|
Activity Number:
|
310
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Tuesday, August 5, 2008 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Business and Economics Statistics Section
|
| Abstract - #300881 |
|
Title:
|
Bartlett-Type Corrections for Score Test Statistics on Gaussian ARMA Models
|
|
Author(s):
|
Bernardo M. Lagos*+ and Pedro A. Morettin
|
|
Companies:
|
University of Concepción and Institute of Marine Science, University of São Paulo
|
|
Address:
|
Department of Statistics, Esteban Iturra s/n, Cocepción, International, 160-C, Chile
|
|
Keywords:
|
ARMA models ; Score statistics ; Bartlett-type correction
|
|
Abstract:
|
In this article we compute the corrected score statistic according to Bartlett-type proposed by Cordeiro and Ferrari (1991), used to test the null hypothesis concerning some parameter vector of interest of an ARMA model, assumed to be Gaussian, stationary and invertible. We also consider the situations where nuisance parameters are present, useful in many practical cases. The formulas are written in matrix form, appropriate for the use of symbolic or numerical languages. The performance of the tests is assessed through an estimate of the rejection probability error and an estimate of the power, via Monte Carlo simulation study. Many cases of Bartlett-type Score test is better than the original version. Particularly cases where the parameters are close to the critical regions, not stationary model.
|
- The address information is for the authors that have a + after their name.
- Authors who are presenting talks have a * after their name.
Back to the full JSM 2008 program |
