This technical note presents the reason for using a binomial logic regression in marketing applications. It is used in Darden's "Big Data in Marketing" course elective. The issues surrounding the use of a linear regression model when the dependent variable is a dummy variable are identified. A consumer-utility-based behavioral rationale is presented for the applicability of the binomial logistic regression for modeling dummy variables. Simulated and real data examples are used to present the mechanics of the logistic regression and the interpretation of the outputs. The relationship between odds ratio and the logistic regression probabilities are presented. Application areas such as brand choice and customer retention are discussed.
Understand the intuition for a logistic regression, learn to interpret the outputs of a logistic regression, and identify marketing analytics scenarios that can benefit from a logistic regression.
