Certified Six Sigma Black Belt Exam Measurement System Analysis (MSA) 5 — Questions and Answers
Question 1: Which of the following is the correct interpretation when a Gauge R&R %GRR (based on study variation) is reported as 8%?
- The gauge is marginal and needs improvement before use
- The gauge is unacceptable and must be replaced
- The gauge is acceptable for most applications (Correct answer)
- The gauge cannot distinguish between any process categories
Correct answer: The gauge is acceptable for most applications
Per AIAG guidelines, a %GRR below 10% indicates an acceptable measurement system suitable for most applications.
Question 2: What is the key difference between 'precision' and 'accuracy' in a measurement system?
- Precision refers to closeness to the true value; accuracy refers to repeatability
- Accuracy refers to closeness to the true value; precision refers to consistency of repeated measurements (Correct answer)
- Precision and accuracy are synonymous in MSA terminology
- Accuracy is measured by Gauge R&R; precision is measured by calibration
Correct answer: Accuracy refers to closeness to the true value; precision refers to consistency of repeated measurements
Accuracy (low bias) means measurements are centered on the true value, while precision (low variation) means repeated measurements cluster tightly together regardless of where they center.
Question 3: When should a Expanded Uncertainty be used instead of a simple %GRR in reporting measurement system performance?
- When the process is in statistical control
- When measurement results are used for legal, regulatory, or customer-contractual purposes (Correct answer)
- When the number of parts in the study is fewer than 5
- When only one operator performs all measurements
Correct answer: When measurement results are used for legal, regulatory, or customer-contractual purposes
Expanded Uncertainty (per ISO/IEC 17025 or GUM) is required for calibration labs and situations where measurement results have legal, regulatory, or contractual significance.
Question 4: A Gauge R&R study is run on a force-testing machine where each tested spring is permanently deformed after measurement. The appropriate study design is:
- Crossed Gauge R&R with 3 operators
- Nested (destructive) Gauge R&R design (Correct answer)
- Type 1 Gauge Study
- Attribute MSA with Kappa analysis
Correct answer: Nested (destructive) Gauge R&R design
Destructive tests require a nested Gauge R&R design because each part can only be measured once, so operators cannot all measure the identical same parts.
Question 5: In MSA, 'effective resolution' differs from 'stated resolution' because it accounts for:
- The number of operators in the study
- The actual measurement variation relative to the readable increment (Correct answer)
- Calibration frequency of the gauge
- The number of replications per operator
Correct answer: The actual measurement variation relative to the readable increment
Effective resolution considers whether the gauge's increments are fine enough relative to actual measurement variation to meaningfully distinguish different values.
Question 6: Which statistical test is typically used to assess whether bias in a measurement system is statistically significant?
- F-test comparing two variances
- One-sample t-test comparing the mean measurement to the reference value (Correct answer)
- Chi-square test of independence
- Anderson-Darling normality test
Correct answer: One-sample t-test comparing the mean measurement to the reference value
A one-sample t-test determines if the average measurement is statistically different from the known reference (true) value, indicating significant bias.
Question 7: When analyzing a Gauge R&R using the ANOVA method, a p-value > 0.25 for the operator-by-part interaction term typically leads to:
- Retaining the interaction term and reporting it separately
- Pooling the interaction into the repeatability (error) term (Correct answer)
- Flagging the study as invalid and repeating it
- Using the interaction variance to inflate the reproducibility estimate
Correct answer: Pooling the interaction into the repeatability (error) term
When the operator-by-part interaction is not statistically significant (p > 0.25 per AIAG guidance), it is pooled into the repeatability error term to produce a cleaner model.
Certified Six Sigma Black Belt Exam Measurement System Analysis (MSA) 5
Which of the following is the correct interpretation when a Gauge R&R %GRR (based on study variation) is reported as 8%?