Design of Experiments (DOE) is a three-day course that begins with first principles and operational definitions. We use a parameter diagram to graphically describe the variables, for example, which knobs are controllable and which knobs we must treat as noise factors.
The quality statistics review considers the importance of process stability and process capability in interpreting results of experimentation. Understanding variation is one of four key concepts Dr. Deming codified into his system of profound knowledge, so we'll follow his leadership, specifically his focus on analytic statistical studies. Measurement systems analysis studies are used to assess the appropriateness of a measurement process prior to that process being used in experimentation. Then we consider Weibull analysis, to understand behavior of distributions, and to assess whether the resulting distribution comes from one or more sets of causal factors.
Knowledge gained will guide us to revise the problem statement, and derive hypotheses to test. We'll consider experimental strategies and designs. We'll study highlights from two classic texts, Statistics for Experimenters (Box, Hunter & Hunter, 1977) and Statistical Problem Solving (Bajaria & Copp, 1991). This class prepares you to experiment successfully.
- Operationally Definite Meaning
- Creating a P Diagram
- Quality Statistics Review
- Profound Knowledge of a System
- Measurement Uncertainty
- Weibull Analysis
- Hypothesis Testing
- Experimental Strategies
- Statistics for Experimenters
- Statistical Problem Solving
Key Course Objectives:
Explicitly Identify Input and Output VariablesComplete the Plan, Do, Check, Act CycleSelect an Experimental StrategyDevelop HypothesesExecute an ExperimentAnalyze Experimental ResultsIdentify Recommended Actions