Business Knowledge Series course

Presented by Mike Patetta, Analytical Training Consultant, Education Division, SAS

This marketing research course shows how to design a discrete choice experiment and how to analyze discrete choice data in SAS software. Analytical advice regarding number of choice sets, the number of alternatives, and number of subjects is also given.

This course includes practice data and exercises.

Learn how to

• design a discrete choice study using the market research macros in SAS software
• analyze discrete choice data using the PHREG procedure
• create interaction plots to illustrate the model results.

Who should attend

Market researchers who want to perform discrete choice analysis using SAS software

Formats available	Duration
e-Learning:	3.5 hours/180 day license

Before attending this course, you should

• be familiar with DATA step programming and basic SAS procedures for producing summary statistics and graphs, such as the MEANS and GPLOT procedures. You can gain this knowledge by completing the SAS编程 1：要点 course.
• understand statistical concepts such as significance probabilities, multiple linear regression, multiple logistic regression, correlation, n-way analysis of variance, and tests of association. You can gain this knowledge by completing the 统计学 1: 方差分析回归与逻辑回归简介 course.

This course addresses SAS/STAT software.

Introduction to a Discrete Choice Analysis

• explain the concept of discrete choice analysis
• show the difference between discrete choice analysis and full-profile conjoint analysis
• describe the model used in discrete choice analysis

Designing the Study

• explain the importance of experimental design for discrete choice analysis
• illustrate the relationship between the linear design and the choice design
• use the market research macros to create a choice design

Performing the Analysis

• illustrate how to fit a discrete choice model in PROC PHREG
• show the functionality of the HAZARDRATIO statement in PROC PHREG
• compute predicted probabilities for the hypothetical products
• create interaction plots