scientific article; zbMATH DE number 1734453

From MaRDI portal
Publication:4328554





Stated Preferences (SP) are preferences elicited from the choices made in a controlled experiment. The book presents arguments why economists and other social scientists should be interested in SP data, which involve choice responses from economic agents evoked in hypothetical/virtual markets. It shows how SP methods can be implemented, and suggests how to combine SP and Revealed Preference (RP) data to get the best from each type. The authors define a major objective in writing the book as: to bring together, in one volume, tools developed over thirty years that allow one to elicit and model consumer preferences, estimate discrete-choice models of various degrees of complexity, apply the models to predict choices. The book consists of 13 chapters, which may be naturally divided into four parts: 1) methodological background of choice modeling and SP methods (Chapters 2-7); 2) combining sources of preference data and implementing SP choice behavior projects (Chapters 8-9); 3) applications (Chapters 10-12); 4) external validity of SP methods (chapter 13). Part one introduces alternative types of data available for studying choices. Random utility theory is used for explaining choice behavior. The behavioral framework is developed to model SP and RP. A family of discrete-choice models is derived from the principle of random utility maximization. The multinomial logit model is developed as a workhorse of choice models. An analytical toolkit to design experiments is presented in some detail, e.g. factorial design, fractional factorials, design coding, main effect plans, orthogonality. Many practical suggestions are based on the accumulated experience of the authors. For the cases, where assumptions on independent and identically distributed errors are too restrictive, a set of models is proposed including nested logit, multinomial probit, mixed logit, random parameter logit models. Design issues are considered for multiple choice problems involving violations of assumption on independent identically distributed errors. Summarizing, part one provides the reader with a large number of tools to design a choice experiment and estimate a discrete-choice model. Part 2 integrates data generation processes and brings together the techniques developed in part 1. The implementation of SP choice projects is structured in the following way: 1) choice problem definition, 2) sample frame definition and sample size, 3) task and survey definition, 4) data collection method, 5) data analysis and modeling, and 6) model use. Many particularities are discussed to safeguard an SP practitioner against real biases that can affect an SP choice study. Part 3 represents applications of SP choice methods. Depending upon the readers interests one of the three chapters should be of greater relevance: marketing, transportation and environmental applications. The final Chapter 13 considers how elicitation of SP and RP can be compared. It is shown through a number of examples that SP and RP preferences seem to match well in different choice contexts, cultures and time periods.




Cited in
(73)


Describes a project that uses

Uses Software






This page was built for publication:

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4328554)