Linear Statistical Regression and Functional Relations
Abstract
An approach to linear least squares or regression with errors of measurement in either one or both variables is covered, giving a computationally convenient procedure for all of the appropriate statistical significance tests or tests of hypotheses concerning the true unknown parameters and the fitted line. The problem of meaningful physical functional relations is also discussed, showing the relation to and a comparison with usual least squares. Pertinent examples illustrating various applications of the theory are also given.
Document Details
- Document Type
- Technical Report
- Publication Date
- Nov 01, 1975
- Accession Number
- ADA018651
Entities
People
- Frank E. Grubbs
Organizations
- Ballistic Research Laboratory