CONTENTS

 Variables and Concepts

 Types of variables

 Hypothesis and Types

 Testing the Hypothesis

 Errors and Types

1
VARIABLE

"An attribute that is observable,

measurable, and has a dimension that
can vary".

For example, temperature is a variable

that is observable, measurable, and
varies from high to low.
2
CONCEPTS VS. VARIABLES
 Concepts are mental images or perceptions
 Meaning vary from individual to individual
 Variables are measurable
 With varying degrees of accuracy

A variable can be measured, a concept

can not be.

 Concepts in your study?

 Need 3
to convert them to variables
CONCEPTS, INDICATORS & VARIABLES
 Concepts in your study – How to measure?
 Identify the indicators
 A set of criteria reflective of the concept
 Convert the indicators to variables
 Example
 Concept – Rich
 Indicator – Income & Assets
 Income(in $) is also a variable
 Assets
 House, Cars, Investements
 Convert each of these into dollars

 Based on total income and total value of assets, decide 4

whether a given person is rich or not.

 CONVERTING CONCEPTS TO VARIABLES

Concept Indicators Variables Decision level

Rich 1. Income 1. Income/year 1. If >$100,000
2. Assets 2. Total value: 2. If > $250,000
1. Home
2. Cars

High academic 1. Marks exam 1. Percentage 1. If > 80%

achievement 2. Marks practical 2. Percentage 2. If > 80%

5
TYPES OF VARIABLES
 Classification can be based on:
 The causal relationship
 The design of study
 The unit of measurement

6
TYPES – FROM VIEW OF CAUSATION
 Change variables – Independent variables
 The cause responsible for bringing about a change in a
phenomenon or situation
 Variable that is believed to cause or influence the dependent
variable

 Outcome variables – Dependent variables

 Variable that is influenced by the independent variable.

 Extraneous variables
 Variables affecting the cause-and-effect relationship
7
EXAMPLES

Does Smoking Cause Lung cancer ?

Does Nursing care Cause Rapid recovery ?

Does Drug (a) Cause Improvement ?

Cause Effect
Independent variable Dependent variable

8
EXAMPLES
Extraneous
• Variables

Variable that confound the relationship between the

dependent and independent variables, thus it needs
to be controlled.

E.g., "air pollution" is an extraneous variable

interferes with studying the relationship between
smoking "independent variable" and lung cancer
"dependent variable".

9
VARIABLES – VIEWPOINT OF STUDY
DESIGN
 Active variables
 Variables that do not pre-exist, so, the researcher has to create
them.
 These variables can be manipulated, changed or controlled.

 Attribute Variables
 A pre-existingcharacteristic or attribute which the researcher
simply observes and measures.
 These variables cannot be manipulated, changed or controlled

10
EXAMPLE
 Study designed to measure the effectiveness of three
teaching models A,B,C

 Researcher may change the teaching model

 No control on the characteristics of the student

population – age, gender or motivation to study

11
VARIABLES – MEASUREMENT
VIEWPOINT
 Categorical Variables (Qualitative)
 Continuous Variables (Quantitative)

12
VARIABLES – MEASUREMENT
VIEWPOINT
 Categorical Variables
 Measured on nominal scales

 Two types
 Dichotomous Variables
 Vary in only two values.
 E.g. alive or dead, day or night etc.

 Polytomous Variables
 More than two categories
 E.g. Religion – Muslim, Christian, Jew

13
VARIABLES – MEASUREMENT
VIEWPOINT
 Continuous Variables
 Continuity in measurement – take any value on the scale
on which they are measured
 E.g. age, income etc.

