INTRODUCTION TO STATISTICS

CHAPTER ONE

INTRODUCTION

1.1 Definition and classification of statistics

The subjects of statistics, as it seems, is not a new discipline but it is as old as the human society
itself. The sphere of its utility, however, was very much restricted.
Definition:
 Statistics is a collection of numerical facts and data.
 Statistics is a mathematical science dealing with the methods of collection, organizing the
collected data, presentation, analysis and interpretation of the data.
 Statistics is a subject that deals with numbers and figures describing certain situations. It
primarily deals with numerical data taken by surveys and summarizes these data in such a
way that this summary gives a good indication about the nature of the data.

The word “statistics” could be singular or plural. The definition given in the second place above
might be taken as the singular form of “statistics”.
Statistics, in its singular sense is a subject area or field of study. It is defined as science, which
deals with the collection, processing, analysis, interpretation and presentation of numerical facts.

 .The word “statistics” is derived from the Latin word “status” indicating the historical
importance of governmental data gathering, which related to demographic information
(military recruitment and tax collecting). Thus, the scope of statistics in the ancient times
was primarily limited to the collection of demographic and property and wealth data of a
country by governments for framing military and fiscal policies.

Now days, statistics is used almost in every field of study, such as natural science, social science
engineering, medicine, agriculture, e t c.

Classification:
Statistics is broadly divided into two categories based on how the collected data are used.
1. Descriptive Statistics
deals with describing data without attempting to infer anything that goes beyond the given set
of data,
Consists of collection, organization, summarization and presentation of data.
For example, tables or graphs are used to organize data, and descriptive values such as the
average score are used to summarize data
2. Inferential Statistics
deals with making inferences and/or conclusions about a population based on data obtained
from a limited sample of observations,
Consists of performing hypothesis testing, determining relationships among variables and
making predictions.
 Examples:
0
a) From past figures, it has been predicted that 31 0 of registered voters will vote in the November
election..
b) To determine the most effective dose of a new medication (on the basis of tests performed with
volunteer patients from selected hospitals)
c) To compare the effectiveness of two reducing diets (based on the weights losses of persons who
were taking the diets)
d) There is a relationship between smoking tobacco and an increased risk of developing cancer.

1.2 Definition of some basic terms

a) Population: Is the totality (collection) of all objects or items under consideration.
Example: All university students of Ethiopia, All staff members of MettuUniversity.
b) Sample: Is a part of a population taken so that some generation about the population
can be made. A sample should be a representative of the population. Example :
c) Parameter: is a descriptive measure of a population, or summary value calculated
from a population. Examples: Average, Range, proportion, variance,
d) Statistic: is a descriptive measure of a sample, or summary value calculated from a
sample.
Example: Average, Range, proportion, variance …

1.3Application and limitation of statistics

Statistics can be applied in any field of study which seeks quantitative evidence. For instance (in
engineering)

 To compare the breaking strength of two types of materials

 To determine the probability of reliability of a product.
 To control the quality of products in a given production process.
 To compare the improvement f yield due to certain additives (fertilizer, herbicides, (wee
decides), e t c
to see some applications
1. agriculture: which types of varieties gives the best yield which feed mixtures should
chickens be fed so that they will gain most weight
2. medical and pharmaceutical research :
However, Statistics has the following limitations.
a) It does not study qualitative characteristics directly Examples: Beauty, honesty, poverty,
and standard of living.
b) It does not study a single individual but deals with aggregate of facts. Example: The
population size of a country for some given year does not help us for comparative
studies.
c) Statistical results are true only on the average. Examples: The probability of getting a
head in tossing a coin is 1|2 the germination percentage of a given variety of seed is 80%
d) It is sensitive for misuse: Examples: The number of car accidents committed in a city in
a particular year by women drivers is 10 while that committed by men drivers is 40.
Hence women drivers are safe drivers.
 1.4 TYPES OF VARIABLES AND MEASUREMENT SCALES

TYPES OF VARIABLES

A variable is a characteristic of an object that can have different possible values.

There are two types of variables.

a) Quantitative variables: are variables that can be quantified or can have numerical
values. Examples: height, area, income, temperature e t c.
b) Qualitative variables: are variables that cannot be quantified directly. Examples: colour,
beauty, sex, location qualitative variables are also called categorical variables. And hence
we have two types of data; quantitative & qualitative data.
Quantitative variables can be further classified as
 Discrete variables, and
 Continuous variables

a) Discrete variables are variables whose values are counts.

Examples: number of students, number of households (family size), Number of pages of a book.
b) Continuous variables are variables that can have any value within an interval.

