Date of Award
Spring 2015
Degree Name
Bachelor of Science
Department
Computer Science & Mathematics; College of Arts & Sciences
First Advisor
Edward F. Wolff
Abstract
This paper firstly provides a conceptual explanation of Hierarchical Linear Modeling. It provides rationale for its development, its applications as well as some of its limitations. It also delves briefly into the topic’s relatively short history. It is the hope of the author that this paper will motivate the reader to learn more about the method and consider its use when appropriate in their practice. Secondarily it provides a Monte Carlo Simulation made by the Author to determine how multilevel data influences the true type I error rate of one sample t-tests under a range of specific conditions. It is work the author believes is novel and could very well be built upon by future students. Lastly the author also hopes this research will help raise awareness of how multilevel data can result in increased type I error in their work and how HLM is one means of dealing with such an issue.
Recommended Citation
Loeb, Franklin F. Jr., "Level Up to HLM: A Way to Handle Multilevel Data" (2015). Capstone Showcase. 1.
https://scholarworks.arcadia.edu/showcase/2026/comp_sci_math/1
Level Up to HLM: A Way to Handle Multilevel Data
This paper firstly provides a conceptual explanation of Hierarchical Linear Modeling. It provides rationale for its development, its applications as well as some of its limitations. It also delves briefly into the topic’s relatively short history. It is the hope of the author that this paper will motivate the reader to learn more about the method and consider its use when appropriate in their practice. Secondarily it provides a Monte Carlo Simulation made by the Author to determine how multilevel data influences the true type I error rate of one sample t-tests under a range of specific conditions. It is work the author believes is novel and could very well be built upon by future students. Lastly the author also hopes this research will help raise awareness of how multilevel data can result in increased type I error in their work and how HLM is one means of dealing with such an issue.