Journal of Medical Statistics and Informatics

Journal of Medical Statistics and Informatics

ISSN 2053-7662
Methodology

Generalizability Theory:Demonstrating the Process and its Utility with EEG Measurements

Adrienne Kline1,2*, Theresa Kline3, Daniel Pittman4, Bradley Goodyear4 and Janet Ronsky1,5

*Correspondence: Adrienne Kline askline1@gmail.com

1. Department of Biomedical Engineering, University of Calgary, Calgary, AB, Canada.

Author Affiliations

2. Cumming School of Medicine, University of Calgary, Calgary, AB, Canada.

3. Department of Psychology, University of Calgary, Calgary, AB, Canada.

4. Department of Radiology and Hotchkiss Brain Institute, University of Calgary, Calgary, AB, Canada.

5. Department of Mechanical and Manufacturing Engineering, University of Calgary, Calgary, AB, Canada.

Abstract

Background: The purpose of this study is to demonstrate the utility of generalizability theory in assessing the reliability of EEG data. Generalizability theory and its relevance for measurement are described, followed by the steps to consider and decisions to be made when conducting a generalizability analysis.

Methods: Using an actual data set, how to conduct generalizability and decision analyses using IBM ® SPSS ® are outlined. Specifically, the beta frequency data collected from the C1 electrode from 16 participants across 60 trials and 2 time periods as they moved their right leg in a stepping motion are used for the demonstration. Data analysis decisions such as number and type of facet, relative versus absolute G-coefficients, follow-up decision-study information, data set-up and analytic commands are all described in detail.

Results: Outputs from the analyses are discussed in terms of their meaning and implications. These include the reliability of each facet, variance accounted for and best next steps for improving data reliability.

Conclusion: Advantages of using the generalizability model are discussed as well as suggestions of various generalizability programs available for use.

Keywords: Generalizability, reliability, EEG

ISSN 2053-7662
Volume 8
Abstract Download