Multivariate Analysis: Dr. Raghuvir Singh

Multivariate Analysis
Dr. Raghuvir Singh

Multivariate Analysis: Cluster,
Factor & Conjoint
These methods do not treat same variables as

independent & same as dependent.
These methods try to identify interdependencies

among a number of variables without treating any
one of them as dependent or independent.
Understanding the following four questions can
help the students better know these methods:
What is the typical problem that can be studied

with the method.
What does the method do?
What types of variables can be analysed?
Does the method identify interdependencies
among a no. of variables without treating any of

them as dependent or interdependent.
Cluster Analysis
Applications in market segments studies.
It identifies different groups (or clusters) of
respondents such that the respondents in any one

cluster are similar to each other but different from
the respondents in the other clusters.
It is typically applied to data consisting of many
variables that have been collected from a large

sample of respondents.
The cluster analysis procedure search through
the data & identify respondents who have given

identical, or least similar, answers to a certain
combination of questions. These respondents are
formed into one cluster
The cluster analysis procedure then searches through the looking

for a second set of respondents, all of whom have given similar

answers to some other combination of questions.
This second set of respondents are similar to each other, but they

are also quite different from the respondents in the first cluster.
This procedure can be continued till all of the different clusters
have been identified.
Types of Variable Used

Cluster analysis is typically applied to data that have been

recorded on interval scales such as 5-,7-, or 10 points but can also

be applied on continuous scale in few cases.
The cluster analysis procedure then searches through the looking

for a second set of respondents, all of whom have given similar

answers to some other combination of questions.
This second set of respondents are similar to each other, but they

are also quite different from the respondents in the first cluster.
This procedure can be continued till all of the different clusters
have been identified.
Types of Variable Used

Cluster analysis is typically applied to data that have been

recorded on interval scales such as 5-,7-, or 10 points but can also

be applied on continuous scale in few cases.
Cluster analysis identifies interdependencies
among variables:
Till now we have been dealing with single variable

to identify a class or a category into which a
respondent belong.
There are situations in which a researcher may
wish to use more than one variable to identify the

class or category in to which a respondent
belongs.
A marketer uses multi-variable classification in
large number of marketing decision making

situations.
2 Step cluster analysis procedure-
Developing a measure to identify similar
respondents, &
Developing a procedure for grouping similar
respondents.
Distance - A commonly used similar measure

(Subscripts)
D = √(x11-x15 )2+(x21-x25 )2+(x31-x35 )2+-----+(xn1-xns )2

15
Distance between respondents 1 & 5 for variables x1,

x2, x3,------------------- xn
Distance is calculated between every pair of
respondents
A procedure for grouping similar
respondents-
•There must be a procedure whereby each respondent is evaluated
to determine if he or she should be placed with a particular group
or cluster. The procedure should have a ‘rule’ for placing a
respondent into a given group.
•Single linkage rule, it states that a respondent will be placed into a
group if the distance between that respondent & any single
respondent already in the group is similar than some pre
established minimum distance.
The procedure for forming clusters is done (explained) with a tree
diagram called a dendogram.
Factor Analysis
•It is used to study a complex product or service in order to identify the major
characteristics (or factors) considered to be important by consumer of the product
or service.
What Factor Analysis Does: Using data from a large sample, F.A. applies an
advanced form of correlation analysis to the responses to number of statements.
•The purpose of analysis is to determine if the response to several of the statements
are highly correlated.
•If the responses to three or more statements are highly correlated, it is believed
that the statement measure some factor common to all & them.
•Some study involves many statements, there are likely to several sets of such
correlated statements.
The statements in any one set are highly
correlated with each other but are not highly
correlated with the statement in any of the other
set.
For each such set of highly correlated statements,
the researchers use their own judgement to
determine what the single “factor” of “ theme” is
that ties the statements together in the minds of
respondents.
Types of variable used:

can only be applied to continuous variables or
intervally scaled variables.
Factor analysis identifies
interdependencies:
•It (cluster) identifies respondents who gave the same
answers to a number of questions.
•F.A. identifies two or more questions that result in set of
responses that are highly correlated.
Factor Analysis Procedure:
a)Three important measures

I.Variance
II.Standerdised score of an individual’s response
III.Correlation coefficients
b. The role of correlation
c. The identification of factors
d. The output of a factor analysis- factor loading s–
associated with a specific factor & a specific statement
is simply the correlation between the factor & that
statement”
e. Evaluating how well the factor fit the data.

Multivariate Analysis: Dr. Raghuvir Singh

Uploaded by

Copyright:

Available Formats

Multivariate Analysis: Dr. Raghuvir Singh

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Multivariate Analysis: Dr. Raghuvir Singh

Uploaded by

Copyright:

Available Formats

Multivariate Analysis

Dr. Raghuvir Singh

These methods do not treat same variables as

These methods try to identify interdependencies

What is the typical problem that can be studied

What types of variables can be analysed?

Does the method identify interdependencies

among a no. of variables without treating any of

respondents such that the respondents in any one

variables that have been collected from a large

the data & identify respondents who have given

for a second set of respondents, all of whom have given similar

Types of Variable Used

recorded on interval scales such as 5-,7-, or 10 points but can also

for a second set of respondents, all of whom have given similar

Types of Variable Used

recorded on interval scales such as 5-,7-, or 10 points but can also

Till now we have been dealing with single variable

wish to use more than one variable to identify the

large number of marketing decision making

Distance - A commonly used similar measure

D = √(x11-x15 )2+(x21-x25 )2+(x31-x35 )2+-----+(xn1-xns )2

Distance between respondents 1 & 5 for variables x1,

Types of variable used:

Factor Analysis Procedure:

a)Three important measures

You might also like