Transcript Marketing Research

1
Marketing Research
Aaker, Kumar, Day and Leone
Tenth Edition
Instructor’s Presentation Slides
2
Chapter Twenty-two
Multidimensional Scaling and
Conjoint Analysis
http://www.drvkumar.com/mr10/
Marketing Research 10th Edition
3
Multidimensional Scaling
Used to:
• Identify dimensions by which objects are
perceived or evaluated
• Position the objects with respect to those
dimensions
• Make positioning decisions for new and old
products
http://www.drvkumar.com/mr10/
Marketing Research 10th Edition
4
Approaches To Creating Perceptual Maps
Perceptual map
Attribute data
Nonattribute data
Preference
Similarity
Factor
analysis
http://www.drvkumar.com/mr10/
Correspondence
analysis
Discriminant
analysis
MDS
Marketing Research 10th Edition
5
Attribute Based Approaches
• Attribute based MDS - MDS used on attribute data
• Assumption
▫ The attributes on which the individuals' perceptions of objects are based
can be identified
• Methods used to reduce the attributes to a small number
of dimensions
▫ Factor Analysis
▫ Discriminant Analysis
• Limitations
▫ Ignore the relative importance of particular attributes to customers
▫ Variables are assumed to be intervally scaled and continuous
http://www.drvkumar.com/mr10/
Marketing Research 10th Edition
6
Comparison of Factor and
Discriminant Analysis
Discriminant Analysis
Factor Analysis
• Identifies clusters of attributes
on which objects differ
• Groups attributes that are
similar
• Identifies a perceptual
dimension even if it is
represented by a single
attribute
• Based on both perceived
differences between objects and
differences between people's
perceptions of objects
• Statistical test with null
hypothesis that two objects are
perceived identically
• Dimensions provide more
interpretive value than
discriminant analysis
http://www.drvkumar.com/mr10/
Marketing Research 10th Edition
7
Perceptual Map of a Beverage Market
http://www.drvkumar.com/mr10/
Marketing Research 10th Edition
8
Perceptual Map of Pain Relievers
Gentleness
. Tylenol
. Bufferin
Effectiveness
. Bayer
. Private-label
aspirin
http://www.drvkumar.com/mr10/
. Anacin
. Advil
. Nuprin
. Excedrin
Marketing Research 10th Edition
9
Basic Concepts of Multidimensional Scaling (MDS)
• MDS uses proximities (value which denotes how similar or how different two
objects are perceived to be) among different objects as input
• Proximities data is used to produce a geometric configuration of points
(objects) in a two-dimensional space as output
• The fit between the derived distances and the two proximities in each
dimension is evaluated through a measure called stress
• The appropriate number of dimensions required to locate objects can be
obtained by plotting stress values against the number of dimensions
http://www.drvkumar.com/mr10/
Marketing Research 10th Edition
10
Determining Number of Dimensions
Due to large increase in the stress values from two dimensions to one,
two dimensions are acceptable
http://www.drvkumar.com/mr10/
Marketing Research 10th Edition
11
Attribute-based MDS
Advantages
Disadvantages
• Attributes can have diagnostic
and operational value
• If the list of attributes is not
accurate and complete, the
study will suffer
• Attribute data is easier for the
respondents to use
• Respondents may not perceive
or evaluate objects in terms of
underlying attributes
• Dimensions based on attribute
data predicted preference better
as compared to non-attribute
data
http://www.drvkumar.com/mr10/
• May require more dimensions
to represent them than the use
of flexible models
Marketing Research 10th Edition
12
Application of MDS With Nonattribute Data
Similarity Data
• Reflect the perceived similarity of two objects from the respondents'
perspective
• Perceptual map is obtained from the average similarity ratings
• Able to find the smallest number of dimensions for which there is a
reasonably good fit between the input similarity rankings and the rankings of
the distance between objects in the resulting space
http://www.drvkumar.com/mr10/
Marketing Research 10th Edition
13
Similarity Judgments
http://www.drvkumar.com/mr10/
Marketing Research 10th Edition
14
Perceptual Map Using Similarity Data
http://www.drvkumar.com/mr10/
Marketing Research 10th Edition
15
Application of MDS With Nonattribute Data (Contd.)
