Title
|
Room
|
* ! IRT Modeling: Applications to Health Policy
|
H-Madison
|
Date / Time
|
Sponsor
|
Type
|
08/07/2001
8:30 AM
-
10:20 AM
|
Section on Health Policy Statistics*, ENAR
|
Invited
|
Organizer:
|
George Papandonatos, Brown Universtiy
|
Chair:
|
Jason Roy, Brown University
|
Discussant:
|
9:50 AM - Jeff Douglas, University of Wisconsin Medical School
|
Floor Discussion
|
10:05 AM
|
Description
Clinical trial data often become available in the form of multiple-item subscales, with each item scored on a likert scale. Most statistical methods for the analysis of such data do not explicitly consider the psychometric features and properties of interest in the original scale development. The aim of this session is to present item response modeling approaches to analyzing longitudinal multiple-item data measured on ordinal scales in a way that maintains fidelity to the psychometrical foundations upon which the original instruments were built.
|