- Cover Page
- Title Page
- Copyright Page
- Contents
- Learn About JMP
- Launch the Fit Model Platform
- Example of the Fit Model Platform
- Specify Different Model Types
- Construct Model Effects
- Fitting Personalities
- Emphasis Options for Standard Least Squares
- Model Specification Options
- Validity Checks
- Fitting Standard Least Squares Models
- Example Using Standard Least Squares
- The Standard Least Squares Report and Options
- Regression Reports
- Estimates
- Effect Screening
- Factor Profiling
- Row Diagnostics
- Save Columns
- Effect Options
- Restricted Maximum Likelihood (REML) Method
- Method of Moments Results
- Singularity Details
- Examples with Statistical Details
- Fitting Stepwise Regression Models
- Overview of Stepwise Regression
- Example Using Stepwise Regression
- The Stepwise Report
- Models with Crossed, Interaction, or Polynomial Terms
- Models with Nominal and Ordinal Terms
- Using the Make Model Command for Hierarchical Terms
- Performing Logistic Stepwise Regression
- The All Possible Models Option
- The Model Averaging Option
- Using Validation
- Fitting Multiple Response Models
- Fitting Generalized Linear Models
- Performing Logistic Regression on Nominal and Ordinal Responses
- Analyzing Screening Designs
- Performing Nonlinear Regression
- Creating Neural Networks
- Modeling Relationships With Gaussian Processes
- Fitting Dispersion Effects with the Loglinear Variance Model
- Recursively Partitioning Data
- Performing Time Series Analysis
- Performing Categorical Response Analysis
- Performing Choice Modeling
- Introduction to Choice Modeling
- Product Design Engineering
- Data for the Choice Platform
- Example: Pizza Choice
- Launch the Choice Platform and Select Data Sets
- Platform Options
- Example: Valuing Trade-offs
- One-Table Analysis
- Segmentation
- Special Data Rules
- Transforming Data
- Logistic Regression for Matched Case-Control Studies
- Correlations and Multivariate Techniques
- Clustering Data
- Analyzing Principal Components and Reducing Dimensionality
- Performing Discriminant Analysis
- Fitting Partial Least Squares Models
- Scoring Tests Using Item Response Theory
- Plotting Surfaces
- Visualizing, Optimizing, and Simulating Response Surfaces
- Comparing Model Performance
- References
- The Response Models
- The Factor Models
- The Usual Assumptions
- Key Statistical Concepts
- Leverage Plot Details
- Multivariate Details
- Power Calculations
- Inverse Prediction with Confidence Limits
- Details of Random Effects
- Index
Chapter 16
Performing Choice Modeling
Using the Choice Platform
The Choice platform is designed for use in market research experiments, where the ultimate goal is to discover the preference structure of consumers. Then, this information is used to design products or services that have the attributes most desired by consumers.
Features provided in the Choice platform include:
• ability to use information about consumer traits as well as product attributes.
• integration of data from one, two, or three sources.
• ability to use the integrated profiler to understand, visualize, and optimize the response (utility) surface.
• provides subject-level scores for segmenting or clustering your data.
• uses a special default bias-corrected maximum likelihood estimator described by Firth (1993). This method has been shown to produce better estimates and tests than MLEs without bias correction. In addition, bias-corrected MLEs ameliorate separation problems that tend to occur in logistic-type models. Refer to Heinze and Schemper (2002) for a discussion of the separation problem in logistic regression.
The Choice platform is not appropriate to use for fitting models that involve:
• ranking or scoring.
• nested hierarchical choices. (PROC MDC in SAS/ETS can be used for such analysis.)
-
No Comment
..................Content has been hidden....................
You can't read the all page of ebook, please click here login for view all page.