AN EXTREME-VALUE PROBLEM RELATED TO A SAMPLING STUDY.
Abstract
This memorandum considers a problem in which a large sample called a collection is taken from a known infinite population and then the collection is sampled. Basically the problem under study is the priori comparison (difference in magnitude) of the upper bound of the collection with the upper bound of the final sample. Briefly the results of the study show that if a small difference is desired, the ratio of sample size to collection size required depends heavily on the nature of the upper tail of the density function of the original population. An increase in the upper bound of the population and an increase in the rate of decay of the density function left of the bound both tend to increase required sample size. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Jun 22, 1964
- Accession Number
- AD0601736
Entities
People
- M. L. Eaton