CSSC Black Belt key terms

A measurable characteristic of a product or service that directly reflects what the customer values most. CTQs translate vague customer desires into specific, testable requirements that improvement projects can target and verify.

The structured collection of customer needs, expectations, and complaints gathered through interviews, surveys, and complaint data. VOC input is analyzed and converted into CTQs and project charters so improvement work stays aligned with real demand.

A framework that classifies product features into must-be, performance, and delighter categories based on how their presence or absence affects satisfaction. It helps teams prioritize improvements that create the greatest customer impact.

A shorthand expression of the core improvement logic: the output Y depends on a set of input variables x. Practitioners identify which inputs drive variation in the output, then control or optimize those inputs.

A high-level process map showing Suppliers, Inputs, Process steps, Outputs, and Customers. It defines project scope, clarifies boundaries, and reveals which supplier and input variables may influence the output.

A visual representation of the sequence of steps, decisions, and handoffs in a process. Detailed maps expose rework loops, non-value-added steps, and variation sources that are invisible in written procedures.

A map that displays material and information flow along with timing data such as cycle time, wait time, and lead time for each step. It highlights where waste accumulates and where flow can be improved.

A step-by-step review that classifies each process activity as value-added, non-value-added but necessary, or pure waste. The classification guides which steps to eliminate, simplify, or redesign.

The eight categories of non-value-adding activity: defects, overproduction, waiting, non-utilized talent, transportation, inventory, motion, and extra processing. Naming waste types helps teams spot opportunities systematically during process observation.

The available production time divided by customer demand, expressed as the pace at which one unit must be completed to meet demand. It sets the rhythm that process cycle times are compared against.

The time required to complete one unit of work at a given process step, measured from start to finish of that step. Comparing cycle time across steps reveals bottlenecks and imbalance.

A relationship stating that average work-in-process equals throughput multiplied by average flow time. It lets practitioners predict how reducing WIP or improving throughput changes lead time.

A five-phase improvement roadmap: Define, Measure, Analyze, Improve, and Control. Each phase has defined deliverables and tollgate reviews that keep projects evidence-driven and prevent premature solution jumping.

A design methodology: Define, Measure, Analyze, Design, and Verify. It is used when a process or product does not yet exist or cannot be fixed by incremental improvement, focusing on designing capability in from the start.

A document that authorizes a project and states its problem, goal, scope, business case, team, milestones, and expected benefits. It aligns stakeholders and defines what is in and out of bounds.

The process of identifying everyone affected by or able to influence a project, then assessing their interest, influence, and likely support. It informs communication and buy-in strategies during Define and beyond.

The structured set of activities used to prepare, support, and sustain people through a process change. It addresses resistance, communication, training, and reinforcement so improvements survive after the project closes.

A set of studies that quantify how much variation in measured data comes from the measurement system itself. It checks resolution, bias, linearity, stability, repeatability, and reproducibility before data are trusted.

A study that separates measurement variation into repeatability, the same operator measuring the same item consistently, and reproducibility, variation between operators. Results are compared against tolerance or process variation to judge acceptability.

A study for pass/fail or categorical data that measures how often appraisers agree with themselves, with each other, and with a known standard. It is the categorical counterpart to a variables Gage R&R.

The complete collection of equipment, procedures, software, operators, and environment used to assign a number or category to a characteristic. Every element can contribute variation, so the whole system is evaluated, not just the instrument.

Indices that compare process spread and centering to specification limits. Cp relates tolerance width to six sigma of spread; Cpk accounts for how far the mean sits from the nearer specification limit.

Capability-like indices computed from the overall standard deviation of all data rather than short-term within-subgroup variation. They describe what the process actually delivered during the study period, including drift between subgroups.

A measure of process capability expressed as the number of standard deviations between the process mean and the nearest specification limit. Higher values mean fewer expected defects per million opportunities.

Defects per million opportunities, calculated as defects divided by total defect opportunities, multiplied by one million. It allows processes with different complexity to be compared on a common scale.

The probability that a unit passes every step of a multi-step process without any defect, found by multiplying the first-pass yields of each step. It exposes hidden factory losses masked by final inspection yields.

The total cost incurred from defects, including scrap, rework, warranty, inspection, and lost goodwill. Quantifying COPQ builds the business case and helps prioritize which problems are worth solving.

A time-ordered plot of a process statistic with statistically derived control limits. It distinguishes common-cause variation from special-cause signals, guiding whether to adjust the process or investigate a specific event.

The routine, inherent variation produced by the ordinary interaction of process inputs. It appears as a stable random pattern; reducing it requires fundamental process changes rather than reacting to individual points.

Variation arising from identifiable, non-routine events such as a bad batch, a new operator, or a machine fault. It produces signals on a control chart and should be investigated and eliminated at its source.

The practice of using control charts and capability studies to monitor and improve process stability over time. It shifts attention from inspecting output to controlling the process that creates it.

A document specifying which process characteristics are monitored, how they are measured, sampling frequency, reaction plans, and responsible owners. It sustains gains after the improvement phase ends.

Documented, agreed-upon methods for performing a task, capturing the current best sequence, timing, and quality checks. It provides a baseline for training, auditing, and further improvement.

A mistake-proofing device or design that prevents an error from occurring or makes it immediately obvious. Simple, low-cost mechanisms stop defects at the source rather than catching them later.

Failure Mode and Effects Analysis, a structured team method that lists potential failure modes, their effects and causes, and scores severity, occurrence, and detection. Risk priority numbers guide which failures to address first.

A planned approach to varying multiple input factors together so their individual and interaction effects on the response can be estimated efficiently. It replaces one-factor-at-a-time testing with statistically valid designs.

A statistical technique that models the relationship between one or more predictor variables and a continuous response. It quantifies effect sizes, tests significance, and supports prediction and input control.

A formal procedure that uses sample data to decide whether an observed difference or relationship is likely real or due to chance. It weighs evidence against a null hypothesis using a chosen significance level.

The probability of observing a result at least as extreme as the sample result if the null hypothesis were true. Small values suggest the data are inconsistent with the null, but they do not measure effect size.

Type I error rejects a true null hypothesis, a false alarm; Type II error fails to reject a false null hypothesis, a missed effect. Their probabilities relate to significance level, power, sample size, and effect size.

Calculating how many observations are needed to detect a meaningful effect with acceptable confidence and power. It prevents underpowered studies that miss real effects and oversized studies that waste resources.