Design of Experiments Guide > Choice Designs
Publication date: 07/08/2024

Choice Designs

Create a Design for Selecting Preferred Product Profiles

A choice (or discrete choice) experiment provides data for modeling discrete preferences. Study participants are presented with sets of potential products (or product profiles) with varying attributes. From each set of profiles, a participant selects a preferred profile. For example, in designing a high-end laptop, a computer company might be interested in the relative importance of key features such as processor speed, hard disk size, screen size, battery life, and price. A choice experiment addresses the relative values of these features to a customer and indicates an optimal set of trade-offs among product features.

The results of a choice experiment are analyzed using conjoint analysis methods. See Choice Models in Consumer Research.

Figure 18.1 A Survey with Eight Choice SetsĀ 

A Survey with Eight Choice Sets

Contents

Overview of Choice Designs

Choice Design Terminology
Bayesian D-Optimality

Example of a Choice Design

Example of a Choice Design with Analysis

Create a Choice Design for a Pilot Study
Analyze the Pilot Study Data
Design the Final Choice Experiment Using Prior Information
Run the Choice Design and Analyze the Results

Build a Choice Design

Attributes
Model
Design Generation
Design
Make Table

Choice Design Options

Statistical Details for Choice Designs

Bayesian D-Optimality and Design Construction
Utility-Neutral and Local D-Optimal Designs
Want more information? Have questions? Get answers in the JMP User Community (community.jmp.com).