Chapter One

Introduction

1.1. Definition of Statistics

Definition: Different individuals have defined the term „statistics‟ differently. However,
the meaning of statistics can be divided in to two major categories. These are
 The plural sense (statistical data/numerical facts)
 The singular sense (statistical method /or as a field of study)

The plural sense: Statistics in its plural sense refers to numerical facts, figures or
statistical data. Example: population statistics, production statistics, etc.

But merely all-numerical data do not fulfill the quality of statistics. Hence, numerical
data must possess the following characteristics in order to be called statistics.
i) Statistics are numerically expressed. But every numerical data may not be
statistics. It should be well defined.
ii) Statistics are aggregates of facts. It should be general and does not concern about
individuals at specific manner.
iii) Statistics are affected to a marked extent by multiplicity of causes.
iv) Statistics are enumerated or estimated according to reasonable standards of
accuracy.
v) Statistics are collected in a systematic manner.
vi) Statistics are collected for predetermined purpose.
vii) Statistics should be placed in relation to each other in time and space. E.g. the
economy grows by 1.9% (Where and when?)

The singular Sense: Statistics is a branch of mathematics or applied research which is

concerned with the development and application of methods and techniques for collecting
organizing, analysing and interpreting quantitative data to come up with sound decisions.
Or, Statistics as a field of study has been defined as the art and science of collecting,
organizing, analysing and interpreting data. According to this definition there are five
steps in numerical investigation.

1
Classification of Statistics
Classification: two categories, namely, descriptive statistics and inferential statistics
Descriptive statistics deals with data collection, summarization, and analysis and give
interpretation (no conclusion or inference about population). We use descriptive
statistics simply to describe what is going on in our data at hand.
Inferential statistics it makes conclusion or inference about population parameter based
on sample statistics. Note that the sample must be taken from the population
about which we make conclusion.

Inferential statistics involves:

1. Estimation of the unknown population parameters

2. Testing hypotheses about these estimates
3. Make decision

1.2. Applications of Statistics

A. Statistics for decision making

Business decisions are much more complex. They aim at the maximization of output, or
minimization of input, or optimization of the outcome in the light of the changing market
environment, changing consumer preferences and intensifying competition. As the
science of inference, statistics is helpful in these respects.

B. Statistics for Market Research

Market research is perhaps the most common area of statistics application in business. It
aims at identifying any gap between customer needs and the products and services being
provided in the mass market. In doing so, it helps in testing the representativeness of the
data or information collected so that vital the survey result is reliable
C. Statistics for Forecasting

2
Limitations of Statistics

Indeed, statistics has greater significance in every branches of knowledge as for as it is

properly applied; but not without limitations. Some of the limitations of statistics are:
 It cannot be applied to all kinds of phenomena since it deals with only those
subjects that can be measured quantitatively and expressed numerically.
 It deals with only aggregates of facts and does not give any importance to
individual items.
 Statistical analysis often is based on sampling, which may not be accurate.
 It can be used to establish wrong conclusions if misused.

1.3. Methods of Data Collection and Data Presentation

1.3.1. Method of Data Collection

A number of methods can be used to gather data from the so called 'respondents'. Some
of the methods of data collection are:
i. Direct personal interview-as the name indicates directly ask the respondents
ii. Indirect personal interviews or observation-ask the 3rd parties about the
respondents
iii. Mail questionnaire
iv. Surveys or schedules through enumerators
v. Telephone interviews
vi. E-mail interviews etc.

A method of data collection depends on:

i) Nature and scope of the problem

ii) Cost, time and resources
iii) Degree of accuracy desired

3
 1.3.2. Data Presentation
The raw data that has been collected should be summarized and presented either using
tabular form or graphical form of data presentation.

A. Tabular Data Presentation

The Frequency Distribution (absolute, relative and cumulative frequency

distributions)
A frequency distribution is a listing of (mutually exclusive and exhaustive) intervals of
possible value for a variable, together with a tabulation of the number of observations in
each interval.
The relative frequency for an interval is the proportion of the sample observations that
f f
fall in that interval rf  or rf  n
where f is the number of times a given element
f
n
i
i 1

repeats itself (absolute frequency) and n is the # of obs.

