|
Activity Number:
|
290
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 14, 2002 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Biometrics Section*
|
| Abstract - #301322 |
|
Title:
|
Comparability of Joinpoint Models
|
|
Author(s):
|
Hyune-Ju Kim*+ and Michael Fay and Eric Feuer and Mike Barrett
|
|
Affiliation(s):
|
Syracuse University and National Cancer Institute and National Cancer Institute and National Cancer Institute
|
|
Address:
|
215 Carnegie Building, Syracuse, New York, 13244, U.S.A.
|
|
Keywords:
|
joinpoint ; comparability ; permutation ; P-value
|
|
Abstract:
|
A regression model with continuous linear phases is often used to describe changes in trend data. Kim et al. (2000, Statistics in Medicine) proposed a permutation procedure to determine the number of unknown joinpoints where each p-value is estimated by using Monte Carlo methods, and the overall asymptotic significance level is maintained through a Bonferroni correction. In this paper, we propose a test to determine the comparability of two or more joinpoint models. Our focus will be on testing if changes occur at the same time points and then if the rates of changes are the same among the groups. We discuss a permutation procedure as well as the classical F-approximation in estimating the P-value of the test in the context of a two-group comparison. We then generalize the idea to a multi-group comparison by using a sequentially rejective Bonferroni procedure proposed by Holm (1979). The performances of these tests are studied via simulations, and some applications to cancer rates will be discussed.
|
