Quantitative

Techniques
Diving into Statistics
Session Data and
Statistics
Segment Introduction to Data
and Statistics
 HOW DATA HAS EVOLVED - A TIMELINE
Ancient
1800s 1990s
World
Ishango bone – tally stick Herman Hollerith’s punch Sir Tim Berners Lee
for numerical records. card machine used to creates hyperlink for
Writing and libraries process US census data . data sharing.
evolve. Internet invented.
Google search launched.

John Graunt studies Fritz Pfleumer invented

London death records to magnetic tape used for
predict life expectancy. recording.
“Data” used in English.
1800s- 1900s-
1600s 1900s
1850s 1950s
 What is Data?

 Data is a collection of facts.

 These facts can be words, numbers, descriptions
or observations.
 It is used to understand, analyse or represent a
process or a phenomenon.
 Examples of Data in Business:
⚫ Financial data
⚫ Customer purchase records
⚫ Inventory data
⚫ Human Resource data
 BASIC STATISTICAL CONCEPTS

What is Statistics?

01
Collection, analysis,
interpretation and presentation
of numerical data
How do we classify Statistics?

Where do we apply it? 03 Two types:

 Descriptive Statistics
 Inferential Statistics

02
Population – large collection of
persons, objects or items

Sample – a small part of

population
Segment Statistics in
Business
NETFLIX: STATISTICAL ANALYSIS OF CONTENT CONSUMPTION

Viewed shows/movies

Language Artists

Location Genre

Viewing
History
 TYPES OF STATISTICAL ANALYSIS

Descriptive Inferential

 Using data that represents  Using a smaller portion of

a defined group to draw data about a group as a
conclusions about the representative sample to
same group draw conclusions about
 Example: Calculating the entire group
student feedback based on  Example: Estimating
a survey taken by the customer satisfaction by
entire class. surveying a random sample
of 100 customers.
 Descriptive Statistics
 Descriptive statistics is used to understand and
describe the characteristics or patterns of a
dataset.
 Using descriptive methods, we can summarise and
present data in an informative way.
 DESCRIPTIVE ANALYSIS IN EVERYDAY LIFE

Car Manufacturing
01 Calculating the time required to finish
manufacturing a car based on extensive factory
floor data.

Cricket
02 Concluding the “Man of the Match” post-game
after studying the performance of all players

Marathons
03 Concluding the “Man of the Match” post-game
after studying the performance of all players

Healthcare
04 Deducing the average time taken by a
participant to complete a marathon after all
participants have crossed the finish line
 Inferential Statistics
 Inferential statistics refers to the process of
selecting a representative sample to draw out
inference for the whole population.
 The conclusions or decisions are made using a
portion of the population from the universe.
 INFERENTIAL ANALYSIS IN EVERYDAY LIFE

Car Manufacturing
01 Estimating the time taken to manufacture based
on a sample of 50 cars from a total of 5000 cars

Cricket
02 Guessing the “Man of the Match” in the game
of cricket based on the past performance of
the players

Marathons
03 Predicting the average time to finish a
Marathon based on the finish times of 30
random runners

Healthcare
04 Estimating the effectiveness of immunisation
by studying a random sample of 100 children
for a state
 APPLICATIONS OF STATISTICS

Weather Forecast Business Decision Stock Market Prediction

Clinical Trials Quality Control Research and Development

 ADVANTAGES OF STATISTICS
Applications in Business

Helps deal with uncertainties by forecasting

1 economic fluctuations

Assists organisations in adopting data-driven approach

2 to optimise profits

Helps in measuring business performance and

3 understand trends for sound decision making
Segment Data: Types and
Measures
 Data Measures: Origin
The first form of early data is tally or tick marks.
They were collected in order to keep track or
record of inventories such as food for ancient
civilisations. Later, the abacus was invented to help
with the calculation of such records.

Abacuses Around the World

 VARIABLE VS. DATA

What is Variable? What is Data?