Absolute Frequency Distribution
i Value of x (xi) Absolute Frequency (fi) Relative Frequency
1 150 1 1/33
2 152 4 4/33
3 154 6 6/33
4 156 2 2/33
5 158 2 2/33
6 160 7 7/33
7 162 2 2/33
8 164 4 4/33
9 166 1 1/33
10 168 3 3/33
11 170 1 1/33

4
Grouped Frequency

The number of items which fall in a class-interval is called its “class frequency”. The
tabulation of raw data by dividing the whole range of observations into a number of
classes and indicating the corresponding class-frequencies against the class-intervals is
called “grouped frequency distribution
Example 1:
A (fictitious) sample was taken of 200 professors at a local college. Each was asked for
his or her weekly salary. The responses ranged from about $520 to $590. If we wanted to
display the data in, say, 7 equal intervals, we would use an interval width of $10

Weekly Earnings Frequency Percentage

520 and under 530 6 3
530 and under 540 30 15
540 and under 550 38 19
550 and under 560 52 26
560 and under 570 42 21
570 and under 580 24 12
From 580 to 590 8 4
Total 200 100%

Several Important Terms in grouped frequency distribution

(a) Class-limits: The maximum and minimum values of a class-interval are called
upper class limit and lower class-limit respectively

(b) Class-mark, or, Mid-value: The class-mark, or, mid-value of the class-interval
lies exactly at the middle of the class-interval and is given by:
(lower class limit + upper class limit)
(i) Class-mark, or, Mid-value 
2
(lower class boundary + upper class boundary)
or
2
1
orLower class-limit  (upper class-limit - lower class-limit)
2

5
 (c) Class boundaries: Class boundaries are the true-limits of a class interval. It is
associated with grouped frequency distribution, where there is a gap between the
upper class-limit and the lower class-limit of the next class. This can be
determined by using the formula:

1
Lower class boundary = lower class-limit - d
2
1
Upper class boundary = upper class-limit + d ,
2
Where d = common difference between the upper class-limit of a class-interval and the
lower class limit of the next higher class interval. The class-boundaries convert a grouped
frequency distribution (inclusive type) into a continuous frequency distribution.

(d) Class-interval:
Class-interval = difference between two successive upper class-boundaries (or, two
successive lower class boundaries) or difference between two successive class marks, or,
mid values

i. Relative frequency in a grouped frequency distribution

The relative frequency of class shows the relative concentration of items in a given class
interval to the other classes of a frequency distribution.

class frequency
Relative frequency 
total frequency

Example 2:

Class Relative Frequency

boundaries Frequency In decimal In percentage
4.5 - 9.5 7 0.11 11%
9.5 - 14.5 10 0.16 16%
14.5 - 19.5 13 0.20 20%
19.5 - 24.5 18 0.28 28%
24.5 - 29.5 8 0.13 13%
29.5 - 34.5 5 0.08 8%
34.5 - 39.5 3 0.05 5%
Totals 64 1.00 100%

Therefore the relative frequency of class-interval 19.5–24.5 in the above table is 18/64 =
0.28 which is 28%.

6
 ii. Cumulative frequency

A frequency distribution becomes cumulative when the frequency of each class-interval

is cumulative. Cumulative frequency of a class-interval can be obtained by adding the
frequency of that class-interval to the sum of the frequencies of the preceding class-
intervals.

Often we want to know the number of cases which fall below, or, above a certain value.
Hence, there are two types of cumulative frequencies, i.e., (1) less than (or, from below)
cumulative frequency, and (2) more than (or, from above) cumulative frequencies. In the
less than type the cumulative frequency of each class-interval is obtained by adding the
frequencies of the given class and all the preceding classes, when the classes are
arranged in the ascending order of the value of the variable. In the more than type the
cumulative frequency of each class-interval is obtained by adding the frequencies of the
given class and the succeeding classes. For grouped frequency distribution, the
cumulative frequencies are shown against the class-boundary points.

