Date of Award

6-15-2024

Document Type

Dissertation

Degree Name

Ph.D.

Organizational Unit

Morgridge College of Education, Research Methods and Information Science, Research Methods and Statistics

First Advisor

Yixiao Dong

Second Advisor

Peter Organisciak

Third Advisor

Nick Cutforth

Fourth Advisor

Stacey Freedenthal

Keywords

Ant colony optimization, Crisis services, Educational measurement, Intraclass correlation coefficient, Item response theory, Suicide prevention

Abstract

Ant Colony Optimization (ACO) is a flexible algorithm designed to solve complex combinatorial problems. While the method was derived from the behavior of ants by researchers in the field of computer science, its application to solving complex combinatorial problems is widespread in a growing number of fields in behavioral science, including psychometrics. Over the last two decades, psychometricians have adapted ACO to measurement model specification problems with the intention of generating measurement models that express measurement model fit and reliability within the standards of what is considered acceptable. Additionally, psychometricians have used ACO to generate shortened versions of existing measures while preserving the integrity of other psychometric properties (e.g., model fit, validity, and reliability). The current study sought to extend the utility of ACO by incorporating the intra-class correlation coefficient (ICC) as an optimization criterion in seeking an optimal (or near-optimal) measurement model solution. The introduction of ICC to ACO procedures is intended to address data that features multiple observations of the same targets from multiple raters and was the first of its kind. The study featured a new measure designed to capture the quality of telephonically delivery crisis intervention services, the Multidimensional Crisis Monitoring Form (MCMF-3). Scores from the ACO-derived measurement models were further tested to examine the relationship between the quality of universal screening questions asked by crisis workers, and the subsequent quality of their approach to the crisis management process as captured by the MCMF-3. Implications for future research and use of AI-driven techniques in behavioral health measurement practices are discussed in response to the study results.

Copyright Date

6-2024

Copyright Statement / License for Reuse

All Rights Reserved
All Rights Reserved.

Publication Statement

Copyright is held by the author. User is responsible for all copyright compliance.

Rights Holder

Mark Leveling

Provenance

Received from ProQuest

File Format

application/pdf

Language

English (eng)

Extent

138 pgs

File Size

1.0 MB

Available for download on Friday, January 31, 2025



Share

COinS