Comparative Analysis of Four Intraclass Correlation Coefficient Estimation Methods on Continuous Outcomes in a Survey Study Open Access

Abstract

This study aimed to compare methods of estimating the intraclass correlation coefficient (ICC) on continuous outcomes in survey research utilizing cluster sampling. Analysis was performed on the dataset obtained from the Power Up for 30 program, which measured all elementary schools from different school districts in Georgia. Grade level teachers completed surveys regarding students' physical activities. ICCs were calculated for responses from each survey question at the district level and the school level using four methods: the ANOVA approach, the random intercept approach, the GEE approach and the design effect approach. The result indicated that different ICC estimation methods led to various results. ICC estimates obtained using the design effect approach greatly diverged from ICC estimates obtained using the other approaches. For all survey questions, ICC estimates at the school level were greater than that at the district level, suggesting that responses toward all questions exhibited greater consistency among different grade level teachers at the school level, compared to at the district level. Future studies should consider ICC estimation methods that take into account the non-normality of the dataset. Also, average number of respondents should be consistent at the district and the school level.

Table of Contents

1 Introduction 1

2 Methods 4

2.1 Dataset 4

2.2 Statistical Methods 4

2.2.1 The ANOVA Estimator 4

2.2.2 The Random Intercept Model Approach 5

2.2.3 The Generalize Estimating Equations Approach 6

2.2.4 The Design Effect Approach 6

2.3 Computation 7

3 Results 8

3.1 District Level 10

3.2 School Level 10

4 Discussion 12

About this Master's Thesis

Rights statement

• Permission granted by the author to include this thesis or dissertation in this repository. All rights reserved by the author. Please contact the author for information regarding the reproduction and use of this thesis or dissertation.

School	Rollins School of Public Health
Department	Biostatistics and Bioinformatics
Subfield / Discipline	Biostatistics - MPH & MSPH
Degree	MSPH
Submission	Master's Thesis
Language	English
Research Field	Statistics
Keyword	Intraclass Correlation
Committee Chair / Thesis Advisor	Weiss, Paul, Emory University
Committee Members	Chang, Howard, Emory University
Partnering Agencies	Georgia state or local agency (not a health department)

Last modified

Primary PDF

Thumbnail	Title	Date Uploaded	Actions
	Comparative Analysis of Four Intraclass Correlation Coefficient Estimation Methods on Continuous Outcomes in a Survey Study ()	2018-08-28 13:57:37 -0400	Press to Select an action Download

Supplemental Files

Thumbnail	Title	Date Uploaded	Actions