Examples: weight, Length, Volume, e t c.

MEASUREMENT SCALES
There are four measurement scales
1. Nominal scale: - “Nominal “is a Latin word for “name” This is a scale for grouping
individuals into different categories.
Examples: red, brown, black, short, tall, pass, fail
 In this scale, one is different from the other
 +, -, *, /, Impossible, comparison(ordering) is impossible
2. Ordinal scale: - “ ordinal” is a Latin word, meaning “order”

 It is a scale for grouping and ordering of individuals in to different

categories.
 Data consisting of an ordering of ranking of measurements are said to be
on an ordinal scale of measurements.
Examples: Faster, taller, shorter, military ranks, ranks in race, e t c.
 One is different from and grater /better/ less than the other.
 +, -, *, / Are impossible, comparison is possible.
Man a weighs more than man B
Ethiopian athletes got 1st and 2nd ranks in the 10,000m women’s final in Sydney.
Ordinal scales data contain and convey more information than the nominal scale data, for relative
magnitudes are known, however, quantitative comparisons are impossible.

3. Interval scale: is a measurement scale in which:

 There is no true zero point (arbitrary zero point)
  There is no physical significance to the zero point.
 There is a constant interval size between any adjacent units on the measurement
scale.
o
Example: c oF (Measuring units of temperature)
 In this measurement scale
One is different, better/greater and by a certain amount of difference than another (Possible to
add and subtract but multiplication and division are not possible)
37Oc – 35oc = 2oc
45oc – 43 oc= 2oc
40oc = 2(20oc) But this does not imply that an object which is 40 oc is twice as hot as an object
which is 20 oc (oF = 9/5, oc +32)
40 oc → 9/5 x 40 oc + 32 = 104 oF
20 oc → 9/5 x20 oc + 32 = 68 oF

o
O
c = 5/9 ( F- 32)
60 oF→ 5/9 (60 – 32) = 15.56 oc
30 oF→ 5/9 (30 - 32) = -1.11 oc
 Interval scale data convey better information than nominal and ordinal scale data.
4. Ratio scale: is a measurement scale in which
 There is a constant interval size between any adjacent units on the measurement scale.
 There exists a zero point on the measurement scale and that there is a physical
significance to this zero point.
Examples: m, cm, kg, km/hr, cm/sec
Year, hour, second, ok, m3, e t c.
 One is different, larger /taller/ better/ less by a certain amount of difference and so much
times than the other.
 (+, -, *, / Are possible on this scale)
 This measurement scale provides better information than interval scale of measurement

 The following present a list of different attributes and rules for assigning numbers to
objects. Try to classify the different measurement systems into one of the four types of
scales. (Exercise)
1. Your checking account number as a name for your account.
2. Your checking account balance as a measure of the amount of money you have in that
account.
3. The order in which you were eliminated in a spelling bee as a measure of your spelling ability.
4. Your score on the first statistics test as a measure of your knowledge of statistics.
5. Your score on an individual intelligence test as a measure of your intelligence.
6. The distance around your forehead measured with a tape measure as a measure of your
Intelligence.
7. A response to the statement "Abortion is a woman's right" where "Strongly Disagree" = 1,
"Disagree" = 2, "No Opinion" = 3, "Agree" = 4, and "Strongly Agree" = 5, as a measure of
attitude toward abortion

Chapter Two
 Methods of Data Collection and Presentation

2.1 Sources of data

Any scientific investigation requires data related to the study. The required data can be obtained
from either a primary source or a secondary source.

Primary source: Is a source of data that supplies first hand information for the use of the
immediate purpose.

 a data obtained from in the following way considered as primary source

a) Field trials
b) Laboratory experiments
c) Surveys – census survey
- Sample survey.
Secondary source: are individuals or agencies, which supply data originally collected for other
purposes by them or others.
- Usually they are published or unpublished materials, records, reports, e t c.
2.2 Classification and Tabulation of Data
Classification: - is the process of arranging items/data into classes or categories according to their
similarities and/or differences.

Classification eliminates inconsistency and also brings out the points of similarity and/or dissimilarity of
collected items/data.

Classification is necessary because it would not be possible to draw inferences and conclusions if we have
a large set of collected [raw] data.

2.3 Frequency Distributions

Frequency: - is the number of times a certain value or set of values occurs in a specific group.

A frequency distribution is a table that presents data according to some criteria with the corresponding
number of items falling in each class (i.e. with the corresponding frequencies.)

Example: A frequency distribution presenting the number of males and females in a class
Sex Frequency
Male 57
Female 39

Generally, there are two basic types of frequency distributions: Ungrouped and Grouped frequency
distributions.

