Design of Experiments (DOE) Training
Design of Experiments (DOE) is one of the most useful applications of statistical methodology. Some of the advantages of good experimental design include:
- superior product designs,
- maximized yields,
- minimized time from product conception to delivery,
- efficient problem solving,
- optimized processes,
- minimized variation, and
- detection of causes of product failure.
Designed experiments are best employed by individuals who have a desire to seek the truth...and to seek the truth in the most efficient manner. DOE is a tool to understand processes quickly and thoroughly.
This 4-day course introduces the most common experimental designs and those designs that are implemented when knowledge of the process is not refined. The purpose is to make many of the introductory DOE concepts accessible. This course will also teach the most common pitfalls to avoid, so that the experimental efforts are not wasted.
Finally, this course emphasizes that DOE is sequential in nature. It is a learning process. We begin with an idea or hypothesis, and as we perform an experiment to test that hypothesis, we generally develop several new thoughts and suspicions, and those can be followed up with other experiments.
Not all experiments are created equal. This course teaches participants how to plan and conduct statistically designed experiments in an effective and efficient manner. Participants gain the fundamental knowledge necessary to design effective experiments, analyze the results, and avoid the common mis-applications in practice. Knowledge of basic algebra is helpful.
Introduction to Experimental Design
- Sequential Experimentation
- Avoiding Common Pitfalls
A Guide to Experimentation
Two Level Factorial Designs
- Design Matrix
- Calculation of Effects
- Geometric Analogy
Identifying Significant Effects
- Variable & Attribute Responses
- Describing Insignificant Location Effects
- Analyzing Replicated and Unreplicated Designs
- Normal Probability Plotting
Developing Mathematical Models
- Developing First Order Models
- Model Validation
- Process Optimization
Fractional Factorial Designs
- Structure of the Designs
- Identifying an "Optimal" Fraction
- Analysis of Fractional Factorials
- Discussion on Reflected Designs
Proportion & Variance Responses
- Identifying Significant Proportion Effects
- Identifying Significant Variance Effects
Exercises and Group Project
- Exercises for each Topic
- Plan, Design, Run, and Analyze an Experiment