 Variable is any attribute or  Data is collected set of information

characteristics that describes the which is used for interpretation and
object or entity (person, place, presentation.
thing) of study. The value of the  In statistical term, it is observed
variable may vary from one entity values of variable or collection of
to another. variables.

Column/ Variable

Row/
Observation

Data
 TYPES OF DATA

Data

Quantitative Qualitative

Discrete Continuous Nominal Ordinal

 QUANTITATIVE AND QUALITATIVE DATA

Quantitative Qualitative

 Numerical data that can be  Descriptive data that can

easily measured. be observed but cannot be
 Example: measured.
⚫ Number of patients  Example:
⚫ Level of blood sugar ⚫ Color of a car
⚫ Viewer rating for a TV
show
 QUANTITATIVE - DISCRETE

Can take only integer values and

1 be counted

Examples:
Number of iPads sold
2 Number of complaints at call center
Number of employees

3
 QUANTITATIVE - CONTINUOUS

Can take any values including

1 decimals and is measured

Examples:
Percentage return on a stock
2 Glucose quantity in blood
Infant mortality rate
 QUALITATIVE - NOMINAL

Data used without a natural order

1 in the values

Examples:
Gender
2 Colour
City
 QUALITATIVE - ORDINAL

Data needs to be placed in an order

1 based on categories before use.

Examples:
Education levels (BS, MS, PhD)
2 Customer rating of item (Poor, Avg, Good)
Mutual fund risk ratings
DICRETE, CONTINUOUS, NOMINAL AND ORDINAL DATA

Age Job Education Level Balance (Dollars) Variables

36 Admin Secondary 336.1

34 Admin Tertiary 242.0
62 Retired Secondary 66.1
27 Teacher Secondary 325.0
47 Manager Primary 524.0
38 Manager Tertiary 897.5
45 Manager Secondary 114.0
38 Teacher Primary -1.0
67 Retired Primary 118.0
72 Retired Primary 222.0

Discrete Nominal Ordinal Continuous

 Levels of Data Measurement
 Defines the relationship between the attribute
and the value of the variable
 Helps decide the statistical analysis that can be
used on the data
 Example: Cricket Jersey numbers
⚫ Attribute: Name; Value: Numbers 1-12
⚫ The numbers just act as a label and don’t signify
anything else.
 LEVELS OF DATA MEASUREMENT

First Level
1 Nominal Level of Measurement

Second Level
2 Ordinal Level of Measurement

Third Level
3 Interval Level of Measurement

Fourth Level
4 Ratio Level of Measurement
 LEVELS OF DATA MEASUREMENT
Nominal Measurement

Purpose
 To classify, name or categorise data

Meaning
 It does not denote any order of classification. Used only as a label
to differentiate one from another

Example
 Which of the following employment classifications best
describes your area of work?
⚫ 1. Educator 2. Construction worker 3. Manufacturing
worker 4. Lawyer 5. Doctor 6. Other
⚫ We can only interpret the name
⚫ Name: Profession
 LEVELS OF DATA MEASUREMENT
Ordinal Measurement

Purpose
 To rank or order data

Meaning
 It denotes the superiority or inferiority among data but does not
denote the amount of the same.

Example
 This online training is:
⚫ 1 - not helpful
⚫ 2 - somewhat helpful
⚫ 3 - moderately helpful
⚫ 4 - very helpful
 We can interpret the name and rank
 Name: Whether or not helpful; Rank: Ascending order from not
helpful to very helpful
 LEVELS OF DATA MEASUREMENT
Interval Measurement

Purpose
 To categorise and rank order data with equal intervals

Meaning
 Gives meaning to distances between consecutive numbers

Example
 Numbers on a temperature scale : 10 Degree Celsius
 We can interpret the name, rank and distance between two values
 Name: Degree Celsius
 Rank: 100o is hotter than 10o
 Distance: Equal interval between 2-4 degrees and 8-10 degrees.
 LEVELS OF DATA MEASUREMENT
Ratio Measurement