1. Ungrouped frequency distribution

Ungrouped frequency distribution is a table of all potential raw scored values that could possibly occur in
the data along with their corresponding frequencies. Ungrouped frequency distribution is often
constructed for small set of data or a discrete variable.

Constructing an ungrouped frequency distribution

To construct an ungrouped frequency distribution, first find the smallest and the largest raw scores in the
collected data. Then make a columnar table of all potential raw scored values arranged in order of
magnitude with the number of times a particular value is repeated, i.e., the frequency of that value. To
facilitate counting method, tallies can be used.

Example: The following data are the ages in years of 20 women who attend health education last year:
30, 41, 39, 41, 32, 29, 35, 31, 30, 36, 33, 36, 32, 42, 30, 35, 37, 32, 30, and 41.
Construct a frequency distribution for these data.
STEP 1. Find the range of the data:
Range=Maximum observation−Minimum observation
STEP 2. Construct a table, tally the data and complete the frequency column. The frequency distribution
becomes as follows.

Age Tally Frequency

29 / 1
30 //// 4
31 / 1
32 /// 3
33 / 1
35 // 2
36 // 2
37 / 1
39 / 1
41 /// 3
42 / 1

2. Grouped frequency distribution

When the range of the data is large, the data must be grouped into classes. Grouped frequency distribution
is a frequency distribution when several numbers of data are grouped into one class.

Some Important Definitions

– Raw data: data collected in original form.
– Array: data arranged, in ascending or descending order.
– Class: the different, on overlapping groups of data.
– Class limits: separate one class in a grouped frequency distribution from another. The limits could
actually appear in the collected data and have gaps between the limit of one class and the lower limit
of the next class.
– Class boundaries: separate one class in a grouped frequency distribution from another. The
boundaries have one more decimal place than the raw data and therefore do not appear in the
collected data. There is on gap between the upper boundary of one class and the lower boundary of
the next class. The lower class boundary (LCB) is found by subtracting 0.5 units of measurement
from the lower class limit (LCL) and the upper class boundary (UCB) is found by adding 0.5 units of
1
measurement to the upper class limit (UCL). That is, LCB=LCL+ 2 U and UCB
1
=UCL + 2 U
– Class width (W): the difference between the upper and lower boundaries of any class or the lower
limits of two consecutive classes, or the upper limits of two consecutive classes.
N.B. Class width is not equal to the difference between UCL and LCL of the same class.
– Class mark (M): the midpoint of a class interval.
 UCBi + LCB i
M=
i.e. 2
– Unit of measurement (U): the smallest difference between any two values of the variable being
measured.
– Cumulative frequency (Cf) less than type: the total frequency of all values (observations) less
than or equal to the upper class boundary for the given class.
– Cumulative frequency (Cf) more than type: The total frequency of all values (observations)
greater than or equal to the lower class boundary for the given class.

A tabular arrangement of class intervals together with their corresponding cumulative frequency
(either less than or more than type; as defined above) is called a cumulative frequency
distribution.
– Relative frequency: the frequency a class divided by the total frequency (i.e. sum of all frequencies)
and, if multiplied by 100, gives the percent of values falling in that class.
Frequency of that class
Re lative frequency of a class=
Total frequency
Note:
 The relative frequency shows what fractional part or proportion of the total frequency belongs to
the corresponding class.
 The sum of all the relative frequencies in the frequency distribution is always 1.
– Relative cumulative frequency (less than type/ more than type): total of the relative frequencies above/
below a class inclusively. Or the cumulative frequency (less than type/more than type) divided by the
total frequency. This gives the percent of values which are less than/more than the upper/lower class
boundary.

Guidelines to construct a grouped frequency distribution

STEP 1. Determine the unit of measurement, U
STEP 2. Find the maximum(Max) and the minimum(Min) observation, and then compute their range, R
Range=Max−Min
STEP 3. Fix the number of classes desired (k). there are two ways to fix k:
– Fix k arbitrarily between 6 and 20, or

Use Sturge’s Formula:

