TS Measurement Systems Analysis (MSA) 2 — Questions and Answers
Question 1: In a standard AIAG crossed GR&R study, how many parts, appraisers, and trials are most commonly recommended?
- 5 parts, 2 appraisers, 3 trials
- 10 parts, 3 appraisers, 2 trials (Correct answer)
- 10 parts, 2 appraisers, 5 trials
- 5 parts, 3 appraisers, 5 trials
Correct answer: 10 parts, 3 appraisers, 2 trials
The AIAG MSA manual recommends 10 parts, 3 appraisers, and 2 trials as the standard design for a crossed GR&R study to provide adequate statistical power.
Question 2: What does the 'Number of Distinct Categories' (NDC) indicate in an MSA study?
- The number of operators required to complete the study
- The number of different measurement tools tested during the study
- The number of non-overlapping data groups the measurement system can reliably distinguish (Correct answer)
- The number of parts that failed measurement system requirements
Correct answer: The number of non-overlapping data groups the measurement system can reliably distinguish
NDC indicates how many distinct groups of parts the measurement system can differentiate; an NDC of 5 or more is generally required for a capable measurement system.
Question 3: Which advantage does the ANOVA method offer over the range method in a GR&R study?
- It requires fewer parts and operators to complete
- It can separate and quantify the part-by-operator interaction effect (Correct answer)
- It eliminates the need for a reference standard
- It produces a lower %GRR result by default
Correct answer: It can separate and quantify the part-by-operator interaction effect
The ANOVA method produces a part-by-operator interaction term showing whether different operators measure different parts differently — an effect the range method cannot detect.
Question 4: What does the %P/T ratio measure in measurement systems analysis?
- The percentage of parts that pass tolerance inspection
- The ratio of gauge variation to the product tolerance (specification range) (Correct answer)
- The proportion of total variation explained by the process
- The percentage of trials completed within the calibration interval
Correct answer: The ratio of gauge variation to the product tolerance (specification range)
The %P/T (Precision-to-Tolerance) ratio compares the measurement system's variation to the specification tolerance, indicating whether the gauge is precise enough for the inspection task.
Question 5: 'Stability' in measurement systems analysis refers to:
- The ability of a gauge to produce zero variation between consecutive readings
- The change in bias of a measurement system over time (Correct answer)
- The consistency of measurements within a single short measurement session
- The maximum allowable difference between two operators' readings
Correct answer: The change in bias of a measurement system over time
Stability (or drift) evaluates whether the measurement system's bias remains constant over an extended period, typically assessed by periodically measuring a master or reference part.
Question 6: In an attribute MSA study, a Type I error (false alarm) occurs when:
- A conforming part is incorrectly classified as nonconforming (Correct answer)
- A nonconforming part is incorrectly classified as conforming
- An operator measures a part outside the tolerance range
- The gauge fails to measure parts at the extreme ends of its range
Correct answer: A conforming part is incorrectly classified as nonconforming
A Type I error (false alarm or false rejection) in attribute MSA occurs when a good part is mistakenly called bad, causing unnecessary rework or scrap costs.
Question 7: What is the primary purpose of conducting 'blind' measurements in an MSA study?
- To prevent appraisers from knowing the gauge's calibration status
- To ensure appraisers measure parts without knowledge of previous readings or part identity, reducing bias (Correct answer)
- To hide measurement results from production supervisors during the study
- To test the gauge under low-light or adverse operating conditions
Correct answer: To ensure appraisers measure parts without knowledge of previous readings or part identity, reducing bias
Blind measurements prevent appraisers from being influenced by previous results or expectations, ensuring observed variation reflects true measurement system performance.
In a standard AIAG crossed GR&R study, how many parts, appraisers, and trials are most commonly recommended?