Square-Root and Cube-Root Estimators. The Developments in Sample Survey Theory
Abstract
When the situation arises that the population being sampled is known, a priori, to have characteristics that place it in a more restricted category, the question rightly may be asked as to whether the condition of uniform unbiasedness might not be dropped and the bias and variance be combined into a single measure of mean square error. The purpose of the research is to investigate a class of estimators, which shall be called root estimators, that will usually have smaller mean square error than the arithmetic mean for certain classes of populations.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 01, 1969
- Accession Number
- AD0742678
Entities
People
- Herman Otto Hartley
- L. J. Ringer
- O. C. Jenkins
Organizations
- Texas A&M University