Department of Biostatistics Seminar/Workshop Series

Choosing the Best Items: An objective multi-model approach to applied psychometrics

Warren Lambert, PhD

Vanderbilt Kennedy Center & Peabody College

Wednesday, January 16, 2008, 1:30-2:30pm, MRBIII Conference Room 1220

Intended Audience: Persons interested in applied statistics, statistical theory, epidemiology, health services research, clinical trials methodology, statistical computing, statistical graphics, R users or potential users

Three popular approaches to psychometrics are classical test theory (CTT, the oldest), confirmatory factor analysis (CFA), and Item Response Theory (IRT). These different approaches can be viewed as complementary models with differing strengths and limitations. In the construction of a test or index, each item candidate can be evaluated by objective statistical criteria from all three models. While this approach sometimes reveals inconsistencies, more often poor items fail several criteria, making it obvious which items need to be dropped or revised. This approach will be illustrated with tests recently developed at Peabody College and Vanderbilt Kennedy Center.

Presenter Information