Example 3: Construct the cumulative frequency distribution (both “less than” and “more
than” types) for example one
Less than cumulative Less than Frequency More than frequency
Less/more than 520 0 200
Less/ more than 530 6 194
Less/more than 540 36 164
Less/more than 550 74 126
Less/more than 560 126 74
Less/more than 570 168 32
Less/more than 580 192 8
Less/more than 590 200 0

Graphical Methods of Data Presentation

Histograms
When dealing with continuous random variables a different kind of graph is required.
This is called a histogram. At first sight these look similar to bar charts. There are,
however, two critical differences:
 The horizontal (x-axis) is a continuous scale. As a result of this there are no gaps
between the bars (unless there are no observations within a class interval);

7
  The height of the rectangle is only proportional to the frequency if the class
intervals are all equal. With histograms it is the area of the rectangle that is
proportional to their frequency.

Initially we will only consider histograms with equal class intervals. Those with uneven
class intervals require more careful thought. Producing a histogram is much like
producing a bar chart and in many respects can be considered to be the next stage after
producing a grouped frequency table. In reality, it is often best to produce a frequency
table first which collects all the data together in an ordered format. Once we have the
frequency table, the process is very similar to drawing a bar chart.
1. Find the maximum frequency and draw the vertical (y–axis) from zero to this
value, including a sensible numeric scale.
2. The range of the horizontal (x–axis) needs to include not only the full range of
observations but also the full range of the class intervals from the frequency table.
3. Draw a bar for each group in your frequency table. These should be the same
width and touch each other (unless there are no data in one particular class).
Example 4: Draw Histogram for example one

Frequency polygons
These are a natural extension of the relative frequency histogram. They differ in that,
rather than drawing bars, each class is represented by one point and these are joined
together by straight lines. The method is similar to that for producing a histogram:

8
 1. Produce a percentage relative frequency table.
2. Draw the axes
 The x-axis needs to contain the full range of the classes used.
 The y-axis needs to range from 0to the maximum percentage relative
frequency.
3. Plot points: pick the mid-point of the class interval on the x-axis and go up until
you reach the appropriate percentage value on the y-axis and mark the point. Do
this for each class.
4. Join adjacent points together with straight lines

Example 5: The Frequency polygon of the above Histogram is given below

Cumulative Frequency Curve (OGIVE)

Cumulative percentage relative frequency is also a useful tool. The cumulative
percentage relative frequency is simply the sum of the percentage relative frequencies at
the end of each class interval (i.e. we add the frequencies up as we go along
The Cumulative frequency curve of the above example is given below

9
 Less than
cumulative

Pie charts (Angular charts)

Pie charts are simple diagrams for displaying categorical or grouped data. These charts
are commonly used within industry to communicate simple ideas, for example market
share. They are used to show the proportions of a whole. They are best used when there
are only a handful of categories to display.

A pie chart consists of a circle divided into segments, one segment for each category. The
size of each segment is determined by the frequency of the category and measured by the
angle of the segment. As the total number of degrees in a circle is 360, the angle given to
a segment is 360◦times the fraction of the data in the category, that is

NumberinCategory
angle  *360
TotalNumberinSample(n)

Example: The following table shows the mode of transportation students‟ use on their
way to school. Represent it using pie chart

Mode Frequency
Car 10
Walk 7
Bike 4
Bus 4
Metro 4
Train 1

10
 Total 30

Measures of Central Tendency

In the previous lesson, we have learnt that the data could be summarized to some extent
by presenting it in the form of a frequency table. We have also seen how data were
represented graphically through bar graphs, histograms and frequency polygons to get
some broad idea about the nature of the data.

Some aspects of the data can be described quantitatively to represent certain features of
the data. The tendency of values of a variable to cluster around the central part of the
distribution is known as central tendency.

An average is one of such representative measures. As average is a number of indicating

the representative or central value of the data, it lies somewhere in between the two
extremes. For this reason, average is called a measure of central tendency. The purpose
of measures of central tendency is to locate the center of the distribution.

The Use of Summation Notation

In statistics the sum of large numbers of terms frequency occur. If a variable „X‟ has the
values X1, X2, X3, X4,……, XN, then X1 + X2 + X3 + X4 +……+ XN can be simplified
using the symbol (∑ Sigma).

