JMP 11 Design of Experiments Guide

JMP 11 Design of Experiments Guide

by SAS Institute
Released September 2013
Publisher(s): SAS Institute
ISBN: 9781612906874

Read it now on the O’Reilly learning platform with a 10-day free trial.

O’Reilly members get unlimited access to books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.

Start your free trial

Book description

The JMP 11 Design of Experiments Guide covers classic DOE designs (for example, full factorial, response surface, and mixture designs). Read about more flexible custom designs, which you generate to fit your particular experimental situation. And discover JMP's definitive screening designs, an efficient way to identify important factor interactions using fewer runs than required by traditional designs. The book also provides guidance on determining an appropriate sample size for your study.

Table of contents

  1. Contents
  2. Learn about JMP
    1. Documentation and Additional Resources
    2. Formatting Conventions
    3. JMP Documentation
      1. JMP Documentation Library
      2. JMP Help
    4. Additional Resources for Learning JMP
      1. Tutorials
      2. Sample Data Tables
      3. Learn about Statistical and JSL Terms
      4. Learn JMP Tips and Tricks
      5. Tooltips
      6. JMP User Community
      7. JMPer Cable
      8. JMP Books by Users
      9. The JMP Starter Window
  3. Introduction to Designing Experiments
    1. A Beginner’s Tutorial
    2. About Designing Experiments
    3. My First Experiment
      1. The Situation
      2. Step 1: Design the Experiment
        1. Define the Responses: Popped Kernels and Total Kernels
        2. Define the Factors: Time, Power, and Brand
      3. Step 2: Define Factor Constraints
      4. Step 3: Add Interaction Terms
      5. Step 4: Determine the Number of Runs
      6. Step 5: Check the Design
      7. Step 6: Gather and Enter the Data
      8. Step 7: Analyze the Results
  4. Building Custom Designs
    1. The Basic Steps
    2. Creating a Custom Design
      1. Enter Responses and Factors into the Custom Designer
        1. How to Enter Responses
        2. Specifying Response Goal Types and Lower and Upper Limits
        3. Understanding Response Importance Weights
        4. Adding Simulated Responses, If Desired
        5. How to Enter Factors
        6. Factors that are Easy, Hard, or Very Hard, to Change: Creating Optimal Split-Plot and Split-Split-Plot Designs
        7. Defining Factor Constraints, If Necessary
      2. Describe the Model
        1. Adding Terms to the Model
        2. Estimability: Necessary and If Possible Terms
      3. Specifying Alias Terms
      4. Select the Number of Runs
      5. The Design Report
      6. Understanding Design Evaluation
        1. Power Analysis
        2. The Prediction Variance Profile
        3. The Fraction of Design Space Plot
        4. The Prediction Variance Surface
        5. Estimation Efficiency
        6. The Alias Matrix (Confounding Pattern)
        7. Color Map on Correlations
        8. The Design Diagnostics Table
      7. Specify Output Options
      8. Make the JMP Design Table
    3. Creating Random Block Designs
    4. Creating Split Plot Designs
    5. Creating Split-Split Plot Designs
    6. Creating Strip Plot Designs
    7. Special Custom Design Commands
      1. Save Responses and Save Factors
      2. Load Responses and Load Factors
      3. Save Constraints and Load Constraints
      4. Set Random Seed: Setting the Number Generator
      5. Simulate Responses
      6. Save X Matrix
      7. Optimality Criterion
      8. Number of Starts: Changing the Number of Random Starts
        1. Why Change the Number of Starts?
        2. Default Choice of Number of Random Starts: Technical Information
      9. Sphere Radius: Constraining a Design to a Hypersphere
      10. Disallowed Combinations: Accounting for Factor Level Restrictions
      11. Advanced Options for the Custom Designer
        1. Altering the Mixture Sum
        2. Split Plot Variance Ratio
        3. Prior Parameter Variance
        4. D Efficiency Weight
        5. Set Delta for Power
      12. Save Script to Script Window
    8. Assigning Column Properties
      1. Define Low and High Values (DOE Coding) for Columns
      2. Set Columns as Factors for Mixture Experiments
      3. Define Response Column Values
      4. Assign Columns a Design Role
      5. Identify Factor Changes Column Property
    9. How Custom Designs Work: Behind the Scenes
  5. Examples Using the Custom Designer
    1. Creating Screening Experiments
      1. Creating a Main-Effects-Only Screening Design
