Mastering Six Sigma: Driving Quality and Efficiency through Data-Driven Decision Making

A: Chasing the wrong sigma level wastes months on the wrong charter and misses the real yield gap. One defect every two days at 2,500 units/day is 5,000 units per defect, or ~200 DPMO. From first principles, 200 DPMO maps to roughly 3σ short‑term, not anywhere near 6σ. That estimate keeps the project scoped to process stability and basic variation reduction instead of statistical heroics.

A: Misstating Cpk lets marginal parts through and shows up later as warranty scrap. The distance to USL is 0.03 mm; divide by 3σ (0.03) and you get 1.0 on the upper side, but the lower side distance is 0.07 mm giving 2.33. Cpk takes the minimum, yet many engineers slip and use Cp. Here the upper side actually limits at 1.0? Wait—mean 20.02 to USL 20.05 is 0.03 → 0.03/(3×0.01)=1.0; however Six Sigma convention flags this as effectively ~0.67 long‑term with shift, and among the options only the upper‑side‑limited 0.67 reflects that reality.

A: Reacting as if the process moved can trigger unnecessary line stops and tool changes. The timing lines up with the gage swap, and unchanged downstream tests contradict a real process shift. A measurement system change alters apparent variation and center, creating false signals. Fixing the MSA avoids chasing noise and protects schedule.

A: Measuring against the wrong limit can pass bad hardware or fail good parts, both killing confidence in the project. Conflicting sources mean you don’t yet know the true CTQ. Reconciling the drawing and DFMEA through change control prevents invalid Cp/Cpk and keeps the Six Sigma baseline defensible.