11
 N
i.e., X1 + X2 + X3 + X4 +……+ XN = x
i 1
i

Example 1: Let X1 = 3, X2 = 9, X3 = 11

3
Find x
i 1
i

Solution

 x = X1 + X2 + X3 = 3 +9 + 11 = 23
i 1
i

Central Tendency Measures

Central tendency refers to the central location of a distribution. The most important
measures of central tendency are (1) the mean, (2) the median, and (3) the mode. We will
be measuring these for populations (i.e., the collection of all the elements that we are
describing) and for samples drawn from a population.

1. Arithmetic Mean/Average

You must have heard people talking about average speed, average rainfall, average
height, average score (marks) etc. If we are told that average height of students is 150 cm,
it does not mean that height of each student is 150 cm. In general, it gives a message that
height of students are spread around 150 cm. Some of the students may have a height less
than it, some may have a height greater than it and some may have a height of exactly
150 cm.

i. Simple Arithmetic Mean ( x )

Arithmetic average of Raw Data:

To calculate the mean of raw data, all the observations of the data are added and their
sum is divided by the number of observations. Let the variable „X‟ has values x1, x2, ....xn,
then, the arithmetic mean of „X‟ of n observations is denoted by x and given by

x1  x2  x3  .......  xn
x
n
n

x i
 i 1
(I)
n

12
Where the symbol “  ” is the capital letter „SIGMA‟ of the Greek alphabet and is used
to denote summation.

To economize the space required in writing such lengthy expression, we use the symbol
Σ, read as sigma.
n
In  xi , is called the index of summation.
i 1

Example 1: The weights of four bags of wheat (in kg) are 103, 105, 102, and 104. Find
the mean weight.

103  105  102  104

Solution: Mean weight ( x )  kg
4

414
 kg  103.5kg
4
Example 2: The enrolment in a school in last five years was 605, 710, 745, 835 and910.
What was the average enrolment per year?

Solution: Average enrolment (or mean enrolment)

605+710+745+835+910 3805
   761
5 5

Arithmetic Mean of Discrete frequency distribution

For the case of discrete frequency distribution in which „X‟ assumes values X1, X2, X3,
X4,……, Xk with respective frequencies f1, f2, f3, f4, ………, fk; then

f1 x1  f 2 x2  f 3 x3  .......  f k xk
x
f1  f 2  f3  ........  f k
(II)
k

fx i i
x i 1
k

f
i 1
i

Example 3: complete the arithmetic mean for the following discrete frequency
distribution.

Values 10 11 12 13
Frequencies 5 5 7 6

13
Solution
4

fx i i
x i 1
4

f
i 1
i

5(10)  5(11)  7(12)  6(13)


5576

267
  11.6
23

Arithmetic Mean of grouped frequency distribution

For the case of grouped frequency distribution we use the same formula as the formula in
equation (II) by taking Xi to represent class mark of the given class.

Example: Compute the arithmetic mean of the following grouped frequency distribution

Marks 10 – 19 20 – 29 30 – 39 40 - 49
No. of Students 4 5 8 6

Solution
The class marks of the given classes are:
LCL  UCL
CM  , then they will be 14.5, 24.5, 34.5, 44.5 respectively.
2
4

fx i i
x i 1
4

f
i 1
i

4(14.5)  5(24.5)  8(34.5)  6(44.5)


4586

723.5
  31.5
23

Mean is the most important measure to find the central tendency than the other two;
median and mode. But does the mean always represent the central portion? No it does
not, because if we take 2, 4, 6, the mean is 2 + 4 + 6/3 = 12/3 = 4, which is the central
portion, but if we take 2, 4, 27, then the mean will be 2 + 4 + 27/3 = 33/3 = 11. Here, the

14
mean is 11 and the central portion is 4. Therefore, the mean does not always represent the
central portion.

ii. Weighted Mean

If W1, W2, W3, …….., WN are the weights assigned to the values X1, X2, X3, ……, XN
respectively, then the weighted mean is given by:

W1 X 1  W2 X 2  W3 X 3  .......  WN X N
xw 
W1  W2  W3  .........  WN

W X i i
xw  i 1
N

W
i 1
i

Example:

A student was registered for five courses with 4, 4, 3, 2, & 3 credit hours. He obtained B,
A, C, D & A grades respectively. The grading system is of the form A=4, B=3, C=2,
D=1, &F=0. Then find the GPA (Grade Point Average) of the student.

