(AandB)=P(A)xP(B/A) or
P(B)xP(A/B);
(Independent events): P(A B)=P(A and B)=P(A)xP(B)
Topic outline
1. Random variable and probability distribution definitions
2. Synopsis
3. Wrap up discussion questions
4. Next session’s assignment
Reading Assignment Discussion:
What is a random variable?
Define probability distribution
What are the requirements for valid probability distribution?
Reading Text:
Depending on the nature of the numeric value of the outcomes, a random variable
could be Discrete or Continuous. The former assumes only certain clearly separated
values resulting from count. The later assume values that fall in a specific interval
infinitely and mostly results from measurement.
E.g. Discrete random variable: number of customers entering a bank (0, 1, 2…); Score
of a football game; A discount offered by a retail store that is always either 5% or 10%;
Continuous random variable: Distance between A.A. & Adama (100km or 100.15 or
100.155…),the time of commercial flight between Nairobi and Addis Ababa; and today’s
outside temperature .
The probability distribution of a discrete random variable possesses the following two
characteristics:
1. The probability assigned to each value of a random variable x lies in the range of
0-1 inclusive. That is, 0 ≤ P(x) ≤ 1 for each value of x.
N.B.: if random variable is X, its probability is denoted by P(x)
2. The sum of the probabilities assigned to all values of outcomes of x is equal to 1.
That is, ΣP(x) =1.0. Remember, if the probabilities are rounded, the sum may
not be exactly 1.0.
It is possible that all numerical values for a discrete random variable can be listed in a
table with accompanying probabilities. There are several standard probability
distributions that can serve as models for a wide variety of discrete random variables
involved in business applications. Some standard discrete probability distributions
known by name are: the binomial probability distributions, hypergeometric probability
distributions, Poisson probability distributions, multinomial probability distributions,
geometric probability distributions and etc. The first three enlisted types of discrete
probability distributions will be discussed in subsequent sections.
On the other hand, for a continuous random variable, all possible fractional values of
the variable cannot be listed, and therefore the probabilities that are determined by a
mathematical function are portrayed graphically by a probability density function, or
probability curve. Several standard probability distributions that can serve as models for
continuous random variables are described in latter sections.
Example 7
Develop a discrete probability distribution for number of heads in three tosses of a fair
coin. (Note: H stands for head, and T stands for tail)
Results from 3 coin tosses (sample space outcomes):
T T T, T TH,THT, HT T, THH, HTH, HHT, HHH
X=The random variable i.e. number of heads obtained, it is a discrete random variable
0 1 2 3
X (# of Heads)
F(x) or Frequency 1 3 3 1 ∑f(x)=8
P(X) (0.5 chance) 1/8=0.125 3/8=0.375 3/8=0.375 1/8=0.375
Note: Each outcome (head and tail) is equally likely; P(H)=0.5 or 1/2 and P(T)=0.5 or
1/2 . For instance, probability of two heads (HTH)=3/8 i.e. calculated as follows:
1/2*1/2*1/2=1/8, and as 2 Heads occur 3 times; 1/8+1/8+1/8=3/8 or 3*1/8=3/8.
Exercise 7
1. Classify the following random variables as discrete or continuous.
a. The number of students in a class
b. The time taken to run a marathon
c. The number of cattle owned by a farmer
d. Amount of today’s rainfall
e. The amount of fuel in the tank
f. The height of soldiers
g. The age of a house
h. The number of new accounts opened at a bank during a month
i. The number of pages in a book that contain at least one error
j. The time spent by a physician examining a patient
2. Develop a discrete probability distribution for number of girls in four births that
are expected.
3. A consumer agency surveyed 2500 families living in the small town to collect
data on the number of television sets owned by them. The following table lists
the frequency distribution of the data collected by this agency.
Number of TV owned: 0 1 2 3 4
Number of families: 850 1000 400 200 50
a. Construct a probability distribution table for the number of TV sets owned
by these families. Draw the (vertical line) graph of the probability
distribution.
b. Determine the probability of the number of TV sets owned by families
are:
Exactly 1
1 to 3
more than 2
less than or equal to 1
at most 2
at least 2
Synopsis
Topic outline
How do you calculate the expected mean, variance and standard deviation for
a probability distribution?
Reading Text:
As we can compute the mean, variance and standard deviation for a frequency
distribution; likewise, we can do so for probability distribution. The formula for finding
the mean (µ) or expected value E(x), variance (σ2) and standard deviation (σ) is given
below respectively:
; ;
Also variance can be calculated as:
Example 8
Compute the mean (expected value) for the probability distribution developed
previously in Example 7.
Table 8.1 Computing Mean, Variance, and Standard deviation for Example 8
Xi (# of Heads) 0 1 2 3 Total
Xi2 0 1 4 9
P(Xi) 1/8=0.125 3/8=0.375 3/8=0.375 1/8=0.125
∑Xi*P(Xi)
Xi*P(Xi) 0 3/8=0.375 6/8=0.75 3/8 =0.375
µ=1.5
Xi-µ (1.5) -1.5 -0.5 0.5 1.5
[Xi-µ]2 2.25 0.25 0.25 2.25 5
[Xi-µ]2.P(Xi) 0.28 0.094 0.094 0.28 ∑Xi-µ)2*P(Xi)
σ2=0.75
X2.P(X) 0 3/8=0.375 1.5 1.125 ∑Xi2*P(Xi)=3
Number of Calls 0 1 2 3 4
Frequency 8 10 22 9 1 =50
Synopsis
o Variance (x)= ;
How is the expected value, variance and standard deviation calculated for a
discrete probability distribution?