An Estimation Problem in Faulty Inspection Sampling. Revised.
Abstract
Suppose samples of size n are chosen at random from lots of size N and examined for the presence (or absence) of a certain type (or types) of defect. If inspection is not perfect, there may be a probability, p, of failing to detect a defect when it is, in fact, present. It is desired to estimate p from records of the number of items found to have the defect(s) on inspection. In the absence of further information, the problem cannot be answered on the basis of single samples, of size n, from each of a number of lots. Some further assumptions are needed - such as supposing the number of defective items per lot is (nearly) constant. Alternatively, some more elaborate sampling - for example, repeated sampling from the same lot - may provide the necessary information. The paper describes several methods of estimation utilizing various forms of additional data and information. Some as yet unsolved problems are briefly discussed.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jun 01, 1984
- Accession Number
- ADA141247
Entities
People
- N. L. Johnson
- S. Kotz
Organizations
- University of Maryland