Chapter

Seventeen

Data Analysis:
Multivariate
Techniques for the
Research Process

Copyright 2006
McGraw-Hill/Irwin

Learning Objectives
1. Define multivariate analysis.
2. Understand how to use multivariate analysis in
marketing research.
3. Distinguish between dependence and interdependence
methods.
4. Define and understand factor analysis and cluster
analysis.
5. Define and use discriminant analysis.
McGraw-Hill/Irwin

Value of Multivariate
Techniques in Data
Analysis

Define multivariate analysis

Multivariate analysis--statistical techniques used

when there are two or more measurements of each
element and the variables are analyzed
simultaneously.
Multivariate techniques are concerned with the
simultaneous relationships among two or more
phenomena.
Important in marketing research because most
business problems are multidimensional

McGraw-Hill/Irwin

Exhibit 17.1

McGraw-Hill/Irwin

Define multivariate analysis

Exhibit 17.2

McGraw-Hill/Irwin

Define multivariate analysis

Classification Multivariate
Techniques

Dependence Method

Dependence Method multivariate technique

appropriate when one or more of the variables can
be identified as dependent variables and the
remaining as independent variables
Dependence techniquesmultiple regression
analysis, discriminant analysis, and MANOVA
Multiple discriminant analysisdependence
technique which predicts customer usage based on
several independent variables
Age, income, peer group, education, lifestyle.
McGraw-Hill/Irwin

Classification Multivariate
Techniques

Interdependence techniques

Interdependence techniques multivariate statistical

techniques in which a whole set of interdependent relationships is

examined
No single variable is defined as dependent or independent
Multivariate procedureanalysis of all variables in the data set
simultaneously
Goal of this methodto group things together
Simplify data
No one variable is predicted or explained by the others
Interdependence techniques factor analysis, cluster analysis,
Perceptual Mapping and multidimensional scaling
McGraw-Hill/Irwin

Classification Multivariate
Techniques

First Step

Nature of the Measurement Scales

Determine which multivariate technique is appropriate to
analyze the data
Dependence vs. Interdependence

Dependent variable
Measured nonmetrically(Nominal)Discriminant analysis,
Conjoint
Measured metrically (ratio or interval) multiple regression,
ANOVA, and MANOVA

McGraw-Hill/Irwin

Classification Multivariate
Techniques
Independent variable
Require metric independent variablemultiple
regression and discriminant analysiscan use
nonmetric dummy variables
Nonmetric independent variablesANOVA and
MANOVA
Metrically measured variables and nonmetric
adaptionsfactor analysis and cluster analysis
McGraw-Hill/Irwin

Interdependence
Techniques

Factor Analysis

Factor Analysisused to summarize information

contained in a large number of variables into a
smaller number of subsets or factors
Purpose of Factor Analysisto simplify the data
No distinction between dependent and
independent variables
all variables under investigation are analyzed
togetherto identify underlying factors
McGraw-Hill/Irwin

Exhibit 17.3

McGraw-Hill/Irwin

Factor Analysis

Interdependence
Techniques

Factor Analysis

Factor Loadingsimple correlation between the variables

Starting Pointinterpreting factor analysis is factor loadings
Factor loadingmeasure of the importance of the variable in
measuring each factor
Like correlationsvary from +1.0 to 1.0
Statistical analysis associated with factor analysis
produces factor loadings between each factor and each of
the original variables

McGraw-Hill/Irwin

Exhibit 17.4

McGraw-Hill/Irwin

Factor Analysis

Interdependence
Techniques

Factor Analysis

Next Step in Factor Analysis

name the resulting factors
Factor 1 Service Quality
Factor 2 Food Quality

Final Aspect of Factor Analysis

the number of factors to retain

McGraw-Hill/Irwin

Exhibit 17.5

McGraw-Hill/Irwin

Factor Analysis

Interdependence
Techniques

Define and understand factor

analysis and cluster analysis

Factor Analysis Applications in Marketing Research

Advertising
to better understand media habits of various customers

Pricing
to identify the characteristics of price-sensitive and prestigesensitive customers

Product
to identify brand attributes that influence consumer choice

Distribution
to better understand channel selection criteria among
distribution channel members

McGraw-Hill/Irwin

Exhibit 17.6

McGraw-Hill/Irwin

Define and understand factor

analysis and cluster analysis

Exhibit 17.7

McGraw-Hill/Irwin

Define and understand factor

analysis and cluster analysis

Exhibit 17.8

McGraw-Hill/Irwin

Define and understand factor

analysis and cluster analysis

Exhibit 17.9

McGraw-Hill/Irwin

Define and understand factor

analysis and cluster analysis

Exhibit 17.10

McGraw-Hill/Irwin

Define and understand factor

analysis and cluster analysis

Interdependence
Techniques

Cluster analysis

Cluster analysismultivariate interdependence technique whose

primary objective is to classify objects into relatively homogeneous
groups based on the set of variables considered

Basic Purpose
To classify or segment objects into groups so that objects within
each group are similar to one another on a variety of variables
To classify segments or objects such that there will be as much
similarity within segments and as much difference between
segments as possible
To identify natural groupings or segments among many variables,
without designating any of the variables as a dependent variable

