Abstract:This technical note introduces business students to the concepts of modeling discrete choice (e.g., a consumer purchasing brand A versus brand B) using logistic regression and maximum-likelihood estimation. It draws the analogy between modeling discrete choice and building a regression model with a dummy dependent variable and on an example illustrates the need for estimating the probability of a choice rather than the choice itself, which leads to a special kind of regression – logistic regression. The note presents the concepts of utility and a random utility choice model, of which the logistic regression model is the most commonly used. It shows how choice probabilities can be constructed from utilities leading to the logit model. It then presents the maximum-likelihood estimation (MLE) method of fitting the logit model to the choice data. Working through a detailed example using Solver and accompanying spreadsheet model, the note gives students deep understanding for how MLE works and how it is similar and different to the standard least-squared estimation in linear regression. The note concludes by presenting the results of estimation using StatTools, a commercial statistical software. The note avoids the use of heavy mathematical machinery but still requires rudimentary knowledge of exponent and logarithmic functions, probability, and optimization with Solver, as well as familiarity with the “standard” linear regression. Applications include building of models for consumer choice, estimating price elasticity, price optimization, product versioning, product line design, and conjoint analysis. Excerpt UVA-QA-0779 Nov. 7, 2011 MODELING DISCRETE CHOICE: CATEGORICAL DEPENDENT VARIABLES, LOGISTIC REGRESSION, AND MAXIMUM LIKELIHOOD ESTIMATION Consider an individual choosing between two or more discrete alternatives: a shopper in a grocery store deciding between apple or orange juice, or a prospective student determining which of several university offers he ought to accept. For the juice manufacturer and the university, the ability to predict the outcome of such choices is of vital importance. In this note, we will discuss how this might be done. The process we will follow bears some similarity to a regular linear regression but also has substantial differences, primarily due to the fact that the choices are discrete; that is, they correspond to a categorical dependent variable in regression. The Concept of Utility . . .

A Constructive Procedure for Modeling Categorical Variables: Log-Linear and Logit Models

Learning Cluster Causal Diagrams: an Information-Theoretic Approach

An Information Analysis on Modeling Interaction Effects in Logistic Regression

D-optimal Designs for Multinomial Logistic Models

Generating contingency tables with fixed marginal probabilities and dependence structures described by loglinear models

Model Building Strategy for Logistic Regression: Purposeful Selection

Structured Mixture of Continuation-ratio Logits Models for Ordinal Regression

Loglinear modeling with mixed numerical and categorical predictor variables through an Extended Stereotype Model

Logistic and multinomial-logit models: A brief review on their modifications and extensions

On the correspondence of deviances and maximum likelihood and interval estimates from log-linear to logistic regression modelling

A binarization approach to model interactions between categorical predictors in Generalized Linear Models

A Tutorial on Testing, Visualizing, and Probing an Interaction Involving a Multicategorical Variable in Linear Regression Analysis

PATH ANALYSIS WITH LOGISTIC REGRESSION MODELS : EFFECT ANALYSIS OF FULLY RECURSIVE CAUSAL SYSTEMS OF CATEGORICAL VARIABLES

LOCALLY D-OPTIMAL DESIGNS FOR HIERARCHICAL RESPONSE EXPERIMENTS

Modeling Discrete Choice: Categorical Dependent Variables, Logistic Regression, and Maximum Likelihood Estimation

Learning Causal Structures Based on Divide and Conquer

Analysis of Ordinal Variables Using Rank-Based Polychoric Correlation

Log-Multiplicative Models for Discrete-Time, Discrete-Covariate Event-History Data

Analysis of Categorical Data with Logistic Regression and the Cochran-Mantel-Haenszel Tests in Biological Experiments

A Random-effects Approach to Regression Involving Many Categorical Predictors and Their Interactions

Discrete-Choice Logit Models