Functional response models for intraclass correlation coefficients

N. Lu*, T. Chen, P. Wu, D. Gunzler, Hui Zhang, H. He, X. M. Tu

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

6 Scopus citations

Abstract

Intraclass correlation coefficients (ICC) are employed in a wide range of behavioral, biomedical, psychosocial, and health care related research for assessing reliability of continuous outcomes. The linear mixed-effects model (LMM) is the most popular approach for inference about the ICC. However, since LMM is a normal distribution-based model and non-normal data are the norm rather than the exception in most studies, its applications to real study data always beg the question of inference validity. In this paper, we propose a distribution-free alternative to provide robust inference based on the functional response models. We illustrate the performance of the new approach using both real and simulated data.

Original languageEnglish (US)
Pages (from-to)2539-2556
Number of pages18
JournalJournal of Applied Statistics
Volume41
Issue number11
DOIs
StatePublished - Nov 2 2014

Keywords

  • domain-sampling model
  • generalized estimating equations
  • linear mixed-effects model
  • SF-36
  • U-statistics-based generalized estimating equations

ASJC Scopus subject areas

  • Statistics and Probability
  • Statistics, Probability and Uncertainty

Fingerprint Dive into the research topics of 'Functional response models for intraclass correlation coefficients'. Together they form a unique fingerprint.

Cite this