14
Hypothesis

15
HYPOTHESIS
 Hypothesis
 Brings clarity, specificity and focus to research problem

 Possible to conduct a study without hypothesis as well

 Hypothesis – how to construct

 Arise from ‘hunches’ or ‘educated guesses’

16
HYPOTHESIS – EXAMPLES
 Betting on a horse race
 Hunch – Horse#6 will win
 Hunch is true or false – Only after the race

 Distribution of smokers
 Hunch – more male smokers at your workplace than female
smokers
 Test the hunch – ask them
 Conclude – hunch was right or wrong

17
HYPOTHESIS – EXAMPLES
 Public health
 A disease is very common in people coming from a
specific sub-group of population
 To find every possible cause – enormous time and
resources
 Narrow down – based on your study identify the
most probable cause e.g. contaminated water
 Perform a study – collect information to verify
your hunch
 Verificiation – hunch correct or not

18
HYPOTHESIS - EXAMPLES
 In example 1
 Waited for event to take place

 In example 2 & 3
 Designed a study to test the validity of your hunch

19
HYPOTHESIS
 Researcher – does not know about a phenomenon, situation
or a condition
 But – does have a hunch, assumption or guess

 Conclude through verification

 Hunch may be
 Right
 Wrong
 Partially right

20
HYPOTHESIS - DEFINITIONS
 A tentative statement about something, the validity of
which is usually unknown

 A proposition that is stated in a testable form and that

predicts a particular relationship between two or more
variables.

 A hypothesis is written in such a way that it can be

proven or disproven by valid and reliable data – it is in
order to obtain these data that we perform our study.
21
22
HYPOTHESIS - CONSIDERATIONS
 A hyothesis should be simple, specific and clear
 No ambiguity in the hypothesis – makes verification difficult
 Unidimensional – should test one relationship at a time
 Must be familiair with the subject area (literature review)
before suggesting the hypothesis

23
HYPOTHESIS - CONSIDERATIONS

The average age of male students in the class is

higher than that of female students

 Clear
 Specific

 Testable

24
HYPOTHESIS - CONSIDERATIONS

Suicide rates vary inversely with social cohesion

 Clear
 Specific

 Testable?
 Difficult – What is social cohesion, how to measure it.

25
HYPOTHESIS - CONSIDERATIONS
 A hypothesis should be capable of verification
 Data collection and analysis
 Hypothesis cannot be tested?
 May forumulate hypothesis for which methods of verification not
available
 You may end up developing a technique

 A hypothesis should be operationalisable

 Expressed in terms that can be measured

26
TYPE OF HYPOTHESIS
 Categories of hypothesis
 Research hypothesis

Your hypothesis which you want

to test

 Alternate hypothesis

Specify the relationship that will be

considered as true in case the research
hypothesis proves to be wrong.
27
WAYS OF FORMULATING HYPOTHESIS
 There is no significant difference in the proportion of male
and female smokers in the study population

 A greater proportion of females than males are smokers in

the study population

 A total of 60% of females and 30% of males in the study

population are smokers

 There are twice as many female smokers as male smokers in

the study population 28
WAYS OF FORMULATING HYPOTHESIS
 Hypothesis of No Difference
 When you formulate a hypothesis stipulating that there is no
difference between two situations, groups or outcomes

 There is no significant difference in the proportion of male

and female smokers in the study population

29
WAYS OF FORMULATING HYPOTHESIS
 Hypothesis of Difference
 A hypothesis in which a researcher stipulates that there will be a
difference but does not specify its magnitude

 A greaterproportion of females than males are smokers in the

study population

30
WAYS OF FORMULATING HYPOTHESIS
 Hypothesis of Point-Prevalence
 A researcher has enough knowledge about the
behaviour/situation

 Able to express the hypothesis in quantitative units

 A totalof 60% of females and 30% of males in the study

population are smokers

31
WAYS OF FORMULATING HYPOTHESIS
 Hypothesis of Association
 Expressed as a relationship

 Twice as many female smokers as male smokers

32
HYPOTHESIS TESTING
 Hypothesis testing - H0
 Null hypothesis
 Usually corresponds to a default "state of nature", for example "this
person is healthy", "this accused is not guilty" or "this product is not
broken".

 Alternate hypothesis
 Negation of null hypothesis, for example, "this person is not healthy",
"this accused is guilty" or "this product is broken ".

 Errors depend directly on null hypothesis.

