Measure Phase

Process characteristics, data collection, measurement systems, basic statistics, probability, and process capability.

5 minutes 5 Questions

The Measure Phase is the second critical stage in the DMAIC (Define, Measure, Analyze, Improve, Control) methodology used in Lean Six Sigma Black Belt training. This phase focuses on establishing baseline performance metrics and collecting reliable data about the current process. Key objectives of…

Concepts covered
Addition and Multiplication RulesPoisson DistributionTouch Time and Cycle TimeHidden Factory ConceptProcess Mapping and FlowchartsQualitative vs Quantitative DataMeasurement Scales (Nominal, Ordinal, Interval, Ratio)Random, Stratified, and Systematic SamplingMeasurement System Analysis (MSA)Bias, Linearity, and Stability in MSAAttribute Measurement System AnalysisCentral Limit TheoremBinomial DistributionDescriptive Statistics (Dispersion and Central Tendency)Normal Probability PlotsBasic Probability ConceptsIndependence and Mutually Exclusive EventsConditional and Complementary ProbabilityNormal DistributionWeibull, Exponential, and Lognormal DistributionsProcess Capability Indices (Cp, Cpk)Process Capability for Attributes DataProcess Capability for Non-Normal DataThroughput and Process ConstraintsGemba WalkContinuous and Discrete DataSampling Concepts and MethodsData Collection Plans and IntegrityMetrology and Calibration SystemsPopulation Parameters vs Sample StatisticsGraphical Methods (Box Plots, Histograms, Scatter Diagrams)Descriptive vs Inferential StatisticsChi-Square, Student's t, and F DistributionsProcess Performance Indices (Pp, Ppk, Cpm)Process Sigma LevelShort-Term vs Long-Term CapabilityProcess Flow Metrics (WIP, WIQ, Takt Time)Value Stream MappingSpaghetti DiagramsGage Repeatability and Reproducibility (R&R)Designing Process Capability StudiesBox-Cox TransformationNatural Process Limits vs Specification LimitsPPM, DPMO, DPU, and Yield Metrics
Test mode:
Start practice test

CSSBB - Measure Phase Example Questions

Test your knowledge of Measure Phase

Question 1

Which of the following is an example of quantitative data?

Correct Answer: Cycle time measured in seconds for each unit produced

Cycle time measured in seconds is quantitative data because it represents numerical measurements on a continuous scale that can be mathematically analyzed, such as calculating means, standard deviations, and performing statistical tests.

Customer satisfaction categories like satisfied or dissatisfied are qualitative (categorical/attribute) data representing discrete classifications. Defect types such as scratch, dent, or crack are also qualitative nominal data used for classification, not measurement. Supplier names are qualitative nominal data with no numerical or measurable meaning.

The key concept tested is the distinction between quantitative (continuous/discrete numerical) data and qualitative (attribute/categorical) data, which is fundamental to selecting the correct analysis tools in Six Sigma.

Question 2

A team is planning an Attribute MSA for a go/no-go inspection at a packaging line. They have 3 inspectors and want to include 50 samples. A colleague suggests using only samples that clearly pass or clearly fail to make the study easier to run. What is the best reason to challenge this sampling approach?

Correct Answer: The study needs borderline samples to truly test the measurement system's ability to discriminate near the specification boundary

The whole purpose of an Attribute MSA is to evaluate whether inspectors can consistently and correctly classify parts near the specification boundary, where real-world disagreements actually occur. Including borderline samples is essential because that is where the measurement system is most challenged and most likely to fail.

Suggesting that only clear samples ensure higher Kappa values is precisely the problem: artificially inflating agreement tells you nothing useful about the system's true capability. Borderline samples do not introduce bias into the Kappa calculation; they reveal the measurement system's true discrimination ability. Clear pass/fail samples alone do not represent the critical decision zone in production.

Key concept tested: proper sample selection strategy in Attribute Measurement System Analysis.

Question 3

A team is setting up a Gage R&R study for a new hardness tester. They have 3 operators available and 10 sample parts selected from recent production. The quality manager asks the team to determine how many trials each operator should perform per part. Given that the team wants reliable results while keeping the study practical, how many trials per operator per part should they plan according to standard Gage R&R practice?

Correct Answer: 2 to 3 trials per operator per part is the standard recommendation for a reliable study

The standard Gage R&R study design recommends 2 to 3 trials (replications) per operator per part, which balances statistical reliability with practical feasibility, typically resulting in 60 to 90 total measurements when using 3 operators and 10 parts.

One trial per operator per part does not allow estimation of repeatability (within-operator variation), which is a fundamental component of the study. Four to five trials per part is excessive and unnecessarily time-consuming for most applications without meaningful improvement in estimates. The claim that more trials are needed with fewer operators conflates two independent components of measurement system variation.

Key concept tested: standard design parameters for a crossed Gage R&R study.

More Measure Phase questions
1185 questions (total)
Practice 100 questions
One session at a time, always new questions