Estimation of Cognitive Ability in Longitudinal IRT using Univariate Composite Likelihood: Application to the Emory Healthy Brain Study Restricted; Files Only

Howard, Nicholas (Spring 2024)

Permanent URL: https://etd.library.emory.edu/concern/etds/5q47rq12x?locale=de
Published

Abstract

Longitudinal cognitive test data contain complex structures, particularly for the Item Response Theory model. Thus, available and accepted estimation methods in longitudinal Item Response Theory suffer from high computational intensity to extract the parameter estimates in the item. Univariate composite likelihood employs a marginal approach that broadens accessibility of estimating longitudinally-based parameters for reduced efficiency. Applying this method to the longitudinal item data from the Rey Auditory Verbal Learning Test, Multilingual Naming Test, and Judgment of Line Orientation test demonstrated that accounting for temporal correlation through robust variance estimators did not identify strong deviations away from the specified model. So, another element of the composite likelihood -- such as the latent trait, itself -- may be informative of the longitudinal aspect of the item data.

Table of Contents

1. Introduction 1

2. Background 3

2.1 Cognitive Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

2.2 Item Response Theory . . . . . . . . . . . . . . . . . . . . . . . . . . 5

2.3 Composite Likelihood . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

3. Methods 8

3.1 Missing Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

4. Results 13

5. Discussion 22

Bibliography 25

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
Department
Subfield / Discipline
Degree
Submission
Language
  • English
Research Field
Stichwort
Committee Chair / Thesis Advisor
Committee Members
Zuletzt geändert Preview image embargoed

Primary PDF

Supplemental Files