Date of Award

Spring 2021

Document Type

Open Access Dissertation

Department

Educational Studies

First Advisor

Xiaofeng Liu

Abstract

Many methods are used to measure interrater reliability for studies where each target receives ratings by a different set of judges. The purpose of this study is to explore the use of hierarchical linear modeling for estimating interrater reliability using the intraclass correlation coefficient. This study provides a description of how the ICC can be estimated using hierarchical linear modeling, recommends an appropriate non-parametric bootstrapping method, illustrates how both can be implemented to obtain a point estimate and an estimate of bias, and explores the viability of using these statistical tools to obtain such estimates. Results indicated that hierarchical linear modeling and the non-parametric bootstrap method can be used on both continuous and binary data to provide point and bias estimates of interrater reliability.

Share

COinS