Purpose
 To derive relation between different parts of data

Meaning
 The ratio of two numbers is meaningful and denotes the relation

Example
 Weight: 80 pounds
 We can interpret the name, rank, distance and it makes sense as a ratio.
 Name: Pound
 Rank: 80 pounds is heavier than 40 pounds
 Distance: Intervals between 20-40 pounds and 60-80 pounds is equal
 Absolute zero: Ratios of two weights is meaningful
Segment Visualising Data
 AN ANALYSIS OF DIGITAL USAGE IN INDIA

Visualized Data Data India Digital Statistics

(figures in billions)
1.6
Mobile Phone Internet Social
Population
1.2 Connections Users media users

0.8

1.37 1.06 0.0687 0.0400

0.4

0.0
Population Mobile Connections

Internet Users Social Media users

 Graphical Representation
 Provides an overall picture of the data and help
reach some useful conclusions merely by studying
the chart or graph
 CONCEPTS IN GRAPHICAL REPRESENTATION

01 02 03

Range Number of classes Width of class interval

Difference between largest and Partitions in the data Approximate width =

smallest number Range / Number of classes
 Frequency Distribution
 Summary of data presented in the form of class
intervals and frequencies
 Used to present data in a classified manner
 CONCEPTS IN FREQUENCY DISTRIBUTION

The value halfway across The proportion of the total The running total of
the class interval and can be frequency that is in any frequencies through the
calculated as the average of given class interval in a classes of a frequency
the two class endpoints. frequency distribution. distribution.
The representative value for The individual class Frequency for present class
each class in most group frequency divided by the interval added to the
statistics calculations. total frequency. preceding cumulative total.

Class Midpoint Relative Frequency Cumulative Frequency

 TYPES OF VISUALISATION

Line Chart Scatter Plot Pie Chart Bar Graph Histogram

 GRAPHICAL REPRESENTATION
Frequency Polygon/Line Chart

Each class frequency is plotted as a dot at Cricket matches held

1 the class midpoint, and the dots are 100
connected by a series of line segments.
80
60
Class mid points on X-axis and frequency on
2 Y-axis 40
20

Connecting all plotted points gives a 0

3 Frequency polygon 1980 1990 2000 2010
India Pakistan Australia

India Pakistan Australia

1980 10 20 30
1990 20 30 90
2000 70 70 50
2010 40 50 60
 GRAPHICAL REPRESENTATION
Scatter plot

80
Two-dimensional graph plot of pairs of
1 points from two numerical variables
60

40
Shows the extent of relation between
2 variables
20

0
0 1 2 3 4 5

India Pakistan
10 20
20 30
70 70
40 50
 GRAPHICAL REPRESENTATION
Pie Charts
India
Circular depiction of data where the area of
1 the whole pie represents 100% of the data
and slices of the pie represent a percentage
breakdown of the sublevels.

Show relative magnitudes with the whole

2 and proportions to be determined for
construction.

Connecting Can lead to less accurate

3 judgments than other graphs.

1980 1990 2000 2010

Year India
1980 10
1990 20
2000 70
2010 40
 GRAPHICAL REPRESENTATION
Bar Graph

Cricket Games Held

Contains two or more categories along one
1 axis and a series of bars, one for each
2010
category, along the other axis.
2000
Length of the bar represents the magnitude
2 of the measure for each category. 1990

1980

0 25 50 75 100
Australia Pakistan India
India Pakistan Australia
1980 10 20 30
1990 20 30 90
2000 70 70 50
2010 40 50 60
 GRAPHICAL REPRESENTATION
Histogram

Cricket games held

A series of contiguous bars or rectangles
1
100
that represent the frequency of data in given
class intervals. 80
60
A useful tool for differentiating the
2 frequencies of class intervals. A quick glance
at a histogram reveals which class intervals
40
20
produce the highest frequency totals.
0
Insights can be gleaned by examining the 1980 1990 2000 2010
3 histogram and show that histograms play an
India Pakistan Australia
important role in the initial analysis of data.

India Pakistan Australia

1980 10 20 30
1990 20 30 90
2000 70 70 50
2010 40 50 60