Abstract #300026

The views expressed here are those of the individual authors
and not necessarily those of the ASA or its board, officers, or staff.

Back to main JSM 2002 Program page
JSM 2002 Abstract #300026
Activity Number: 366
Type: Invited
Date/Time: Thursday, August 15, 2002 : 8:30 AM to 10:20 AM
Sponsor: Section on Statistics in Epidemiology*
Abstract - #300026
Title: Bayesian Hierarchical Transformation Models for Receiver Operating Characteristic Curve Analysis
Author(s): Alistair O'Malley*+ and Kelly Zou
Affiliation(s): Harvard Medical School and Harvard Medical School
Address: 180 Longwood Avenue #304A, Boston, Massachusetts, 02115-5899, USA
Keywords: Discriminant analysis ; Regression ; Bayesian analysis ; Receiver operating characteristic (ROC) curve ; Box-Cox transformation ; Hierarchical modeling
Abstract:

Bayesian hierarchical modeling is developed for Receiver Operating Characteristic (ROC) curve analysis based on clustered continuous diagnostic data with covariates, after an appropriate monotone transformation. Because the mean, variance, transformation, and summary ROC measures are all modeled parametrically, a flexible class of hierarchical transformation models is proposed. The choices of the models and transformations are compared. We compare this model to those in logistic regression analysis in which the gold standard is a binary outcome. We also consider the problem of identifying a composite test based on multivariate outcome data, where the two types of models can be combined. Results are illustrated on prostate cancer biopsy data from a multi-center clinical trial.

  • The address information is for the authors that have a + after their name.
  • Authors who are presenting talks have a * after their name.

Back to the full JSM 2002 program

JSM 2002

For information, contact meetings@amstat.org or phone (703) 684-1221.

If you have questions about the Continuing Education program, please contact the Education Department.

Revised March 2002