|
Activity Number:
|
113
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, August 3, 2009 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Statistical Computing
|
| Abstract - #303580 |
|
Title:
|
Tests for Comparing Several Poisson Means
|
|
Author(s):
|
Jie Peng*+ and Kalimuthu Krishnamoorthy
|
|
Companies:
|
St. Ambrose University and University of Louisiana-Lafayette
|
|
Address:
|
518 West Locust street, Davenport, IA, 52803,
|
|
Keywords:
|
Exact conditional test ; Power ; Type I error ; Variance stabilizing transformation
|
|
Abstract:
|
In this talk, we consider the problem of testing equality of several Poisson means. Two tests, one based on the conditional distribution of the sample counts, and another based on the bootstrap samples from estimated Poisson distributions (PB test) are proposed. These tests are compared with some asymptotic tests such as the likelihood ratio test, chi-square test, Neyman-Scott test and the Brown-Zhao test. Our comparison studies indicate that the asymptotic tests perform satisfactorily only when the Poisson means are large. The conditional test and the PB test work satisfactorily even for small samples and/or small values of Poisson mean. The PB test offers more power than other tests in some situations. The tests are illustrated using some numerical examples.
|
