Estimators for the truncated beta-binomial distribution
Let X have a beta-binomial(m,p,theta) distribution, truncated such that X > t for t = 0 or 1. Suppose that independent observations of X are available. A consistent estimator of (p,theta) is given, based on the first three sample moments. This may be used as a start for maximum likelihood estimation or jackknifing. The standard assumptions for a C(..cap alpha..) is truncated binomial do not hold; however, a test is proposed based on jackknifing the sample variance of X. Some Monte Carlo comparisons are given. For moderately small data sets, these comparisons show that the moment estimator is often superior to the MLE, and the C(..cap alpha..) test is superior to other proposed tests, in spite of its lack of theoretical justification. 3 figures, 5 tables.
- Research Organization:
- EG and G Idaho, Inc., Idaho Falls (USA)
- DOE Contract Number:
- AC07-76ID01570
- OSTI ID:
- 5359803
- Report Number(s):
- CONF-800644-1
- Country of Publication:
- United States
- Language:
- English
Similar Records
Accuracy functions for confidence bounds: a basis for sample size determination
Properties of parameter estimation techniques for a beta-binomial failure model. Final technical report