Solution

Let W1=4, W2=4, W3=3, W4=2 and W5=3

Then X1=3, X2=4, X3=2, X4=1, and X5=4

Therefore the GPA of the student equals:

5

W X i i
xw  i 1
5

W
i 1
i

(4*3)  (4* 4)  (3* 2)  (2*1)  (3* 4)


4 43 23

48

16

3.00

15
 iii. Combined Mean

Let there be two sets of observations on the variable X. let N1 and N2 be numbers of
observations in the 1st and 2nd set respectively. Let x1 and x2 be the means of observations
in the 1st and 2nd set respectively. Then the mean of the combined (pooled) set N1 + N2
observations on X, denoted by x12 is given by:

N1 x1  N 2 x2
x12 
N1  N 2

Example

In a set given to two sections of a statistics course section 1(A) has a mean of 57.30,
section 2(B) a mean of 65.30. If there are 27 and 23 students in section 1 and 2
respectively, what is the mean of marks for the two sections combined together?

Solution

Let x1 be mean marks of section 1

x2 be mean marks of section 2
x12 be mean marks of section 1 and 2.

N1 x1  N 2 x2 27*57.3  23*65.3 3049

Therefore, x12     60.98
N1  N 2 27  23 50

Generally, if we have „K‟ different sets of data with N1, N2, N3, N4, ……. Nk number of
observations and x1 , x2 , x3 ,..........xk arithmetic means respectively, then the arithmetic
mean for the combined set of observations is given by the relation

N1 x1  N 2 x2  N3 x3 .........  N k xk
x123 
N1  N 2  N3  ........N k

Example

The mean mark obtained by 300 candidates in statistics is 46. The mean of the top 100 of
them was found to be 70 and the mean of the last 100 was known to be 20. What is the
mean of the remaining 100 candidates?

Solution

Here there are three different sets of observation with N1 = N2 = N3 =100

x123  46, x1  70, x3  20, x2  ?

16
 N1 x1  N 2 x2  N3 x3
x123 
N1  N 2  N3

100*70  100* x2  100* 20

46 
300

13800  7000  100 x2  2000

13800  9000  100 x2
100 x2  13800  9000
x2  48

2. Geometric Mean

Let „X‟ be a variable with values X1, X2, X3… XN , then the geometric mean of X
denoted by Mg or MG is defined as:
1
GM   X1 * X 2 * X 3 * .* X N  N

= N X 1 * X 2 * X 3 * . * X N for x i  0

Example:

Let X has value 2, 4, 6, 8, and then the geometric mean is

1
GM   2 * 4* 6 * 8  4

=4 2 * 4 * 6 * 8
= 4 384
=4.4267

There are two main uses of geometric mean

1. To average percent‟s, indexes and relatives
2. To determine the percentage increase in sales, in production or other business or
economic, series from one period to another.

Example

Suppose the profits earned by certain company on five projects were 3, 4, 4, 6, and 5
clusters respectively. What is the geometric mean profit?
Then the GM is equals to

17
 =5 3 * 4 * 4 * 6 * 5
= 5 1440
=4.28%
3  4  4  6  5 22
But the arithmetic mean profit is x    4.4% , here what we have to
5 5
know is for any positive number GM  Arithmetic mean, but if all values are equal GM
is also equal to AM.

