CLSSGB-001 practice question 23 of 30

Conducting a viable Capability Analysis using Attribute Data one must obtain a fairly large sample set to be statistically sound.

Full set: 200 questions
Answer options
A: True Correct answer
B: False
Correct Answer:
A. True
Explanation

The statement is true. Capability analysis for attribute data (e.g., pass/fail, go/no-go) relies on estimating proportions, such as the fraction of nonconforming items. Compared to variable data, which provides a measurement for each unit, attribute data is less informative. To achieve a statistically sound estimate of the true process proportion with a reasonable level of confidence and precision, a much larger sample size is required. A small sample may not capture any defects, leading to an inaccurate and overly optimistic assessment of process capability. Therefore, a large sample set is crucial to ensure the observed defect rate is representative and the analysis is viable.

Why Incorrect

B. False: This is incorrect. Using a small sample set for attribute data leads to high uncertainty and wide confidence intervals for the capability estimate, making the analysis statistically unreliable and potentially misleading.

References

1. Montgomery, D. C. (2013). Introduction to Statistical Quality Control (7th ed.). Wiley.

Reference: In Chapter 9, Section 9.5, "Process Capability Analysis with Attribute Data," Montgomery explains that when dealing with proportions (attribute data), especially for low-defect processes, "very large sample sizes are required to estimate the process fraction nonconforming p with adequate precision." This directly supports the need for a large sample set.

2. Pyzdek, T., & Keller, P. A. (2018). The Six Sigma Handbook (5th ed.). McGraw-Hill Education.

Reference: Chapter 19, "Process Capability for Attribute Data," states, "Attribute data provides less information than variables data. Because of this, attribute analysis requires relatively large sample sizes to provide the desired degree of statistical confidence." This is a foundational text often used in university-level quality management courses.

3. The Pennsylvania State University. (n.d.). STAT 501: Regression Methods Course Notes. Eberly College of Science.

Reference: In the lesson on Inference for Proportions, the course notes discuss the conditions for using the normal approximation to the binomial distribution, which is fundamental for many attribute data analyses. The condition often cited is that both np and n(1-p) should be greater than 5 or 10, where n is the sample size and p is the proportion. This mathematical requirement inherently forces a large sample size (n) when the proportion of defects (p) is small.

Scroll to Top