ANSI/ASQ Z1.9 (formerly MIL-STD 414) is a generally accepted sampling standard for products with real-number (variables) data. Acceptance or rejection depends on the sample statistics. The sampling plan itself is a function of the lot size, acceptable quality level (AQL), and sampling level. Switching rules similar to those for ANSI/ASQ Z1.4 (MIL-STD 105) determine whether sampling is at the normal, tightened, or reduced level.
- Know that these plans rely on the assumption that the critical to quality characteristic follows the normal (bell curve) distribution.
- The required sample size is much lower when the process standard deviation is known (as it might be from a process capability study) than when it must be estimated from the sample statistics.
- Estimate, where necessary, the nonconforming fraction at each specification limit to make the acceptance/rejection decision.
- Use of the range method; an important takeaway is that ranges are for groups of 5 parts and not the total sample. If the sample size is 15, the range is the average range for three groups of five parts as opposed to one range for 15 parts.
Who Will Benefit:
Manufacturing and quality engineers, technicians, and inspectors with responsibility for inspections that involve quantitative measurements of parts.
William A. Levinson, P.E., FASQ, CFPIM, is the owner of Levinson Productivity Systems PC. He holds professional certifications from the American Society for Quality, APICS, and Society of Manufacturing Engineers.