McGraw-Hill/Irwin

Exhibit 17.11

McGraw-Hill/Irwin

Cluster analysis

Interdependence
Techniques

Cluster analysis

Statistical Procedures for Cluster Analysis

Degree of similarity between objects
determined through a distance measure
Distance between any pair of points is
positively related to how similar the
corresponding individuals are when the two
variables are considered together
McGraw-Hill/Irwin

Interdependence
Techniques

Cluster analysis Scatter Plots

Clustersdeveloped from scatter plots

This is a very complex, trial and error process
Requires the use of computer algorithms

McGraw-Hill/Irwin

Interdependence
Techniques

Cluster analysis

Applications in Marketing Research

New product researchto examine product offerings relative to
competition
Test marketingto group test cities into homogeneous clusters
for test marketing purposes
Buyer behaviorto identify similar groups of buyers who have
similar choice criteria
Market segmentationto develop distinct market segments on
the basis of geographic, demographic, psychographic, and
behavioral variables
McGraw-Hill/Irwin

Exhibit 17.12

McGraw-Hill/Irwin

Define and understand factor

analysis and cluster analysis

SPSSexercise
Use the Santa Fe database
Find different subgroups of customers
with different levels of commitment
Use Variables 22, 23,24
Anaylse-classify-hierarchical cluster
Select wards method
Save box select 2
This procedure takes time
McGraw-Hill/Irwin

Exhibit 17.13

McGraw-Hill/Irwin

Define and understand factor

analysis and cluster analysis

Exhibit 17.14

McGraw-Hill/Irwin

Define and understand factor

analysis and cluster analysis

Exhibit 17.15

McGraw-Hill/Irwin

Define and understand factor

analysis and cluster analysis

EndHere

McGraw-Hill/Irwin

Analysis of
Dependence

Define and use discriminant

analysis

Discriminant Analysismultivariate procedure

used for predicting group membership on the basis
of two or more independent variables
Purposeto classify objects or groups by a set of
independent variables
Dependent variablenonmetric or categorical
Independent variablesmetric
McGraw-Hill/Irwin

Exhibit 17.17

McGraw-Hill/Irwin

Define and understand factor

analysis and cluster analysis

Analysis of
Dependence

Define and use discriminant

analysis

Purpose of discriminant analysisprediction of a categorical

variable by studying the direction of group differences based on
finding a linear combination of independent variables
Discriminant functionlinear combination of independent
variables developed by discriminant analysis which will best
discriminate between the categories of the dependent variable

Discriminate analysisstatistical tool for determining linear

combinations of those independent variables and using this to
predict group membership

McGraw-Hill/Irwin

Analysis of
Dependence

Define and use discriminant

analysis

Discriminant score (Z-score)basis for predicting to which group

the particular individual belongs and is determined by a linear
function
Zi =
b1X1i + b2X2i + bnXni

Zi

ith individuals discriminant score

bn

Discriminant coefficient for the nth variable

Xni =

Individuals value on the nth independent variable

Discriminant scorethe score of each respondent on the

discriminant function
McGraw-Hill/Irwin

Analysis of
Dependence

Define and use discriminant

analysis

Discriminant function coefficients

estimates of the discriminatory power of a
particular independent variable
multipliers of variables in the discriminant
function when variables are in the original units
of measurement

Coefficientscomputed by means of the

discriminant analysis software
McGraw-Hill/Irwin

Analysis of
Dependence

Define and use discriminant

analysis

Important goal of discriminant analysis

classification of objects or individuals into groups
Classification (Prediction) Matrixto determine
whether the estimated discriminant function is a
good predictor
Classification (or prediction) matrixclassification
matrix in discriminant analysis contains the number of
correctly classified and misclassified cases
McGraw-Hill/Irwin

Exhibit 17.18

McGraw-Hill/Irwin

Define and understand factor

analysis and cluster analysis

Analysis of
Dependence

Define and use discriminant

analysis

Applications for Marketing Research

Product researchto distinguish between heavy, medium, and
light users of a product in terms of their consumption habits and
lifestyles
Image researchto discriminate between customers who exhibit
favorable perceptions of a store or company and those who do not
Advertising researchIn distinguishing how market segments
differ in media consumption habits
Direct marketingin distinguishing characteristics of consumers
who respond to direct marketing solicitations and those who dont

McGraw-Hill/Irwin

Exhibit 17.20

McGraw-Hill/Irwin

Define and understand factor

analysis and cluster analysis

Exhibit 17.21

McGraw-Hill/Irwin

Define and understand factor

analysis and cluster analysis

Exhibit 17.22

McGraw-Hill/Irwin

Define and understand factor

analysis and cluster analysis

Exhibit 17.23

McGraw-Hill/Irwin

Define and understand factor

analysis and cluster analysis

Exhibit 17.24

McGraw-Hill/Irwin

Define and understand factor

analysis and cluster analysis

Exhibit 17.25

McGraw-Hill/Irwin

Define and understand factor

analysis and cluster analysis

Summary
Value of Multivariate Techniques in Data
Analysis
Classification Multivariate Techniques
Interdependence Techniques
Analysis of Dependence

McGraw-Hill/Irwin

The End

Copyright 2006 McGraw-Hill/Irwin

McGraw-Hill/Irwin