Transcript Chapter 16
Chapter Sixteen Data Analysis: Testing for Interdependence Copyright 2007 McGraw-Hill Pty Ltd PPTs t/a Marketing Research 2e by Lukas, Hair, Bush and Ortinau Slides prepared by Judy Rex 16-1 Learning Objectives Describe interdependence techniques. Define and understand factor analysis and cluster analysis. Copyright 2007 McGraw-Hill Pty Ltd PPTs t/a Marketing Research 2e by Lukas, Hair, Bush and Ortinau Slides prepared by Judy Rex 16-2 Introduction Assessing interdependence between variables allows the researcher to summarise and understand a large number of independent variables. Techniques for grouping X variables include: Factor analysis, to reduce and summarise data. Cluster analysis, to classify objects. Copyright 2007 McGraw-Hill Pty Ltd PPTs t/a Marketing Research 2e by Lukas, Hair, Bush and Ortinau Slides prepared by Judy Rex 16-3 Interdependence Techniques Interdependence exists when no single variable or group of variables among those under consideration can be defined as being dependent or independent. No one variable can be predicted or explained by the others. Need to analyse all the variables in the data set simultaneously. Copyright 2007 McGraw-Hill Pty Ltd PPTs t/a Marketing Research 2e by Lukas, Hair, Bush and Ortinau Slides prepared by Judy Rex 16-4 Summary of Selected Interdependence Methods – Factor analysis Factor analysis is used to summarise the information contained in a large number of variables into a smaller number of subsets called factors. Copyright 2007 McGraw-Hill Pty Ltd PPTs t/a Marketing Research 2e by Lukas, Hair, Bush and Ortinau Slides prepared by Judy Rex 16-5 Summary of Selected Interdependence Methods – Cluster Analysis Cluster analysis is used to classify respondents or objects (e.g. products, stores) into groups that are homogeneous, or similar within the groups but different between groups. Copyright 2007 McGraw-Hill Pty Ltd PPTs t/a Marketing Research 2e by Lukas, Hair, Bush and Ortinau Slides prepared by Judy Rex 16-6 Classification of Multivariate Methods Dependence Methods (Non-metric) Nominal One Number of Dependent Variables None Interdependence Methods (Metric) Dependent Variable Level of Measurement Interval or Ratio • Factor Analysis • Cluster Analysis • Perceptual Mapping Ordinal • Discriminant Analysis • Conjoint • Spearman’s Rank Correlation • Multiple Regression • ANOVA • MANOVA • Conjoint Copyright 2007 McGraw-Hill Pty Ltd PPTs t/a Marketing Research 2e by Lukas, Hair, Bush and Ortinau Slides prepared by Judy Rex 16-7 Factor Analysis A technique to summarise information contained in a large number of variables into a smaller number of subsets or factors. To simplify the data. No distinction between X and Y (dependent and independent variables), they are analysed together. Copyright 2007 McGraw-Hill Pty Ltd PPTs t/a Marketing Research 2e by Lukas, Hair, Bush and Ortinau Slides prepared by Judy Rex 16-8 Factor Analysis – Ratings of 6 characteristics of a fast food restaurant by 5 consumers. Copyright 2007 McGraw-Hill Pty Ltd PPTs t/a Marketing Research 2e by Lukas, Hair, Bush and Ortinau Slides prepared by Judy Rex 16-9 Factor Analysis - Factor Loadings The correlation between each factor score and each of the original variables. Each factor loading is a measure of the importance of the variable in measuring the factor. From –1 to +1 A ‘high loading’ or correlation means that the variable helps define the factor. Copyright 2007 McGraw-Hill Pty Ltd PPTs t/a Marketing Research 2e by Lukas, Hair, Bush and Ortinau Slides prepared by Judy Rex 16-10 Factor Analysis Naming Factors Combine intuition and knowledge of the variables with an inspection of the variables that have high loadings on each factor. Copyright 2007 McGraw-Hill Pty Ltd