GLOBAL MARKETING RESEARCH V Kumar 1 Dr V
GLOBAL MARKETING RESEARCH V. Kumar 1 © Dr V. Kumar (www. drvkumar. com/gmr) Global Marketing Research
CHAPTER 15 Multivariate Data Analysis 2 © Dr V. Kumar (www. drvkumar. com/gmr) Global Marketing Research
Chapter Overview An In-depth look at Multivariate Data Techniques: q Interdependence Techniques q Dependence Techniques 3 © Dr V. Kumar (www. drvkumar. com/gmr) Global Marketing Research
Interdependence Techniques 4 © Dr V. Kumar (www. drvkumar. com/gmr) Global Marketing Research
Interdependence Techniques Most frequently used q Factor Analysis q Cluster Analysis q Multidimensional Scaling 5 © Dr V. Kumar (www. drvkumar. com/gmr) Global Marketing Research
Factor Analysis Definition: Technique in which researchers look for a small number of factors that could explain the correlation between a large number of variables Factor Analysis is used for: q Data Reduction and Transformation q Evaluating personality scales q Identifying key product attributes 6 © Dr V. Kumar (www. drvkumar. com/gmr) Global Marketing Research
Factor Analysis Factor Eigenvalue • Variable or a construct that is not directly observable • Must be inferred from the input variables • The amount of variance in the original variables associated that is associated with the factor 7 © Dr V. Kumar (www. drvkumar. com/gmr) Global Marketing Research
Factor Analysis Scree Plot Criteria Percentage of Variance Criteria • Scree: Plot of eigenvalues against the number of factors • For factors with large eigenvalues this plot has a steep slope. • The number of factors extracted is determined so that the cumulative percentage of variance extracted by the variance reaches a satisfactory level. 8 © Dr V. Kumar (www. drvkumar. com/gmr) Global Marketing Research
Factor Analysis Factor Interpretation Factor Score Communality • The input to a factor analysis program is a set of variables for each individual or object in the sample. • The value of each factor for all respondents • The percentage of a variable’s variance that contributes to the correlation with other variables or is “common” to other variables 9 © Dr V. Kumar (www. drvkumar. com/gmr) Global Marketing Research
Factor Analysis Variance Explained Factor Rotation • Summary measure indicating how much of the total original variance of all five variables the factor represents • One of the several solutions (loadings and factor scores) generated by Factor Analysis for any data set. 10 © Dr V. Kumar (www. drvkumar. com/gmr) Global Marketing Research
Disadvantages of Factor Analysis q Highly subjective method q Does not make any use of standardized statistical tests 11 © Dr V. Kumar (www. drvkumar. com/gmr) Global Marketing Research
Cluster Analysis Definition: Variables are placed in subgroups or clusters. q Group objects into clusters based on the attributes they possess. q Objects that are similar are placed in one group q Groups have minimum withingroup variability and maximum between-group variability 12 © Dr V. Kumar (www. drvkumar. com/gmr) Global Marketing Research
Cluster Analysis The six stages of Cluster Analysis include: Problem Definiti on Measur es of Similarit y Decide how to group Objects Decide # of clusters 13 © Dr V. Kumar (www. drvkumar. com/gmr) Global Marketing Research Evaluate and Profile Clusters Validate Clusters
Multidimensional Scaling q Definition: Encompasses a set of computational procedures that can summarize an input matrix of associations between variables or objects. Used by marketing researchers to analyze: § Consumer perceptions and preferences of various brands § Identify attributes important to consumers 14 © Dr V. Kumar (www. drvkumar. com/gmr) Global Marketing Research
Dependence Techniques 15 © Dr V. Kumar (www. drvkumar. com/gmr) Global Marketing Research
Dependence Techniques The dependence techniques most frequently used in global marketing are: Discriminant Analysis • Used for predictions and descriptions • Used to classify objects into two or more alternative groups on the basis of a set of measurements Conjoint Analysis • Used to provide a quantitative measure of the relative importance of one attribute as opposed to another 16 © Dr V. Kumar (www. drvkumar. com/gmr) Global Marketing Research
- Determine linear combinations of predictor variables that would form groups - Develop procedures that help in placing new objects in one of the previously formed groups. - Test if there are significant differences between groups - Determine variables that contribute the most to the group variations. Conjoint Analysis Discriminant Analysis: Characteristics of Dependence Techniques 17 © Dr V. Kumar (www. drvkumar. com/gmr) Global Marketing Research - Helpful in predicting the buying or usage of a new product that is still in concept form. - The greater the difference between the highest- and the lowest-valued levels of the attribute, the more important is the attribute
End of Chapter 15 18 © Dr V. Kumar (www. drvkumar. com/gmr) Global Marketing Research
- Slides: 18