The Empirical Distribution Function with Arbitrarily Grouped, Censored, and Truncated Data
Abstract
This paper is concerned with the nonparametric estimation of a distribution function F, when the data are incomplete due to grouping, censoring and/or truncation. The situation occurs frequently in survivorship, reliability, and recidivism analysis. Using the idea of self-consistency, a simple algorithm is constructed and shown to converge monotonically to yield a maximum likelihood estimate of F. The procedure compares favourably with the more cumbersome Newton-Raphson method. A test is proposed for comparing two distributions when data on one or both is incomplete and some other applications of the empirical distribution function are indicated.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jul 01, 1976
- Accession Number
- ADA030940
Entities
People
- Bruce W. Turnbull
Organizations
- Cornell University