TOLERANCE LIMITS FOR SMALL LOTS WHEN SAMPLE PIECES ARE SOMETIMES MISCLASSIFIED

Technical Report · Wed Sep 20 04:00:00 EDT 1961

OSTI ID:4828361

Burrows, G L

Tables are presented that show the minimum number of pieces that must be examimed from a lot in order to insure that a certain amount of the pieces meet desired specifications. Three different tables are presented in order that the correct sampling plan may be chosen. The first table gives the minimum sample size required from finite lots to provide assurance against an excessive number of defectives. The second table shows the increase in sample size necessary to obtain the same assurance against defectives as in the first table when defectives sampled from the lot may be erroneously classified as non-defective. The third table shows that possible errors in misclassifying both defectives and non-defectives lead to reductions rather than increases in sample size. Illustrations are given showing that protection against excessive defectives is achieved with reduced sample size only by a marked reduction in the probability of accepting acceptable lots. Illustrations are also given on how the minimum sample size criterion can be modified to get a sampling plan that provides assurance against an excessive number of defectives while at the same time accepting lots with low numbers of defectives with fairly high probability. (N.W.R.)

• Research Organization: Knolls Atomic Power Lab., Schenectady, N.Y.
• DOE Contract Number: W-31-109-ENG-52
• NSA Number: NSA-16-003384
• OSTI ID: 4828361
• Report Number(s): KAPL-M-GLB-6
• Country of Publication: United States
• Language: English

Related Subjects

ERRORS
MATHEMATICS
MATHEMATICS AND COMPUTERS
SAMPLING
STATISTICS