[ | Next | Previous | Up ]

Sampling Plan for Attribute Identification

From: Roy
Date: 26 Nov 2003
Time: 13:41:26

Comments

Our company makes investment castings. After the foundry operations, we inspect the batches of castings 100% to either scrap, salvage, or pass "as is". The operators are also required to record attribute defect data of the scrap. There may be 5 to 7 different attributes recorded. This attribute data is used to trouble shoot the foundry. The problem is; too many inspectors produce too many opinions and thus incorrect data is supplied to process engineering. I am in the process of setting up a sampling plan to record the attribute data by a smaller more accurate group of inspectors. I will require MSA (measurement system analysis)for this smaller group. My question is, what would be the correct sample size to capture the foundry defect attributes for a batch of 1134 pieces when the average scrap per batch has been 12% with a 95% confidence interval?


Last changed: November 20, 2007