CSSYB cheat sheet

Six Sigma targets 3.4 defects per million opportunities (DPMO), equivalent to a 6-sigma process with a 1.5-sigma long-term shift.

Sigma-level reference: 3σ ≈ 66,807 DPMO; 4σ ≈ 6,210 DPMO; 5σ ≈ 233 DPMO; 6σ ≈ 3.4 DPMO.

Y = f(X): the output Y is a function of process inputs X; controlling critical Xs improves Y.

DMAIC improves existing processes; DMADV (DFSS) designs new products or processes.

Lean attacks waste and lead time; Six Sigma reduces variation and defects.

Eight wastes (TIMWOODS): Transport, Inventory, Motion, Waiting, Overproduction, Overprocessing, Defects, Skills underutilized.

Belt roles: Champion/Sponsor, Master Black Belt, Black Belt, Green Belt, Yellow Belt, Process Owner.

Rolled Throughput Yield = product of individual step yields = e^(−DPU).

Cost of poor quality includes internal failure, external failure, and appraisal costs; prevention cost is not part of poor quality.

Project charter elements: problem statement, business case, goal statement, scope, team, milestones, metrics, and sponsor approval.

SMART goal: Specific, Measurable, Achievable, Relevant, Time-bound; it includes baseline, target, and deadline.

Problem statement quantifies the pain (what, where, when, magnitude) and avoids stating a cause or solution.

SIPOC stands for Suppliers, Inputs, Process, Outputs, Customers; it defines start and stop boundaries.

VOC → CTQ → operational definition → metric; CTQs are measurable customer requirements.

Pareto principle: roughly 80% of effects come from 20% of causes; use it to prioritize.

Scope control uses an in-scope versus out-of-scope list; unclear scope leads to scope creep.

Stakeholder power/interest grid: high power plus high interest means manage closely.

DPMO = (defects ÷ (units × opportunities)) × 1,000,000.

DPU = defects ÷ units; RTY = e^(−DPU).

Cp = (USL − LSL) ÷ (6σ); Cpk = min[(USL − μ), (μ − LSL)] ÷ (3σ).

Cpk ≥ 1.33 is generally acceptable; Cpk = 1.00 means the process just fits the specification width when centered.

Gage R&R: <10% acceptable, 10–30% conditionally acceptable, >30% unacceptable.

Attribute agreement Kappa: ≥0.70 acceptable, ≥0.90 excellent.

Measurement scales: nominal, ordinal, interval, ratio; ratio has a true zero.

Sampling methods include simple random, stratified, systematic, and cluster; random selection reduces bias.

Data collection plan defines what, how, who, when, where, and operational definitions.

Hypothesis test decision rule: p-value < α → reject H0; α is commonly 0.05.

Type I error = α (false positive); Type II error = β (false negative); power = 1 − β.

t-test compares one or two means; ANOVA compares three or more means.

Chi-square test evaluates independence or goodness-of-fit for categorical data.

Correlation r ranges from −1 to +1; r near 0 indicates no linear relationship; correlation does not prove causation.

Scatter plot displays the relationship between two continuous variables.

5 Whys drills down to root cause; fishbone (Ishikawa) organizes potential causes into categories.

FMEA RPN = Severity × Occurrence × Detection; higher RPN gets priority for action.

Regression R² is the proportion of variation in Y explained by the model.

Solution selection criteria include impact, effort, cost, risk, cycle time, and feasibility.

Poka-yoke prevents or detects errors at the source; prevention is preferred over detection.

5S: Sort, Set in order, Shine, Standardize, Sustain.

Kaizen is continuous incremental improvement involving the people who do the work.

SMED reduces changeover time, ideally to single-digit minutes.

DOE varies factors at set levels to identify main effects and interactions.

PDCA/PDSA: Plan, Do, Check/Study, Act; used to pilot and refine changes.

Update FMEA after improvements; a lower RPN confirms risk reduction.

Control plan: characteristic, measurement method, frequency, owner, control limits, and reaction plan.

Control limits = process mean ± 3σ; specification limits come from customer requirements.

Common cause variation is inherent and stable; special cause variation is assignable and signals a process change.

X̄-R chart for continuous data in subgroups; p chart for proportion defective; c chart for defect counts.

Run rules flag nonrandom patterns such as a point beyond a control limit, consecutive points on one side, or a sustained trend.

OCAP defines containment, investigation, and corrective action for out-of-control events.

Standard work and SOP updates sustain gains; training and audits verify adherence.

Project closure includes handoff to the process owner, lessons learned, and financial validation.

Visual management makes abnormal conditions visible at a glance.