–
k =1+3 .32 log N
10 where N is the total frequency. And round
this value of k up to get an integer number.
STEP 4. Find the class widths (W) by dividing the range by the number of classes and round the number up
R
W=
to get an integer value. K
STEP 5. Pick a suitable starting point less than or equal to the minimum value. This starting point is the
lower limit of the first class. Continue to add the class width to this lower limit to get the rest of
the lower limits.
STEP 6. Find the upper class limits. To find the upper class limit of the first class, subtract one unit of
measurement from the lower limit of the second class. Then continue to add the class width to this
upper limit so as to get the rest of the upper limits.
STEP 7. Compute the class boundaries as: LCB=LCL− 12 U and UCB=UCL+ 12 U
Where LCL = lower class limit, UCL= upper class limit, LCB= lower class boundary and UCB= upper
class boundary. The class boundaries are also half way between the upper limit of one class and the lower
limit of the next class.
STEP 8. Tally the data.
 STEP 9. Find the frequencies.
STEP 10. (If necessary) Find the cumulative frequencies (more than and less than types).
Example: The number of hours 40 employees spends on their job for the last 7 working days is given
below.
62 50 35 36 31 43 43 43
41 31 65 30 41 58 49 41
37 62 27 47 65 50 45 48
27 53 40 29 63 34 44 32
58 61 38 41 26 50 47 37
Construct a suitable frequency distribution for these data using 8 classes.
STEP 1. Unit of measurement; U= 1year
STEP 2. Max = 65, Min = 26 so that R = 65-26 = 39
STEP 3. It is already determined to construct a frequency distribution having 8 classes.
39
W= = ¿5
STEP 4. Class width 8
STEP 5. Starting point = 26 = lower limit of the first class. And hence the lower class limits become
26 31 36 41 46 51 56 61
STEP 6. Upper limit of the first class = 31-1 = 30. And hence the upper class limits become
30 35 40 45 50 55 60 65
The lower and the upper class limits (Steps 5 and 6) can be written as follows.
Class limits
26 – 30
31 – 35
36 – 40
41 – 45
46 – 50
51 – 55
56 – 60
61 – 65
STEP 7. By subtracting 0.5 units of measurement from the lower class limits and by adding 0.5 units of
measurement to the upper class limits, we can get lower and upper class boundaries as follows.

Class
boundaries
25.5 – 30.5
30.5 – 35.5
35.5– 40.5
40.5– 45.5
45.5– 50.5
50.5– 55.5
55.5– 60.5
60.5– 65.5
STEPS 8, 9 and 10 are displayed in the following table (columns 3, 4 and 5&6 respectively).
Class limits Class Tally frequency Cumulative Cumulative
boundaries frequency (less frequency
than type) (more than type)
26 – 30 25.5 – 30.5 ///// 5 5 40
31 – 35 30.5 – 35.5 ///// 5 10 35
36 – 40 35.5– 40.5 ///// 5 15 30
41 – 45 40.5– 45.5 ///// //// 9 24 25
 46 – 50 45.5– 50.5 ///// // 7 31 16
51 – 55 50.5– 55.5 / 1 32 9
56 – 60 55.5– 60.5 // 2 34 8
61 – 65 60.5– 65.5 ///// / 6 40 6

2.3 Diagrammatic and Graphic Presentation of Data

The data that is presented by a frequency distribution can also be displayed diagrammatically or
graphically.
Diagrams and graphs:
- are techniques for presenting data in visual displays using geometric figures;
- are visual aids which give a bird’s eye view about a given set of numerical data;
- have greater attraction than mere figures (numbers);
- facilitate comparison of data;
- are easily understandable by anyone who does have no statistical background
Usually diagrams are appropriate for presenting discrete data, whereas graphs are appropriate for
presenting continous types of data.
There are three common diagrammatic presentations of data: bar-diagram/charts, pie-chart and
pictograms, as well as three common graphic presentations of data: histogram, frequency polygon, and
cumulative frequency polygon (ogive).

I. Bar-diagrams/ Bar-charts
- Bar-diagram is a series of equally spaced bars having equal width and the height of each bar
representing the magnitude or frequency of observations in each group.
- Bar-diagrams are usually used to represent one way or simple frequency distribution.
- Bar-diagrams can be drawn either horizontally or vertically. Usually horizontal bar-diagrams are used
for qualitatively classified data whereas vertical bar-diagrams are used for quantitatively classified
data.
Example: Horizontal bar-diagram.

There are a number of bar-diagrams. The most common being:

- Simple bar-diagrams
- Deviation (two-way) bar-diagrams
- Broken bar-diagrams
- Component (subdivided) bar-diagrams
- Multiple bar-diagrams
1. Simple bar-diagrams
Simple bar-diagrams are used to depict data of single variable or one-way variable.
Example: The following frequency distribution shows sales of production (in million birr) of three
products for 2004 production year.
Produc Sale (in million)
t
 A 14
B 21
C 9
D 17
The bar-diagram presentation for these data is given below.

2. Deviation bar-diagrams
When the data take both positive and negative values (for instance data on profit, net export, percent
change, etc) deviation bar-diagrams are appropriate.

Example: Present the following data using a suitable bar-diagram.

