The Optimal Size of a Preliminary Test for Linear Restrictions when Estimating the Regression Scale Paramter

This paper considers the choice of critical value for a pre-test of exact linear restrictions when estimating the regression error variance. We calculate the critical value according to a mini-max risk regret criterion and compare the resulting risk functions with those generated by using the critical value which minimises the pre-test risk function. The results suggest that the latter approach is generally preferable.


Issue Date:
Feb 02 1991
Publication Type:
Working or Discussion Paper
Language:
English
Total Pages:
15
Series Statement:
9102




 Record created 2017-09-12, last modified 2017-09-12

Fulltext:
Download fulltext
PDF

Rate this document:

Rate this document:
1
2
3
 
(Not yet reviewed)