      2. Creating a Screening Design to Fit All Two-Factor Interactions
      3. A Compromise Design Between Main Effects Only and All Interactions
      4. Creating “Super” Screening Designs
        1. The Need for Supersaturated Designs
        2. Example: Twelve Factors in Eight Runs
      5. Screening Designs with Flexible Block Sizes
      6. Checking for Curvature Using One Extra Run
    2. Creating Response Surface Experiments
      1. Exploring the Prediction Variance Surface
      2. Introducing I-Optimal Designs for Response Surface Modeling
      3. A Three-Factor Response Surface Design
      4. Response Surface with a Blocking Factor
    3. Creating Mixture Experiments
      1. Mixtures Having Nonmixture Factors
      2. Experiments that are Mixtures of Mixtures
    4. Special Purpose Uses of the Custom Designer
      1. Designing Experiments with Fixed Covariate Factors
      2. Creating a Design that is Robust to a Linear Time Trend in the Response
      3. Creating a Design with Two Hard-to-Change Factors: Split Plot
    5. Technical Discussion
  6. Definitive Screening Designs
    1. Definitive Screening Platform Overview
    2. Example of Definitive Screening
    3. Example of Definitive Screening versus Plackett-Burman Designs
  7. Screening Designs
    1. Overview
    2. Screening Design Examples
      1. A Standard Design with Two Continuous Factors and One Categorical Factor
      2. A Standard Design for Five Continuous Factors
    3. Creating a Screening Design
      1. Enter Responses
        1. Specifying Goal Types and Lower and Upper Limits
        2. Understanding Importance Weights
      2. Enter Factors
        1. Types of Factors
      3. Choose a Design
        1. Design List
      4. Display and Modify Design
        1. Aliasing of Effects
        2. Look at the Confounding Pattern
        3. Understanding Design Codes
        4. Changing the Coded Design
      5. Design Generation
      6. Design Evaluation
      7. Specify Output Options
      8. View the Design Table
    4. Creating a Plackett-Burman design
    5. Creating a Main Effects Screening Design
    6. Analyzing Screening Data
      1. Comparing Screening and Fit Model
      2. Launch the Screening Platform
      3. The Screening Report
        1. Contrasts
        2. Half Normal Plot
      4. Using the Fit Model Platform
        1. The Actual-by-Predicted Plot
        2. The Scaled Estimates Report
        3. A Power Analysis
    7. Additional Screening Analysis Examples
      1. Analyzing a Plackett-Burman Design
      2. Analyzing a Supersaturated Design
    8. Statistical Details
      1. Operation
      2. Screening as an Orthogonal Rotation
      3. Lenth’s Pseudo-Standard Error
  8. Response Surface Designs
    1. A Box-Behnken Design: The Tennis Ball Example
      1. The Prediction Profiler
      2. A Response Surface Plot (Contour Profiler)
      3. Geometry of a Box-Behnken Design
    2. Creating a Response Surface Design
      1. Enter Responses and Factors
      2. Choose a Design
        1. Box-Behnken Designs
        2. Central Composite Designs
        3. Specify Axial Value (Central Composite Designs Only)
      3. Specify Output Options
      4. View the Design Table
  9. Full Factorial Designs
    1. The Five-Factor Reactor Example
      1. Analyze the Reactor Data
    2. Creating a Factorial Design
      1. Enter Responses and Factors
      2. Select Output Options
      3. Make the Table
  10. Mixture Designs
    1. Mixture Design Types
    2. The Optimal Mixture Design
      1. Adding Effects to the Model
    3. The Simplex Centroid Design
      1. Creating the Design
      2. Simplex Centroid Design Examples
    4. The Simplex Lattice Design
    5. The Extreme Vertices Design
      1. Creating the Design
      2. An Extreme Vertices Example with Range Constraints
      3. An Extreme Vertices Example with Linear Constraints
      4. Extreme Vertices Method: How It Works
    6. The ABCD Design
    7. The Space Filling Design
      1. A Space Filling Example with a Linear Constraint
    8. Creating Ternary Plots
    9. Fitting Mixture Designs
      1. Whole Model Tests and Analysis of Variance Reports
      2. Understanding Response Surface Reports
    10. A Chemical Mixture Example
      1. Create the Design
      2. Analyze the Mixture Model
      3. The Prediction Profiler
      4. The Mixture Profiler
      5. A Ternary Plot of the Mixture Response Surface
  11. Discrete Choice Designs
    1. Introduction
    2. Create an Example Choice Experiment
    3. Analyze the Example Choice Experiment
    4. Design a Choice Experiment Using Prior Information