Preference Data
• An ideal object is the combination of all customers' preferred
attribute levels
• Location of ideal objects is to identify segments of customers who
have similar ideal objects, since customer preferences are always
heterogeneous
http://www.drvkumar.com/mr10/
Marketing Research 10th Edition
16
Issues in MDS
• Perceptual mapping has not been shown to be reliable
across different methods
• The effect of market events on perceptual maps cannot be
ascertained
• The interpretation of dimensions is difficult
• When more than two or three dimensions are needed,
usefulness is reduced
http://www.drvkumar.com/mr10/
Marketing Research 10th Edition
17
Conjoint Analysis
• Technique that allows a subset of the possible combinations of
product features to be used to determine the relative importance of
each feature in the purchase decision
• Used to determine the relative importance of various attributes to
respondents, based on their making trade-off judgments
• Uses:
▫ To select features on a new product/service
▫ Predict sales
▫ Understand relationships
http://www.drvkumar.com/mr10/
Marketing Research 10th Edition
18
Inputs in Conjoint Analysis
• The dependent variable is the preference judgment that a
respondent makes about a new concept
• The independent variables are the attribute levels that need
to be specified
• Respondents make judgments about the concept either by
considering
▫ Two attributes at a time - Trade-off approach
▫ Full profile of attributes - Full profile approach
http://www.drvkumar.com/mr10/
Marketing Research 10th Edition
19
Outputs in Conjoint Analysis
• A value of relative utility is assigned to each level of an
attribute called partworth utilities
• The combination with the highest utilities should be the
one that is most preferred
• The combination with the lowest total utility is the least
preferred
http://www.drvkumar.com/mr10/
Marketing Research 10th Edition
20
Applications of Conjoint Analysis
• Where the alternative products or services have a number of
attributes, each with two or more levels
• Where most of the feasible combinations of attribute levels do not
presently exist
• Where the range of possible attribute levels can be expanded beyond
those presently available
• Where the general direction of attribute preference probably is
known
http://www.drvkumar.com/mr10/
Marketing Research 10th Edition
21
Steps in Conjoint Analysis
1
2
3
• Choose product attributes (e.g. size, price, model)
• Choose the values or options for each attribute
• Define products as a combination of attribute options
4
• A value of relative utility is assigned to each level of an
attribute called partworth utilities
5
• The combination with the highest utilities should be
the one that is most preferred
http://www.drvkumar.com/mr10/
Marketing Research 10th Edition
22
Utilities for Credit Card Attributes
Source: Paul E. Green, ‘‘A New Approach to Market Segmentation,’’
http://www.drvkumar.com/mr10/
Marketing Research 10th Edition
23
Utilities for Credit Card Attributes (Contd.)
http://www.drvkumar.com/mr10/
Marketing Research 10th Edition
24
Full-profile and Trade-off Approaches
Source: Adapted from Dick Westwood, Tony Lunn, and David Bezaley, ‘‘The Trade-off Model and Its Extensions’’
http://www.drvkumar.com/mr10/
Marketing Research 10th Edition
25
Conjoint Analysis - Example
http://www.drvkumar.com/mr10/
Make
Price
MPG
Door
0
Domestic
$22,000
22
2-DR
1
Foreign
$18,000
28
4-DR
25
Marketing Research 10th Edition
26
Conjoint Analysis – Regression Output
Model Summaryc
Model
1
R
R Square
.785 b
.616
Adjus ted
R Square
.488
Std. Error of
the Es timate
6.921
b. Predictors : Door, MPG, Price, Make
c. Dependent Variable: Rank
ANOVAc
Model
1
Sum of
Squares
921.200
574.800
1496.000
Regress ion
Res idual
Total
df
4
12
16
Mean Square
230.300
47.900
F
4.808
Sig.
.015 a
a. Predictors : Door, MPG, Price, Make
c. Dependent Variable: Rank
Coefficientsa, b
Model
1
Make
Price
MPG
Door
Uns tandardized
Coefficients
B
Std. Error
1.200
3.095
4.200
3.095
5.200
3.095
2.700
3.095
Standardized
Coefficients
Beta
.088
.307
.380
.197
t
.388
1.357
1.680
.872
Sig.
.705
.200
.119
.400
a. Dependent Variable: Rank
b. Linear Regres sion through the Origin
http://www.drvkumar.com/mr10/
Marketing Research 10th Edition
27
Part-worth Utilities
1.4
4.5
1.2
4
3.5
3
Utility
Utility
1
0.8
0.6
2.5
2
1.5
0.4
1
0.5
0.2
0
0
Foreign
Domestic
18,000
Price
6
3
5
2.5
4
2
Utility
Utility
Make
3
1.5
2
1
1
0.5
0
0
28
22
4-Dr
MPG
http://www.drvkumar.com/mr10/
22,000
2-Dr
Door
27
Marketing Research 10th Edition
28
Relative Importance of Attributes
Attribute
Part-worth Utility
Relative
Importance
Make
1.2
9%
Price
4.2
32%
MPG
5.2
39%
Door
2.7
20%
http://www.drvkumar.com/mr10/
28
Marketing Research 10th Edition
29
Limitations of Conjoint Analysis
Trade-off approach
• The task is too unrealistic
• Trade-off judgments are being made on two attributes,
holding the others constant
Full-profile approach
• If there are multiple attributes and attribute levels, the
task can get very demanding
http://www.drvkumar.com/mr10/
Marketing Research 10th Edition