Data: Net profit (in thousands birr) in oil sales for five years
Year Profit (in
thousands)
1997 12
1998 -5
1999 14
2000 9
2001 -6
The deviation bar-diagram for the data looks like the following.

3. Broken bar-diagrams
This kind of bar-diagram is used to present data involving a few extreme values where it will be difficult
to accommodate the magnitude of the bars corresponding to these values within the graph paper. In this
case we use pieces of bars with each piece starting with a jump on the numerical scale.
Example: Data: - Amount of production per a day for four products of a factory.

Product Quantity
produced (kg/day)
A 14
B 35
C 23
D 109
4. Component bar-diagrams
When it is desired to show how a total (an aggregate) is divided into component parts, we use component
bar diagram. In such type of bar-diagrams, the bars represent aggregate value of a variable with each
aggregate broken into its component parts and different colors or designs are used for identification.

Example: Represent the following data using bar-charts

Data: Yields of production of farmers in Southern Ethiopia.
Year  1990 EC 1991 EC 1992 EC 1993 EC
Crop
Barley 14 15 26 19
Wheat 10 15 14 25
Maize 2 6 10 3
Total 26 36 50 47

The component bar-diagram for this table is as follows

5. Multiple bar-diagrams
Multiple bar-diagrams are used to display data on more than one variable. They are used for comparing
different variables at the same time.

Example: The data given in the above example can be presented by using multiple bar-diagram as below.
II. Pie-charts
A pie-chart is a circle that is divided into sections or wedgrs according to the percentages of frequencies
in each category of the distribution. The angle of the sector of a class is obtained by multiplying the ratio
of the frequency of the class to the total frequency by 3600.
frequency of the class
i.e. sec tor angle of a class= ×3600
total frequency
Note that pie-charts are usually used for depicting nominal level data.

Example: A survey showed that a car owner spends birr 2,950 per year on operating expenses. Below is
the breakdown of the various expenditure items. Draw an appropriate chart to portray the data.

Expenditure item Amount (in

birr)
Fuel 603
Interest on car loan 279
Repairs 930
Insurance and license 646
Depreciation 492
Total 2,950

How to draw a pie-chart

- First find the percentages of each class
- Next calculate the degree measures for each class
- Finally, using a protractor, put each sector /degree measure/ in a circle and give a key for
explanation.

Expenditure item Amount (in Percentage Degree

birr) (approx) (approx)
Fuel 603 20 74
Interest on car loan 279 9 34
Repairs 930 32 113
Insurance and license 646 22 79
Depreciation 492 17 60
Total 2,950 100 360

Now we can draw the pie-chart for the data.

 Key
Fuel
Insurance and license
Repairs
Interest on car loan
Depreciation

III. Pictograms
In pictograms, we represent the data by means of some picture symbols. Here we decide a suitable picture
to represent a definite number of units in which the variable is measured.

Example: Draw a pictorial diagram to present the following data (number of students in a certain school
for four years.)

Year 1992 1993 1994 1995

No. of students 2000 3000 5000 7000

Let a single picture () represents one thousand students.

1995 
1994  Key: = 1000 students
1993 
1992 

IV. Histogram
A histogram is another way of data presentation which is more suitable for frequency distributions with
continuous classes.
In drawing a pictogram, we put the class boundaries of each class on the horizontal axis and its respective
frequency on the vertical axis.
Example: Draw a histogram presenting the following data.

Frequency Cumulative Cumulative

Class Boundaries Class Mark Frequency (less Frequency (more
than type) than type)
5.5 – 11.5 8.5 2 2 20
11.5 – 17.5 14.5 2 4 18
17.5 – 23.5 20.5 7 11 16
23.5 – 29.5 26.5 4 15 9
29.5 – 35.5 32.5 3 18 5
35.5 – 41.5 38.5 2 20 2
V. Frequency Polygon
A frequency polygon is a line graph drawn by taking the frequencies of the classes along the vertical axis
and their respective class marks along the horizontal axis. Then join the cross points by a free hand curve.

Example: Present the data in the previous example using a frequency polygon.

VI. Cumulative Frequency Polygon (Ogive)

Cumulative frequency polygon can be traced on less than or more than cumulative frequency basis. Place
the class boundaries along the horizontal axis and the corresponding cumulative frequencies (either less
than or more than cumulative frequencies) along the vertical axis. Then join the cross points by a free
hand curve.
Example: the data in the previous example can be presented using either a less than or a more than
cumulative frequency polygon as given below (i) and (ii) respectively.
(i) Less than type cumulative frequency polygon

(ii) More than type cumulative frequency polygon