Mason Archival Repository Service

Within-Cluster Resampling Methods for Clustered Receiver Operating Characteristic (ROC) Data

Show simple item record

dc.contributor.advisor Tang, Liansheng Miao, Zhuang
dc.creator Miao, Zhuang 2015-02-12T02:59:54Z 2015-02-12T02:59:54Z 2014 en_US
dc.description.abstract The diagnostic studies in which each patient has several diseased and nondiseased observations generate clustered ROC data. Within the same cluster, observations are naturally correlated, and the cluster size may be random. The traditional ROC methods on clustered data can result in a biased variance estimator and subsequently lead to incorrect statistical inference. We introduce resampling methods on clustered ROC data to account for the within-cluster correlation. The within-cluster resampling ROC methods work as follows. First, one observation is randomly selected from each patient/cluster, and then the traditional ROC methods are applied on the resampled data to obtain resampled ROC estimates. These steps are performed many times and the average of resampled ROC estimates is the final estimator. The proposed methods do not require a specific within-cluster correlation structure and yield valid estimators while accounting for the within-cluster correlation. We compare the proposed methods with existing methods in extensive simulation studies and apply the proposed methods to two eye rating examples.
dc.format.extent 149 pages en_US
dc.language.iso en en_US
dc.rights Copyright 2014 Zhuang Miao en_US
dc.subject Statistics en_US
dc.subject Biostatistics en_US
dc.subject Clustered ROC Data en_US
dc.subject ROC Curve en_US
dc.subject Within-Cluster Resampling en_US
dc.title Within-Cluster Resampling Methods for Clustered Receiver Operating Characteristic (ROC) Data en_US
dc.type Dissertation en Doctoral en Statistical Science en George Mason University en

Files in this item

This item appears in the following Collection(s)

Show simple item record

Search MARS


My Account