Two Items Concerning Directional Data
Abstract
A discussion is given of confidence procedures for the model vector A of a von Mises or Fisher distribution. Slight revisions of confidence levels are suggested because the usual procedure strictly gives the axis but not the direction of A. The likelihood ratio test is given, with percentage points, for testing A = A sub 0, for the Fisher distribution. A good approximation is offered for the similar test for the von Mises distribution.
Document Details
- Document Type
- Technical Report
- Publication Date
- Apr 03, 1972
- Accession Number
- AD0741702
Entities
People
- M. A. Stephens
Organizations
- Stanford University