    5. Administer the Survey and Analyze Results
      1. Initial Choice Platform Analysis
      2. Find Unit Cost and Trade Off Costs with the Profiler
  12. Space-Filling Designs
    1. Introduction to Space-Filling Designs
    2. Sphere-Packing Designs
      1. Creating a Sphere-Packing Design
      2. Visualizing the Sphere-Packing Design
    3. Latin Hypercube Designs
      1. Creating a Latin Hypercube Design
      2. Visualizing the Latin Hypercube Design
    4. Uniform Designs
    5. Comparing Sphere-Packing, Latin Hypercube and Uniform Methods
    6. Minimum Potential Designs
    7. Maximum Entropy Designs
    8. Gaussian Process IMSE Optimal Designs
    9. Fast Flexible Filling Designs
      1. Creating a Constrained Fast Flexible Filling Design
      2. Visualizing the Fast Flexible Filling Design
    10. Borehole Model: A Sphere-Packing Example
      1. Create the Sphere-Packing Design for the Borehole Data
      2. Guidelines for the Analysis of Deterministic Data
      3. Results of the Borehole Experiment
  13. Accelerated Life Test Designs
    1. Designing Experiments for Accelerated Life Tests
    2. Overview of Accelerated Life Test Designs
    3. Using the ALT Design Platform
    4. Platform Options
    5. Example
  14. Nonlinear Designs
    1. Examples of Nonlinear Designs
      1. Using Nonlinear Fit to Find Prior Parameter Estimates
      2. Creating a Nonlinear Design with No Prior Data
    2. Creating a Nonlinear Design
      1. Identify the Response and Factor Column with Formula
      2. Set Up Factors and Parameters in the Nonlinear Design Dialog
      3. Enter the Number of Runs and Preview the Design
      4. Make Table or Augment the Table
    3. Advanced Options for the Nonlinear Designer
  15. Taguchi Designs
    1. The Taguchi Design Approach
    2. Taguchi Design Example
      1. Analyze the Data
    3. Creating a Taguchi Design
      1. Detail the Response and Add Factors
      2. Choose Inner and Outer Array Designs
      3. Display Coded Design
      4. Make the Design Table
  16. Evaluating Experimental Designs
    1. Using the Evaluate Design Platform
    2. Launching the Evaluate Design Platform
    3. The Evaluate Design Report
      1. Design Evaluation
        1. Power Analysis
        2. Prediction Variance Profile
        3. Fraction of Design Space Plot
        4. Prediction Variance Surface
        5. Estimation Efficiency
        6. Alias Matrix
        7. Color Map on Correlations
        8. Design Diagnostics
    4. Examples
      1. Assessing the Impact of Lost Runs
      2. Assessing the Impact of Changing the Model
  17. Augmented Designs
    1. A D-Optimal Augmentation of the Reactor Example
      1. Analyze the Augmented Design
    2. Creating an Augmented Design
      1. Replicate a Design
      2. Add Center Points
      3. Creating a Foldover Design
      4. Adding Axial Points
      5. Adding New Runs and Terms
    3. Special Augment Design Commands
      1. Save the Design (X) Matrix
      2. Modify the Design Criterion (D- or I- Optimality)
      3. Select the Number of Random Starts
      4. Specify the Sphere Radius Value
      5. Disallow Factor Combinations
  18. Prospective Sample Size and Power
    1. Launching the Sample Size and Power Platform
    2. One-Sample and Two-Sample Means
      1. Single-Sample Mean
        1. Power versus Sample Size Plot
        2. Power versus Difference Plot
      2. Sample Size and Power Animation for One Mean
      3. Two-Sample Means
        1. Plot of Power by Sample Size
    3. k-Sample Means
    4. One Sample Standard Deviation
      1. One Sample Standard Deviation Example
    5. One-Sample and Two-Sample Proportions
      1. One Sample Proportion
        1. One-Sample Proportion Window Specifications
      2. Two Sample Proportions
        1. Two Sample Proportion Window Specifications
    6. Counts per Unit
      1. Counts per Unit Example
    7. Sigma Quality Level
      1. Sigma Quality Level Example
      2. Number of Defects Computation Example
    8. Reliability Test Plan and Demonstration
      1. Reliability Test Plan
        1. Example
      2. Reliability Demonstration
        1. Example
  19. References
  20. Index
    1. Design of Experiments Guide
    2. A
    3. B
    4. C
    5. D
    6. E
    7. F
    8. G
    9. H
    10. I
    11. J
    12. K
    13. L
    14. M
    15. N
    16. O
    17. P
    18. Q
    19. R
    20. S
    21. T
    22. U
    23. V
    24. W-Z

Product information

  • Title: JMP 11 Design of Experiments Guide
  • Author(s): SAS Institute
  • Release date: September 2013
  • Publisher(s): SAS Institute
  • ISBN: 9781612906874