A Nonparametric Test for Trend Alternatives,
Abstract
Let (X sub j; j=1,2,...) be a sequence of independently distributed real valued random variables. A non-parametric test of the hypothesis that the X sub j's are identically distributed against the alternative that the X sub j's are stochastically increasing, is given. The asymptotic power of the test for certain regression alternatives is given, and its efficiency is compared with the most powerful test for the normal alternative. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Oct 15, 1971
- Accession Number
- AD0736168
Entities
People
- James R. Thompson
- Khursheed Alam
Organizations
- Clemson University