Robust C(Alpha)-type Tests for Linear Models.
Abstract
A new class of tests that can be viewed as a generalization of Neyman's optimal C(Alpha) test is introduced. An otpimally robust test which maximizes the aymptotic power within this class, subject to a bounded influence function, is selected. It is shown that it is equivalent to a certain asymptotically minimax test proposed by Wang (1981). Finally, some numerical results on the asymptotic behavior of robust C(Alpha)-type tests under several errors' and carriers' distributions are discussed. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Feb 01, 1984
- Accession Number
- ADA143177
Entities
People
- E. Ronchetti
Organizations
- Princeton University