CSSGB key terms

A disciplined, data-driven approach to reducing variation and defects in processes, aiming for no more than 3.4 defects per million opportunities while improving customer satisfaction and business results.

The five-phase improvement cycle - Define, Measure, Analyze, Improve, Control - used to fix an existing process that is not meeting customer requirements or performance targets.

The money an organization loses because work is not done right the first time, including scrap, rework, warranty, inspection, and complaint handling; often hidden and far larger than expected.

A set of principles and tools aimed at removing waste - anything that does not add value for the customer - while shortening lead time and improving the flow of work.

The unreported extra work created by defects, such as rework, retesting, expediting, and extra inspections, which consumes capacity and cost without appearing in standard process metrics.

A shorthand model stating that the output result (Y) is determined by a set of input process variables (x); improvement comes from finding and controlling the vital few x's.

A formal document that authorizes a Six Sigma project by stating the problem, goal, scope, business case, team, schedule, and expected benefits, and that must be approved by the project sponsor.

A high-level map showing the Suppliers, Inputs, Process steps, Outputs, and Customers of a process, used to agree on scope and boundaries before detailed analysis begins.

The collected needs, expectations, and complaints of customers, gathered through interviews, surveys, and complaint data, that define what the process must deliver.

A measurable characteristic of a product or service that matters most to customers and must be met; CTQs translate vague customer wants into specific, testable requirements.

A way of classifying customer requirements into basic (must-have), performance (more is better), and delighter (unexpected) categories so improvement effort is spent where it pays off.

Identifying everyone who affects or is affected by a project, then assessing their influence and interest so a communication and buy-in plan can be built.

A precise, agreed description of what a characteristic is and exactly how it is measured, so that different people measuring the same item get consistent results.

A study of the measuring system itself - equipment, method, appraisers, and environment - to quantify how much observed variation comes from measurement error rather than from the process.

A study that splits measurement variation into repeatability (same appraiser, same part) and reproducibility (different appraisers), showing whether the gauge can be trusted for decisions.

A number comparing how well a stable process fits inside its specification limits, accounting for both spread and centering; 1.33 is a common minimum for a capable process.

Defects per million opportunities - total defects divided by total defect opportunities, multiplied by one million - used to compare processes of different complexity on a common scale.

The probability that a unit passes every step of a process without a defect, found by multiplying the first-pass yields of all steps; it exposes hidden losses in multi-step processes.

Counted or classified observations, such as pass/fail, number of defects, or defect type, which can take only whole-number values and require different statistical tools than continuous data.

A visual tool that organizes possible causes of a problem into branches - commonly people, machine, method, material, measurement, and environment - to guide the investigation.

A questioning technique that repeatedly asks why a problem occurred, drilling from the visible symptom down to a root cause that can actually be addressed.

A statistical procedure that uses sample data to decide whether an observed difference or relationship is real or could plausibly have arisen from ordinary random variation.

The probability of obtaining a result at least as extreme as the one observed if the null hypothesis were true; small values suggest the observed effect is unlikely to be chance.

Rejecting a true null hypothesis - concluding there is a real difference when there is none; its risk is set by the alpha level, commonly 0.05.

Analysis of variance: a test that compares the means of three or more groups by examining how much variation lies between the groups versus how much lies within them.

A method that models the relationship between one or more input variables and a continuous output, giving an equation for prediction and showing how much variation the inputs explain.

A chart that displays variation from several sources at once - within a piece, between pieces, and over time - helping locate where variation mainly lives before formal testing.

A structured set of planned tests in which input factors are deliberately varied together, so their individual and combined effects on a response can be measured efficiently.

An input variable deliberately set at different values in an experiment to see whether changing it shifts the response.

A situation where the effect of one factor on the response depends on the level of another factor, so the two factors cannot be judged separately.

Continuous, incremental improvement carried out by the people who do the work, often through short focused events that eliminate waste in a specific area.

A device or design feature that makes it impossible, or immediately obvious, to perform a step incorrectly, preventing defects rather than detecting them later.

A team-based method that lists potential failure modes, rates their severity, occurrence, and detection, and prioritizes actions to reduce the highest-risk failures.

A document listing process steps, the measurements to monitor, the control method, sample size and frequency, and the action to take when a result falls outside requirements.

A time-ordered plot of a process measure with statistically calculated limits, used to distinguish normal variation from signals that the process has changed.

Normal, everyday variation built into a process by many small sources; reducing it requires changing the process itself rather than reacting to individual data points.

Unusual variation from an identifiable source outside the normal process, such as a bad batch, a new operator, or a machine fault, which appears as a signal on a control chart.

A pair of control charts for continuous data: one tracks subgroup averages (X-bar) and the other tracks subgroup ranges (R), monitoring both the centering and the spread of a process.

A pre-agreed set of steps telling operators whom to notify and what to check, contain, and record when a control chart signals an out-of-control condition.