33
34
HYPOTHESIS TESTING

True state of nature

H0 is True H0 is False
Your Decision

Reject H0

Accept H0

35
HYPOTHESIS TESTING

True state of nature

H0 is True H0 is False
Your Decision

Reject H0 Type I error Correct Decision

Accept H0 Correct Decision Type II error

36
 H0 is True H0 is False

Reject H0 Type I error Correct Decision

Accept H0 Correct Decision Type II error

HYPOTHESIS TESTING
 H0 = This person is healthy

Telling the person that he is sick when infact he was healthy Type I error

Telling the person that he is sick when infact he was sick Correct

Telling the person that he is healthy when infact he was sick Type II error

Telling the person that he is healthy when infact he was healthy Correct

Traditionally probability of type I errors is denoted by α

and that of type II errors by β
37
HYPOTHESIS TESTING

H0 = Defendent is Innocent 38
EXAMPLE – AIRPORT TRAVELERS

True state of nature

Innocent Terrorist
Your Decision

Terrorist False positive True positive

(rejected)

Innocent True Negative False negative

(Accepted)

- truth/falseness of the null hypothesis and outcomes of the test

- Positive means you rejected the null hypothesis
39
- Negative means you accepted the null hypothesis
EXAMPLE: FACE DETECTION
True Positives

False Negative

False Positives

40
True Negative (Rest of the image) Neg- Acceptance of null
Pos- Rejection of null
Null- majority not a face
PERFORMANCE MEASURES
Recall
TP
R
TP  FN

Precision
TP
P
TP  FP
F measure

Precision . Recall
F  2. 41
Precision + Recall
 TP TP
P R
TP  FP TP  FN

EXAMPLE: FACE DETECTION

How many faces
were detected out of
total?

Recall = 3/4= 75%

Did system detected

extra objects other
than faces?

Precision = 3/6 = 50%

42
EXAMPLE - BIOMETRICS
 Biometric access control system
 Finger print, iris, face, hand geometry etc.

 Enrollment
 Enrollall the authorized users – take their finger prints, facial
images or iris scans etc.

 Validation
 A person arrives
 Take data (finger print, iris, face)
 Compare with database
 If matched with an individual – Allow 43

 Else - Decline
 EXAMPLE - BIOMETRICS
Enrollment

What kind of errors the system can make?

44

http://www.idteck.com/support/biometrics.asp
 The FRR is the frequency that an
authorized person is rejected access
EXAMPLE

45
The FAR is the frequency that a non authorized person is
accepted as authorized
EXAMPLE - BIOMETRICS
 Challenge
How to find a similarity threshold value for
acceptance/rejection
Find system response to a large number of inquires from
authorized as well as unauthorized users.
Record similarity scores of authorized and unauthorized
cases
Plot respective histograms/distributions

46
EXAMPLE - BIOMETRICS

47
EXAMPLE - BIOMETRICS

48
EXAMPLE - BIOMETRICS

49
EXAMPLE - BIOMETRICS

50
Move the decision boundary (threshold) to the right

FAR will decrease and FRR will increase 51

Move the decision boundary (threshold) to left

FAR will increase and FRR will decrease 52

Which boundary to chose?

Depends upon your application – Which errors are less serious 53

HOW TO QUANTIFY SYSTEM
PERFORMANCE
 On different thresholds system has different values of
FAR and FRR
 If some one asks you what is the performance of your
system – how to answer?

54
HOW TO QUANTIFY SYSTEM
PERFORMANCE
 Equal Error Rate - EER
 Change the value of threshold and plot FAR and FRR

 The point where both are equal is the EER

55
 HOW TO QUANTIFY SYSTEM
PERFORMANCE
 The Receiver Operating Characteristic (ROC) Curve

50

High security – 40
cannot afford
FAR 30
FRR (%)

Balance
20
User comfort – Lesser
10 False Rejections

0
0 10 20 30 40 50
FAR (%) 56
The material in these slides is based on the following resources.

REFERENCES

 Research Methodology, Ranjit Kumar, Chapter 6

 http://en.wikipedia.org/wiki/Type_I_and_type_II_errors
 http://www.intuitor.com/statistics/T1T2Errors.html
 http://www.fingerprint-it.com
 http://fingerchip.pagesperso-orange.fr

57