## DOE:  Screening Experiments

### Course outline

#### Lesson 1

###### Why DOE?
• Limitations of OATs (one-at-a-time) experimentation.

• How designed experiments overcome the limitations of OATs and are a more effective and efficient way to characterize and improve processes and products.

#### Lesson 2

###### DOE Terminology
• An explanation of the key terms used in designed experiments.

#### Lesson 3

###### Types of Designed Experiments
• Full Factorials

• Fractional Factorials

• Screening Experiments

• Response Surface Analysis

• EVOP

• Mixture Experiments

#### Lesson 4

###### Tests of Significance
• Alpha and Beta Risks

• Degrees of Freedom

• Hypothesis Tests

• t-Tests

• F-Tests

#### Lesson 5

###### Setting Up a Designed Experiment
• Design & Communicate the Objective

• Define the Process

• Select a Response and Measurement System

• Select Factors to be Studied

• Select the Experimental Design

• Set Factor Levels

• Final Design Considerations

#### Unit Test

###### Challenge

An assessment of the learner’s progress in this unit.

#### Lesson 1

###### Plackett-Burman Matrices
• The derivation of Plackett–Burman designs.

• Types of Plackett–Burman matrices.

• Ways to determine the experimental error.

• Techniques for analyzing experimental results.

#### Lesson 2

###### Calculating Statistical Significance
• Multiple techniques for testing the statistical significance of factor effects.

• Using graphical techniques to analyze responses and interactions.

#### Lesson 3

###### Calculating a Prediction Equation
• Developing a prediction equation using factor effects.

• Using the prediction equation to optimize the process or product.

##### Lesson 4
###### Analyzing for Effect on Variation
• How to analyze variation as a response.

• Creating a scree diagram to graphically analyze factor effects on variation.

#### Lesson 5

###### When Bad Things Happen to Good Experiments
• The need for good planning to prevent problems.

• Some techniques for salvaging an experiment if data are lost or suspect.

#### Unit Test

###### Challenge
• An assessment of the learner’s progress in this unit.

#### Taguchi Concepts

• The concept of robustness.

• The Taguchi Loss Function.

• Signal to noise ratios.

#### Taguchi Matrices

• Taguchi designs for two-level experiments.

• Use of Taguchi Interaction Tables.

#### Taguchi Experimental Analysis

• Multiple techniques for testing the statistical significance of factor effects.

• Using graphical techniques to analyze responses and interactions.

#### Determining Where to Set Factors

• Developing a prediction equation.

• Use the mean, signal to noise ratio, and variation effects to determine where to set factors.

#### When Bad Things Happen to Good Experiments

• The need for good planning to prevent problems.

• Some techniques for salvaging an experiment if data are lost or suspect.

#### Challenge

An assessment of the learner’s progress in this unit.