Design of experiments, piloting solutions, control plans, and statistical process control used in the Improve and Control phases.
29 cards · basic cards · AI-written, checked twice. Edit anything.
- What is the primary purpose of Design of Experiments?
- To systematically vary input factors and study their effects on an output variable to identify the vital few factors influencing the response.
- In DOE, what is a factor?
- An input variable that is deliberately changed or manipulated during an experiment to study its effect on the response.
- In DOE, what are levels?
- The different values or settings of a factor being tested in an experiment (e.g., temperature at 150°C, 200°C, and 250°C).
- What is a full factorial design?
- An experiment testing every possible combination of all factor levels (e.g., 2^3 = 8 runs for three factors at two levels each).
- What is a fractional factorial design?
- A DOE using a subset of all possible factor combinations to reduce experimental runs while still identifying main effects.
- What does resolution mean in fractional factorial designs?
- The ability to distinguish between main effects and interaction effects; higher resolution (III, IV, V) provides better separation of effects.
- Distinguish between a main effect and an interaction effect in DOE.
- A main effect is how one factor alone affects the response; an interaction occurs when two factors together produce an effect different from their individual effects.
- What is blocking in DOE?
- Grouping experimental runs to minimize the effect of known nuisance variables that cannot be directly controlled.
- What is confounding in DOE?
- When the effect of one factor becomes entangled with another factor, making it impossible to separate their individual contributions to the response.
- What is Response Surface Methodology?
- A DOE technique using polynomial regression to map the relationship between factors and response over a region of interest, often to optimize the response.
- What is the purpose of center points in fractional factorial or response surface designs?
- To detect curvature in the response surface and provide an estimate of pure error independent of the fitted model.
- Why is randomization important in the run order of a DOE?
- To prevent bias from time-related factors and ensure observed effects are due to the factors studied, not confounding variables that change over time.
- What is a pilot run in Six Sigma?
- A small-scale trial of a proposed solution to verify feasibility, refine implementation details, and gather data before full deployment.
- What must be validated before conducting a pilot run?
- The measurement system must be validated through a Gage R&R study to ensure it can accurately capture the effect of the proposed solution.
- What is the purpose of a run chart during the pilot phase?
- To track the response variable over time during the pilot to detect shifts or trends and confirm the solution produces the desired effect.
FlashKeepers