ABSTRACT
Chapter 6 addresses discriminant analysis and logistic regression, suitable techniques to measure the effects of independent metric variables on a categorical dependent variable divided into two or more groups. For instance, imagine a service provider with high rates of customer defection (high churn rate). Using these techniques, it is possible to identify which variables are causing exit or permanence of customers, such as working hours, personnel, service quality, price, etc. Initially, it presents how to perform a discriminant analysis and interpret its coefficients, explanatory power, and classification matrix. Next comes logistic regression, with the same steps. The techniques are illustrated with theoretical description, followed by an example with the SPSS commands and the results tables with comments. The chapter also includes exercises, such as a road map to perform the analysis, an interpretative exercise with results tables, and a market context to guide a research design.
