|
Activity Number:
|
228
|
|
Type:
|
Invited
|
|
Date/Time:
|
Tuesday, July 31, 2007 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
SSC
|
| Abstract - #307928 |
|
Title:
|
Improved Likelihood Inference for Vector Parameters with Discrete Data
|
|
Author(s):
|
Nicola Sartori*+
|
|
Companies:
|
University of Venice
|
|
Address:
|
San Giobbe, Cannaregio 873, Venezia, 30121, Italy
|
|
Keywords:
|
Contingency table ; Discrete data ; Likelihood asymptotics ; P-value ; Vector parameter
|
|
Abstract:
|
We consider the testing of compound hypotheses or vector component parameters in contingency tables and other discrete data models. In general models, asymptotic inference from the observed likelihood provides just first order accuracy. For the discrete data case Davison, Fraser and Reid (JRSSB, 2006) develop second order methods for assessing a scalar component parameter, which makes use of a data based reparameterization of exponential type. We extend this approach to obtain second order p-values for the assessment of compound or vector hypotheses. This makes use of marginal likelihoods and directional departure measures. The approach is illustrated on examples.
|
- 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 2007 program |
