期刊
COMPUTER METHODS AND PROGRAMS IN BIOMEDICINE
卷 109, 期 3, 页码 293-304出版社
ELSEVIER IRELAND LTD
DOI: 10.1016/j.cmpb.2012.09.002
关键词
Concordance correlation coefficient; Linear mixed models; Repeated measures; U-statistics; Variance components
类别
资金
- MEDU Grant [CGL2008-05448-C02-02/BOS]
The concordance correlation coefficient is one of the most common approaches used to assess agreement among different observers or instruments when the outcome of interest is a continuous variable. A SAS macro and R package are provided here to estimate the concordance correlation coefficient (CCC) where the design of the data involves repeated measurements by subject and observer. The CCC is estimated using U-statistics (UST) and variance components (VC) approaches. Confidence intervals and standard errors are reported along with the point estimate of the CCC. In the case of the VC approach, the linear mixed model output and variance components estimates are also provided. The performance of each function is shown by means of some examples with real data sets. (c) 2012 Elsevier Ireland Ltd. All rights reserved.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据