7/28/2019 Analysis Factor Analysis Cluster Analysis
1/18
Discriminant Analysis,Factor Analysis, Cluster
Analysis, Multidimensional
scaling and Conjoint
analysis
7/28/2019 Analysis Factor Analysis Cluster Analysis
2/18
Discriminant Analysis
A technique for analyzing
marketing research data when
the criterion/dependent variable
is categorical and the
predictor/independent variable
are interval in nature
7/28/2019 Analysis Factor Analysis Cluster Analysis
3/18
Objectives of discriminant
analysis
1. Development of discriminant functions(linear combinations of the predictorwhich will best discriminate between thecategories of the criterion)
2. Examination whether significantdifferences exist among the group
3. Determination of which predictorvariables contribute to most of theintergroup differences
4. Classification of cases to one of thegroups based on the values of thepredictor variables
5. Evaluation of the accuracy ofclassification
7/28/2019 Analysis Factor Analysis Cluster Analysis
4/18
Similarities and Differences
among ANOVA, Regression and
Discriminant Analysis
Similarities:
Number of dependent variables -
1
Number of independent variables -
Multiple
Differences
Nature of dependent variable:Metric, Metric, Categorical
Nature of independent variable:
categorical, metric, metric
7/28/2019 Analysis Factor Analysis Cluster Analysis
5/18
Steps involved in Discriminant
Analysis
1. Formulate the problem
2. Estimate the discriminate
function coefficient
3. Determine the significance of
the discrimination function
4. Interpret the results5. Assess the validity of
discriminant analysis
7/28/2019 Analysis Factor Analysis Cluster Analysis
6/18
Marketing Research Applications
In term of demographic characteristics, howdo customers who exhibit store loyalty differfrom those who do not
Do heavy, medium and light users of softdrink differ in terms of their consumption of
frozen food What psychographic characteristics help
differentiate between price-sensitive andnon-price sensitive buyers of groceries
Does the various market segment differ in
their media consumption habits In terms of lifestyles, what are the differences
between heavy patrons of general stores andMall stores
7/28/2019 Analysis Factor Analysis Cluster Analysis
7/18
Factor analysis
A class of procedures primarily used for datareduction and summarization
It is use in the following circumstances
1. To identify underlying dimensions/factors
that explain correlations among a set ofvariables
2. To identify a new, smaller set ofuncorrelated variables to replace theoriginal set of correlated variables insubsequent multivariate analysis
3. To identify a smaller set of salientvariables from a larger set for use insubsequent multivariate analysis
7/28/2019 Analysis Factor Analysis Cluster Analysis
8/18
Steps involved in factor
analysis
1. Formulate the problem
2. Construct the correlation matrix
3. Determine the method of factor
analysis
4. Determine the number of factors
5. Rotate the factors
6. Interpret the factors1. Calculate the factor scores
2. Select the surrogate variables
7.
Determine the model fit
7/28/2019 Analysis Factor Analysis Cluster Analysis
9/18
7/28/2019 Analysis Factor Analysis Cluster Analysis
10/18
Cluster Analysis
A class of techniques used to classifyobjects/cases into relatively homogeneousgroups called clusters.
Objects in each cluster tend to be similar to
each other and dissimilar to other cluster A.k.a. classification analysis/numerical
taxonomy
Both cluster analysis and discriminantanalysis is used for classification but incluster analysis there is no a prioriinformation about the group/clustermembership
7/28/2019 Analysis Factor Analysis Cluster Analysis
11/18
Steps involved in cluster
analysis
Formulate the problem
Select a distance measure
Select a clustering procedure Decide on the number of cluster
Interpret and profile cluster
Assess the validity of clustering
7/28/2019 Analysis Factor Analysis Cluster Analysis
12/18
Marketing Research
Application
Segmenting the market
Understanding buyer behaviour
Identifying new productopportunities
Selecting test markets
Reducing data
7/28/2019 Analysis Factor Analysis Cluster Analysis
13/18
Multidimensional scaling
A class of procedures for representingperceptions and preferences ofrespondents spatially by means of avisual display.
MDS has been used in marketing toidentify:
1. The number and nature of dimensionsconsumer use to perceive differentbrands in the market place
2. The positioning of current brands onthese dimensions
3. The positioning of consumers idealbrand on these dimensions
7/28/2019 Analysis Factor Analysis Cluster Analysis
14/18
Steps in MDS
Formulate the problem
Obtain input data
Select an MDS procedure Decide on the number of
dimensions
Label the dimensions andinterpret the configurations
Assess reliability and validity
7/28/2019 Analysis Factor Analysis Cluster Analysis
15/18
Marketing Research
Application
1. Image measurement
2. Market segmentation
3. New product development
4. Assessing advertisingeffectiveness
5. Pricing analysis
6. Channel decisions
7. Attitude scale construction
7/28/2019 Analysis Factor Analysis Cluster Analysis
16/18
Conjoint Analysis
A technique that attempts todetermine the relative importanceconsumers attach to salientattributes and the utilities they attachto the levels of attributes
It seeks to develop the part-worth/utility functions describing theutility consumers attach to the levelsof each attribute.
It is complementary to MDS
7/28/2019 Analysis Factor Analysis Cluster Analysis
17/18
Steps involved in Conjoint
Analysis
Formulate the problem
Construct the stimuli
Decide on the form of input data Select a conjoint analysis
procedure
Interpret the resultsAssess reliability and validity
7/28/2019 Analysis Factor Analysis Cluster Analysis
18/18
Market Research Application
Determining the relative importanceof attributes in the consumer choiceprocess
Estimate market share of brands thatdiffer in attribute levels
Determining the composition of mostpreferred brand
Segmenting the market based onsimilarity of preferences for attributelevels