Tests and efficiencies of separate regression models

1 August 1978

journal article
Published by Oxford University Press (OUP) in Biometrika

Vol. 65 (2) , 319-327
https://doi.org/10.1093/biomet/65.2.319

Abstract

The results of Cox (1961, 1962) on tests of separate families of hypotheses are used to develop tests for some log linear regression models. The consequences of using one model when another is true are studied. The probability limits and asymptotic efficiencies of the estimators of the regression coefficients when using a false model in relation to the true model are investigated

Keywords

EFFICIENCY
FALSE MODEL
LIKELIHOOD RATIO
LOG LINEAR REGRESSION MODEL
PROBABILITY LIMIT
SEPARATE FAMILIES OF HYPOTHESES

This publication has 0 references indexed in Scilit: