A simple method for probability proportional to size (. pi. ps) sampling without replacement based on ranks
There can be gains in estimation efficiency over equal probability and unequal probability with replacement sampling methods when one makes use of auxiliary information for probability proportional to size without replacement (..pi..pswor) sampling methods. Such methods are often complex to execute, particularly when the sample size n exceeds 2. Also, appropriate auxiliary information may not be available. When the population units can be ranked according to the variable of interest, we demonstrate that there can be gains in estimation efficiency with a very simple ..pi..pswor sampling strategy based on ranks for all sample sizes n where the population size N = 2kn (k a positive integer).
- Research Organization:
- Oak Ridge National Lab., TN (USA)
- DOE Contract Number:
- AC05-84OR21400
- OSTI ID:
- 6504729
- Report Number(s):
- ORNL/TM-10398; ON: DE87010146
- Country of Publication:
- United States
- Language:
- English
Similar Records
An empirical investigation of the use of ranks in. pi. ps sampling
Effect of initial seed and number of samples on simple-random and Latin-Hypercube Monte Carlo probabilities (confidence interval considerations)
(An empirical investigation of the use of ranks in. pi. ps sampling)
Conference
·
Wed Dec 31 23:00:00 EST 1986
·
OSTI ID:6006690
Effect of initial seed and number of samples on simple-random and Latin-Hypercube Monte Carlo probabilities (confidence interval considerations)
Conference
·
Thu May 04 00:00:00 EDT 2000
·
OSTI ID:756096
(An empirical investigation of the use of ranks in. pi. ps sampling)
Technical Report
·
Wed Sep 30 00:00:00 EDT 1987
·
OSTI ID:6799313