CSSGB practice question 24 of 30

The perfect sample size is the minimum number of data points required to provide exactly 6% overlap or risk if one wants a 95% confidence level.

Full set: 148 questions

Domain Measure Phase

Answer options
A: True
B: False Correct answer
Correct Answer:
B. False
Explanation

This statement is false. A 95% confidence level directly corresponds to a 5% risk of a Type I error (alpha, α), which is the probability of incorrectly rejecting a true null hypothesis (α = 1 - 0.95 = 0.05). The figure of 6% is incorrect. Furthermore, the "perfect" or required sample size is not determined by confidence level alone. It is a function of four key parameters: the confidence level (α), the desired statistical power (1-β, the probability of detecting an effect when it is present), the size of the effect one wishes to detect (delta, Δ), and the variation within the population (standard deviation, σ). The concept of "overlap" is related to these factors, not a fixed percentage tied only to the confidence level.

Why Incorrect

A. True: This is incorrect because a 95% confidence level implies a 5% alpha risk, not 6%. The calculation for sample size is also dependent on power, effect size, and variation, not a fixed "overlap" value.

References

1. The Certified Six Sigma Green Belt Handbook, 2nd Edition. (2015). ASQ Quality Press. In Chapter 17, "Hypothesis Testing," the section on "Determining Sample Size" (pp. 318-320) explicitly shows that sample size n is a function of the risks (α and β), the standard deviation (σ), and the size of the difference to be detected (Δ). It also defines the confidence level as 1-α, confirming that 95% confidence corresponds to α=0.05.

2. Montgomery, D. C. (2013). Introduction to Statistical Quality Control, 7th Edition. John Wiley & Sons. Chapter 4, "Statistical Process Control," and Chapter 9, "Hypothesis Testing," define the confidence level and its relationship to the Type I error rate (α). Specifically, on page 89, it states, "For example, if we set the probability of a type I error to α = 0.05, it is equivalent to constructing a 95 percent confidence interval." This directly contradicts the 6% figure in the question.

3. Pennsylvania State University. (n.d.). STAT 500: Applied Statistics, Lesson 3.1.1 - The basics of Confidence Intervals. Penn State Eberly College of Science. Retrieved from https://online.psu.edu/stat500/lesson/3/3.1/3.1.1. The courseware states, "The confidence level is the associated probability... A 95% confidence level is common. The corresponding alpha-risk is α = 1 - 0.95 = 0.05." This academic source confirms the correct risk percentage is 5%, not 6%.

Scroll to Top