APMG LSS Green Belt key terms

The structured collection of customer needs, expectations and complaints, gathered through interviews, surveys and observation, that defines what a process must deliver. It anchors project goals in real requirements rather than internal assumptions.

A measurable characteristic that translates a customer need into a specific, verifiable requirement, such as delivery within three days or defect rate below one percent. CTQs turn vague expectations into targets a project can test.

A high-level map showing Suppliers, Inputs, Process steps, Outputs and Customers for a process. It scopes a project by clarifying where the process begins and ends and who is affected.

A formal document authorising a improvement project, stating the problem, goal, scope, business case, team and timeline. It secures sponsor commitment and prevents scope drift.

A concise description of the gap between current and desired performance, including where, when and how much, without stating a cause or a solution. A well-written one keeps the team focused on facts.

The process of identifying everyone affected by or able to influence a project, then assessing their interest and influence to plan communication and engagement. Neglected stakeholders are a common cause of project failure.

A table assigning each project task or decision to those Responsible, Accountable, Consulted and Informed. It removes ambiguity about ownership and prevents decisions stalling.

A comparison of current performance against a required or benchmark level, quantifying the difference to be closed. It justifies the project and later shows whether improvement has been achieved.

The money lost through defects and failures, including rework, scrap, warranty claims, inspection and lost goodwill. Quantifying it builds the business case for improvement work.

The core Lean Six Sigma model expressing that an output Y is determined by input and process variables x. Improvement work identifies and controls the few x variables that most affect Y.

A precise description of how a characteristic is measured, including the instrument, method and decision criteria, so that different people measure it the same way. It underpins reliable data collection.

A document specifying what data to collect, why, from where, by whom, how often and with what method. It prevents wasted effort and ensures data answer the project question.

A set of studies, such as gauge repeatability and reproducibility, that quantify how much variation comes from the measurement system rather than the process. Unacceptable systems must be fixed before trusting data.

Repeatability is variation when one operator measures the same item repeatedly; reproducibility is variation between different operators measuring the same items. Together they reveal the main sources of measurement error.

A systematic offset between the average measured value and the true value, caused by instrument calibration, worn tooling or operator habit. Unlike random error it does not average out with more readings.

Accuracy is how close a measurement is to the true value; precision is how closely repeated measurements agree with each other. A system can be precise yet inaccurate if it is consistently offset.

Selecting a subset of units or time periods for measurement instead of every one. Good sampling plans are random or systematic enough to represent the population and large enough to support the analysis.

Splitting data into subgroups such as shift, machine, operator or product type so patterns hidden in combined figures become visible. It often reveals that averages mask very different subgroup behaviour.

The recorded performance of a process before improvement, used as the reference for judging whether changes worked. It must be measured with the same definitions used later.

A diagram of the sequence of steps, decisions and handoffs in a process, often built with the people who do the work. It exposes complexity, loops and non-value-adding steps.

A map of material and information flow through a process, with cycle times, wait times and inventory at each step. It distinguishes value-adding activity from waste and highlights delay.

The classic set of non-value-adding losses: defects, overproduction, waiting, non-utilised talent, transportation, inventory, motion and extra processing. Naming them helps teams spot improvement opportunities.

Examining each process step to decide whether the customer would pay for it, whether it physically changes the product or service, and whether it is done right first time. Steps failing all three are candidates for removal.

A systematic investigation to find the underlying reasons a problem occurs, rather than treating symptoms. Tools include the five whys, cause-and-effect diagrams and fault tree analysis.

A cause-and-effect chart that branches possible causes into categories such as people, machine, method, material, measurement and environment. It organises brainstorming before data confirm which causes matter.

Repeatedly asking why a problem occurred, typically about five times, to move from a symptom to a fundamental cause. It works best when each answer is verified with evidence.

A statistical procedure that judges whether an observed difference or relationship is likely real or explainable by chance, using a p-value compared with a chosen significance level. It supports or refutes suspected causes.

The probability of observing data at least as extreme as obtained if the null hypothesis were true. A small p-value suggests the observed effect is unlikely to be pure chance.

A Type I error rejects a true null hypothesis, a false alarm; a Type II error fails to detect a real effect, a missed opportunity. Their risks trade off against sample size and significance level.

Correlation measures the strength of linear association between two variables; regression models how a response changes with one or more predictors. Correlation alone does not establish causation.

A bar chart of causes ordered by frequency or cost, with a cumulative line, illustrating that a few causes account for most of the effect. It directs effort to the largest contributors.

A structured team review that lists ways a process or product can fail, rates severity, occurrence and detection, and prioritises actions by risk priority number. It prevents problems before they reach customers.

A mistake-proofing device or procedure that makes an error impossible or immediately obvious, such as a fixture that only accepts the correct part. It prevents defects rather than detecting them later.

Continuous, often small-scale improvement driven by the people doing the work, frequently through structured team events. It builds sustained capability rather than relying on occasional large projects.

A repeating cycle of Plan, Do, Check and Act used to test changes on a small scale, study the results and then standardise or adjust. It makes improvement systematic and evidence-based.

A trial of a proposed change on a limited scale or in one area before full implementation. It reveals practical problems and provides evidence that the change delivers the expected result.

Documented, agreed and repeatable methods for performing a task, including sequence, timing and key points. It provides the baseline from which further improvement is measured.

A document listing the process parameters and product characteristics to monitor, the measurement method, sampling frequency, limits and the response when limits are breached. It keeps gains in place after the project ends.

A time-ordered plot of a process measure with statistically derived limits that separate normal variation from signals of special causes. It tells operators when to act and when to leave the process alone.

Common cause variation is the inherent, stable noise of a process; special cause variation arises from identifiable, unusual events. Only special causes justify local intervention, while common causes require process redesign.

A measure such as Cp or Cpk of how well a stable process meets specification limits relative to its variation. It indicates whether the process can consistently produce within tolerance.

The time taken to complete one unit of work from start to finish of a process step or the whole process. Reducing it shortens lead time and often exposes hidden waste.

A relationship stating that average lead time equals work in progress divided by throughput. It shows that reducing work in progress shortens delivery time without changing capacity.

The unreported rework, expediting and correction work that keeps output flowing but never appears in standard metrics. It consumes capacity and inflates true cost of poor quality.