Key Lean Six Sigma Black Belt terms in plain English. It covers the shared DMAIC vocabulary and the advanced statistics that set the Black Belt apart - hypothesis testing, multiple regression and design of experiments. Understanding what each tool does, which DMAIC phase it belongs to, and which data situation it fits is exactly what the exam tests.
| Term | Definition |
|---|---|
| Six Sigma | A data-driven methodology for reducing defects and variation in a process to improve quality. |
| Lean | An approach focused on maximising value and eliminating waste in a process. |
| DMAIC | The core improvement method: Define, Measure, Analyze, Improve, Control. |
| Define | The first DMAIC phase: framing the problem, scope, customer needs and project charter; at Black Belt level, scoping and leading larger projects. |
| Measure | The DMAIC phase that maps the process and quantifies current performance with data. |
| Analyze | The DMAIC phase that identifies root causes; at Black Belt level it relies heavily on inferential statistics and hypothesis testing. |
| Improve | The DMAIC phase that develops, pilots and implements solutions; at Black Belt level it adds multiple regression and design of experiments. |
| Control | The DMAIC phase that sustains improvements with control plans and monitoring. |
| Project charter | A document that defines a project’s problem, scope, goals, team and timeline. |
| Voice of the Customer (VoC) | The expressed needs and expectations of the customer that drive requirements. |
| Critical to Quality (CTQ) | The specific, measurable customer requirements a process or product must meet. |
| SIPOC | A high-level process map of Suppliers, Inputs, Process, Outputs and Customers. |
| Defect | Any output that fails to meet a customer requirement (a CTQ). |
| DPMO | Defects Per Million Opportunities: a standardised measure of defect rate. |
| Process capability (Cp, Cpk) | Indices that compare how well a process meets its specification limits. |
| Measurement System Analysis (MSA) | A check that the way data is measured is accurate and consistent. |
| Variation | The natural (common-cause) or special-cause spread in process outputs that Six Sigma works to reduce. |
| Descriptive statistics | Summaries of data such as the mean, median and standard deviation that describe a sample. |
| Inferential statistics | Methods that draw conclusions about a wider population from a sample, central to the Black Belt Analyze phase. |
| Hypothesis test | A statistical test used to decide whether a difference or effect is real or due to chance. |
| Null hypothesis | The default assumption of no difference or no effect that a test tries to disprove. |
| Alternative hypothesis | The claim that there is a real difference or effect, accepted only if the data are strong enough. |
| P-value | The probability of seeing the data (or more extreme) if the null hypothesis were true; small values suggest a real effect. |
| Alpha (significance level) | The threshold (often 0.05) below which a p-value is treated as statistically significant. |
| Type I error | Rejecting a true null hypothesis - a ‘false alarm’. |
| Type II error | Failing to reject a false null hypothesis - a ‘missed effect’. |
| Confidence interval | A range that is likely to contain the true value, expressing the uncertainty in an estimate. |
| Normal distribution | A symmetric, bell-shaped distribution that many statistical tests assume. |
| Non-normal data | Data that does not follow a normal distribution, requiring different tests or transformations - a Black Belt concern. |
| ANOVA | Analysis of Variance: a test for whether the means of three or more groups differ. |
| Correlation | A measure of how strongly two variables move together (not proof of causation). |
| Regression | A statistical method that models the relationship between variables. |
| Multiple regression | Regression that models an output from several input variables at once. |
| Design of Experiments (DOE) | A structured way to test several input factors at once to see how they affect the output. |
| Factor | An input variable that is deliberately changed in a design of experiments. |
| Fishbone diagram | A cause-and-effect (Ishikawa) diagram for brainstorming root causes. |
| Pareto chart | A bar chart ordering causes by frequency to find the vital few. |
| FMEA | Failure Mode and Effects Analysis: a method to anticipate and prioritise risks. |
| Poka-yoke | Mistake-proofing: designing a process so errors are hard or impossible to make. |
| Statistical Process Control (SPC) | Using control charts to monitor a process and detect unusual variation over time. |
| Control plan | A document setting out how the improved process will be monitored and maintained. |
| Kaizen | A philosophy and practice of continuous, incremental improvement. |
| Black Belt | A practitioner who leads larger improvement projects full-time, applies advanced statistics, and coaches Green Belts. |