3. The Harmonic Mean

The harmonic mean is the reciprocal of the arithmetic mean of the reciprocal of series of
observations i.e., if „X‟ has values of X1, X2, X3, ….., XN, then the harmonic mean for the
raw data will be computed as:
N
HM  N
1

i 1 X i

Whereas the harmonic mean for discrete and grouped frequency distribution is computed
as:
k

f i
HM  i 1
k
1
X
i 1 i
Example: Find the harmonic mean of the following data 2, 4, 6, and 8.
Solution:
N 4 4 24
HM  N  4   4*  3.84
1 1 1 1 1 1
     25
i 1 X i i 1 xi 2 4 6 8

Advantages of the MEAN:

 takes into account all observations
 Can be used for further statistical calculations and mathematical manipulations.

Disadvantages of the MEAN:

 Easily affected by extreme values.
 Cannot be computed if there are missing values due to omission or non-response.
 In grouped data with open-ended class intervals, the mean cannot be computed

18
Positional Measures of Central Tendency

Positional measures of central tendency are measures which are chosen because of their
position. The most popular positional measure is Median.

1. The Median

The median is a measure of central tendency that identifies the value of the middle case
when the cases have been placed in order or in line from low to high. The middle of the
line is as far from being extreme as you can get. There are as many cases in line in front
of the middle case as behind the middle case. The median is the attribute used by that
middle case. When you know the value of the median, you know that at least half the
cases had that value or a higher value, while at least half the cases had that value or a
lower value.

Be careful, here. The median is not about putting the attributes in order from low to high
and identifying the middle attribute. It is about putting the cases in order from low to
high based on their attributes and identifying the attribute used by the middle case. For an
ordinal variable with seven attributes, the median won‟t necessarily be the fourth
attribute. It might be, but it could be any of the seven attributes. It depends on how the
cases are distributed across the attributes.

In general if the given values of a variable ‘X’ are arranged in an increasing order, then,
the middle most value in this arrangement is called median of ‘X’ (denoted by Md or x )

Median for ungrouped data

There are two cases in which the median of ungrouped data is determined.
i. When the number of values, N, is odd, the middle most value is the
 N 1 
th

  value in the arrangement will be the unique median of X.

 2 
 N 1 
th

i.e., Md=the   value after arrangement.

 2 

ii. When N is even, there will be no unique median, then the median is given by the
th th
N N 
  value    1 value
Md     
2 2
formula:
2

19
Example

A large department store collects data on sales made by each of its salespeople. The
number of sales made on a given day by each of20 salespeople is shown below.

9 6 12 10 13 15 16 14 14 16 17 16 24 21 22 18 19 18 20 17

What is the median?

The sorted values: 6 9 10 12 13 14 14 15 16 16 16 17 17 18 18 19 20 21 22 24

Since n=20 is even,
th th th th
N N   20   20 
  value    1 value   value    1 value
Md    2  = Md     2 
2 2
=
2 2

The median for grouped frequency distribution

For grouped frequency distribution, the median is given by the following formula.
N 
  CFpMd 
Md  x  LCBMd    CW
2
Md
f Md
Where:
 LCBMd= Lower Class Boundary of Md Class
 N = Total number of observation (summation of frequencies)
 fMd= frequency of Md class
 CFpMd = cumulative frequency of pre-median class
 CWMd = Class Width of median class

Example:
Find the mean and the median for the following data.

No. of absent days No. of students

0–5 29
5 – 10 195
10 – 15 241
15 – 20 10
Total 475

20
Solution:

First let us calculate the less than cumulative frequency distribution for our observation
as follows.

S.N No. of absent No. of Less than

days students CFD
1 0–5 29 29
2 5 – 10 195 224
3 10 – 15 241 465
4 15 – 20 10 475

N
Here the total observation N, is 475 and  237.5 . Therefore, the median class is the
2
class which contains the value 237.5th observation which is the 3rd class. Then

 LCBMd = 10,
 CFpMd = 224,
 CWMd = 5
 fMd = 241

N 
  CFpMd 
  CW
2
Therefore x  LCBMd Md
f Md

x  10 
 237.5  224  5
241

 10  0.28

 10.28days

2. The Mode

Mode is another measure of position average. The mode of a distribution is the

value that has the greatest concentration of frequencies. It is thus a typical value
or measure of central tendency in as much as it is the most probable value in a
series. The mode is the value of the X variable in a frequency distribution that
occur the greatest number of times.

21
Measures of Variations

22