Certified Six Sigma Black Belt Exam Measurement System Analysis 4 — Questions and Answers
Question 1: The AIAG MSA manual recommends that %Tolerance (P/T ratio) be used when the primary concern is:
- Understanding the measurement system relative to process variation
- Determining if the gage is adequate to hold the product specification (Correct answer)
- Estimating the number of distinct categories
- Evaluating operator training needs
Correct answer: Determining if the gage is adequate to hold the product specification
The P/T ratio (precision-to-tolerance) compares gage variation to the specification tolerance, making it the right metric when conformance decisions are the focus.
Question 2: A measurement system has a %R&R of 22% and an ndc of 3. Which statement is most accurate?
- The system is acceptable because %R&R is below 30%
- The system is unacceptable primarily because ndc is below 5 (Correct answer)
- Both metrics indicate the system is marginally acceptable
- The system should be evaluated solely on ndc, not %R&R
Correct answer: The system is unacceptable primarily because ndc is below 5
An ndc below 5 means the system cannot distinguish enough categories within the process variation, making it inadequate regardless of %R&R being marginally acceptable.
Question 3: In an MSA study for a vision inspection system, 50 parts (25 good, 25 defective) are presented to 3 operators twice. This study design best evaluates:
- Variable measurement system capability
- Attribute agreement analysis (Kappa) for an automated or human visual inspector (Correct answer)
- Gage linearity across the measurement range
- Gage stability over a 30-day period
Correct answer: Attribute agreement analysis (Kappa) for an automated or human visual inspector
Presenting known-status parts to multiple appraisers in replicated trials is the standard attribute agreement analysis design, evaluated using Kappa statistics.
Question 4: Which of the following is a correct interpretation of 'bias' in a measurement system?
- The spread of repeated measurements on the same part
- The systematic difference between the observed average measurement and the true (reference) value (Correct answer)
- Variation caused by different operators
- The inability of a gage to detect small changes
Correct answer: The systematic difference between the observed average measurement and the true (reference) value
Bias is the systematic (consistent) error — the offset between the measurement system's average reading and the accepted reference value.
Question 5: A Six Sigma team discovers that calibration records for a critical measurement device are 6 months overdue. What should be done FIRST before using data collected during that period?
- Discard all data and repeat all measurements
- Conduct an immediate bias and stability study to assess the magnitude of potential error (Correct answer)
- Recalibrate the device and assume all prior data is valid
- Increase sample size to compensate for potential measurement error
Correct answer: Conduct an immediate bias and stability study to assess the magnitude of potential error
Assessing the current bias and reviewing historical stability data helps quantify the risk before deciding whether prior data is usable or must be discarded.
Question 6: In a Gage R&R ANOVA table, which F-ratio is used to test whether part-to-part variation is statistically significant?
- F = MS(Part) / MS(Repeatability)
- F = MS(Operator) / MS(Repeatability)
- F = MS(Part) / MS(Operator×Part) (Correct answer)
- F = MS(Operator×Part) / MS(Repeatability)
Correct answer: F = MS(Part) / MS(Operator×Part)
In a crossed Gage R&R ANOVA, Part variation is tested against the Operator×Part interaction term as the appropriate error term.
Question 7: A temperature sensor is used to monitor a process over the range 50°C–250°C. A linearity study reveals the bias is negligible at 50°C but reaches +4°C at 250°C. What action is most appropriate?
- Accept the sensor since average bias across the range is low
- Apply a correction factor at high temperatures or replace the sensor (Correct answer)
- Reduce the operating range to below 150°C
- Increase calibration frequency to monthly
Correct answer: Apply a correction factor at high temperatures or replace the sensor
Non-constant bias across the operating range (poor linearity) requires either a correction factor at affected values or replacing the sensor with one that performs linearly.
Certified Six Sigma Black Belt Exam Measurement System Analysis 4
The AIAG MSA manual recommends that %Tolerance (P/T ratio) be used when the primary concern is: