Certified Six Sigma Black Belt Exam Certified Six Sigma Black Belt Measurement System Analysis (MSA) 1 — Questions and Answers
Question 1: In Measurement System Analysis, what does 'bias' specifically refer to?
- The variation observed when a single appraiser measures the same part multiple times
- The difference between the observed average measurement and the accepted reference value (Correct answer)
- The change in measurement system performance over an extended period
- The inconsistency of measurements across the full operating range of the gauge
Correct answer: The difference between the observed average measurement and the accepted reference value
Bias (also called accuracy) is the systematic error component — it quantifies how far the average of repeated measurements deviates from the true or reference value. It is determined by comparing the mean of multiple readings on a reference standard to its known value.
Question 2: Which source of variation is captured by the ANOVA method in a Gage R&R study that the traditional Range (Average & Range) method cannot detect?
- Equipment variation (repeatability)
- Appraiser variation (reproducibility)
- Appraiser-by-part interaction (Correct answer)
- Part-to-part variation
Correct answer: Appraiser-by-part interaction
The ANOVA method partitions variance into equipment variation, appraiser variation, AND the interaction between appraisers and parts. The Range method combines the interaction term into reproducibility and therefore cannot isolate it. Significant interaction means different appraisers are inconsistent specifically on certain parts.
Question 3: A Gage R&R study reports that the %Contribution of total Gage R&R to total variation is 6%. What does this value indicate about the measurement system?
- The measurement system is unacceptable because %Contribution should exceed 10%
- The measurement system is acceptable because measurement error accounts for only 6% of total observed variance (Correct answer)
- The measurement system requires conditional approval and further cost-benefit analysis
- The number of distinct categories (ndc) will be less than 2, making the system unfit for use
Correct answer: The measurement system is acceptable because measurement error accounts for only 6% of total observed variance
%Contribution (variance-based) below 1% is ideal, and below 9% is generally considered acceptable per AIAG MSA guidelines. At 6%, the measurement system is acceptable — gauge error contributes very little to the observed variation, leaving the majority attributable to real part-to-part differences.
Question 4: In an Attribute Agreement Analysis, a Kappa value of 0.52 is calculated for an appraiser compared to the standard. How should this result be interpreted?
- Excellent agreement — the measurement system is fully acceptable
- Moderate agreement — the system may require improvement before use in production (Correct answer)
- Poor agreement — the appraiser must be retrained and the study repeated immediately
- The Kappa statistic is not applicable to attribute measurement systems
Correct answer: Moderate agreement — the system may require improvement before use in production
Kappa values are typically interpreted as: <0.40 = poor, 0.40–0.75 = moderate/fair, >0.75 = excellent. A Kappa of 0.52 falls in the moderate range, signaling that the attribute measurement system has meaningful disagreement with the standard and should be improved before relying on it for critical decisions.
Question 5: When is a nested (hierarchical) Gage R&R design required instead of a crossed Gage R&R design?
- When the number of appraisers exceeds five
- When the measurement process is destructive and the same part cannot be remeasured (Correct answer)
- When the gauge has a known linearity problem across its range
- When parts come from more than one production line
Correct answer: When the measurement process is destructive and the same part cannot be remeasured
A crossed design requires each appraiser to measure every part multiple times, which is impossible when measurement destroys the part (e.g., pull-strength testing, chemical titration). In a nested design, each appraiser measures a unique set of parts, so parts are 'nested' within appraisers, allowing repeatability and reproducibility to still be estimated.
Question 6: A measurement system's stability study reveals a gradual upward drift in readings on a reference standard over several weeks. What is the most appropriate corrective action?
- Increase the number of appraisers to offset the drift through averaging
- Investigate and address root causes such as tool wear, environmental changes, or calibration drift, then recalibrate the gauge (Correct answer)
- Switch from the Range method to the ANOVA method to better quantify the drift
- Widen the specification limits to accommodate the unstable measurement system
Correct answer: Investigate and address root causes such as tool wear, environmental changes, or calibration drift, then recalibrate the gauge
Stability (or drift) indicates the measurement system is changing over time, which violates a fundamental MSA assumption. The correct response is to identify the cause — worn gauge components, temperature fluctuations, or calibration interval too long — and recalibrate or repair the system. Masking instability with wider specs or more appraisers does not fix the root cause.
Certified Six Sigma Black Belt Exam Certified Six Sigma Black Belt Measurement System Analysis (MSA) 1
In Measurement System Analysis, what does 'bias' specifically refer to?