|
Activity Number:
|
154
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, July 30, 2007 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Biopharmaceutical Section
|
| Abstract - #310322 |
|
Title:
|
Exact Inference for Nominal and Rating Scale Data with Repeated Measurements in Clinical Trials with Parallel Group Design
|
|
Author(s):
|
James Lee*+ and John Dar Shong Hwang
|
|
Companies:
|
Daiichi Sankyo Pharma Development and B.R.S.I.
|
|
Address:
|
399 Thornall St., Edison, NJ, 08837,
|
|
Keywords:
|
Exact inference ; nominal data ; treatment*time interaction ; treatment effect ; conditional distribution ; rating scale
|
|
Abstract:
|
For a clinical trial with two treatment arms having repeated measurements of nominal data as the response, this paper proposes a model-free exact inference on hypotheses commonly encountered in practice. The method is an extension of the binary data case previously reported. Rating scale data is considered as a transformation of the nominal data. Construction and justification of parameters which may be used to represent Treatment by Time Interaction, common Treatment Effect, common Time Effect for the data with 3 categories of response is developed. Existence and actual derivation of a conditional distribution which depends on desired parameters and is devoid of nuisance parameters is illustrated using two repeated time points. Generally the distributions may be simplified to become product of basic binomial or hypergeometric distributions.
|