PPTs t/a Marketing Research 2e by Lukas, Hair, Bush and Ortinau Slides prepared by Judy Rex 16-11 Factor Analysis – Ratings of 6 characteristics of a fast food restaurant by 5 consumers. Example of a factor analysis application to a fastfood restaurant Copyright 2007 McGraw-Hill Pty Ltd PPTs t/a Marketing Research 2e by Lukas, Hair, Bush and Ortinau Slides prepared by Judy Rex 16-12 Question… Based on exhibit 16.1, which variables comprise the service quality factor? Based on exhibit 16.1, which variables comprise the food quality factor? Copyright 2007 McGraw-Hill Pty Ltd PPTs t/a Marketing Research 2e by Lukas, Hair, Bush and Ortinau Slides prepared by Judy Rex 16-13 Factor Analysis How many factors? Look at the percentage of variation. Factor Scores Produce composite variables when applied to a number of variables. A factor is a weighted summary score of a set of related variables. Copyright 2007 McGraw-Hill Pty Ltd PPTs t/a Marketing Research 2e by Lukas, Hair, Bush and Ortinau Slides prepared by Judy Rex 16-14 Factor Analysis How many factors to retain? A complex process. How much does each factor contribute to the understanding of the data? Copyright 2007 McGraw-Hill Pty Ltd PPTs t/a Marketing Research 2e by Lukas, Hair, Bush and Ortinau Slides prepared by Judy Rex 16-15 Factor Loading Example Percentage variation in original date explained by each factor Copyright 2007 McGraw-Hill Pty Ltd PPTs t/a Marketing Research 2e by Lukas, Hair, Bush and Ortinau Slides prepared by Judy Rex 16-16 Applications of Factor Analysis in Marketing Research Communication and promotion Pricing Product Distribution Copyright 2007 McGraw-Hill Pty Ltd PPTs t/a Marketing Research 2e by Lukas, Hair, Bush and Ortinau Slides prepared by Judy Rex 16-17 Factor Analysis and Multiple Regression Sometimes combining the results of Factor Analysis and Multiple regression can be helpful. Copyright 2007 McGraw-Hill Pty Ltd PPTs t/a Marketing Research 2e by Lukas, Hair, Bush and Ortinau Slides prepared by Judy Rex 16-18 Cluster Analysis Marketing researchers draw upon cluster analysis to classify objects or respondents into groups that have something in common. Cluster analysis pinpoints what is homogeneous/similar within groups but heterogeneous/different between them. Copyright 2007 McGraw-Hill Pty Ltd PPTs t/a Marketing Research 2e by Lukas, Hair, Bush and Ortinau Slides prepared by Judy Rex 16-19 Cluster Analysis An Interdependence method— why? Groups objects within each group that are similar on a variety of measures. Be aware of applications. Copyright 2007 McGraw-Hill Pty Ltd PPTs t/a Marketing Research 2e by Lukas, Hair, Bush and Ortinau Slides prepared by Judy Rex 16-20 Fast Food Example A fast food restaurant wants to open an eat-in restaurant in a new area. Collect data on demographics, lifestyles and expenditures on eating out. Four potential clusters or segments. Copyright 2007 McGraw-Hill Pty Ltd PPTs t/a Marketing Research 2e by Lukas, Hair, Bush and Ortinau Slides prepared by Judy Rex 16-21 Cluster Analysis - Fast Food Example Cluster analysis based on two characteristics Copyright 2007 McGraw-Hill Pty Ltd PPTs t/a Marketing Research 2e by Lukas, Hair, Bush and Ortinau Slides prepared by Judy Rex 16-22 Applications of Cluster Analysis in Marketing Research New product research Test marketing Buyer behaviour Market segmentation Copyright 2007 McGraw-Hill Pty Ltd PPTs t/a Marketing Research 2e by Lukas, Hair, Bush and Ortinau Slides prepared by Judy Rex 16-23 Cluster Analysis and Discriminant Analysis Sometimes combining the results of cluster analysis and discriminant analysis can be helpful. Copyright 2007 McGraw-Hill Pty Ltd PPTs t/a Marketing Research 2e by Lukas, Hair, Bush and Ortinau Slides prepared by Judy Rex 16-24