Dissertation
A hierarchical Bayesian model to group geographical regions into health categories
University of Iowa
Doctor of Philosophy (PhD), University of Iowa
Autumn 2024
DOI: 10.25820/etd.007723
Abstract
At the beginning of each decade since 1980, the Healthy People national objectives have been launched to provide a set of priorities for communities, individuals, and organizations to improve the health and well-being of the American population. Understanding community health can aid in identifying potential factors that contribute to better health, as well as potential factors that contribute towards poorer health and need improvement. In this work, a Bayesian hierarchical modeling framework is used to develop a model for assessing region-level health, utilizing cluster analysis methods to group together regions that are most similar in terms of overall health. The proposed cluster-approach methodology is assessed alongside a latent classification model, also implemented in a Bayesian hierarchical framework, to highlight advantages of the proposed cluster-approach model. A data analysis is performed separately for Iowa and Texas to demonstrate how the cluster analysis model and the latent classification model identify health categories of counties within each state. The results of each of these models is compared to analyses conducted annually by the County Health Rankings & Roadmaps program sponsored by the University of Wisconsin Population Health institute.
Details
- Title: Subtitle
- A hierarchical Bayesian model to group geographical regions into health categories
- Creators
- Lauren Mudd
- Contributors
- Jacob J Oleson (Advisor)Emine O Bayman (Committee Member)Brian J Smith (Committee Member)Joseph E Cavanaugh (Committee Member)
- Resource Type
- Dissertation
- Degree Awarded
- Doctor of Philosophy (PhD), University of Iowa
- Degree in
- Biostatistics
- Date degree season
- Autumn 2024
- DOI
- 10.25820/etd.007723
- Publisher
- University of Iowa
- Number of pages
- xiv, 138 pages
- Copyright
- Copyright 2024 Lauren Mudd
- Language
- English
- Date submitted
- 12/09/2024
- Description illustrations
- illustrations (some color), maps (some color)
- Description bibliographic
- Includes bibliographical references (pages 119-124).
- Public Abstract (ETD)
At the beginning of each decade since 1980, the Healthy People national objectives have been launched to provide a set of priorities for communities, individuals, and organizations to improve the health and well-being of the American population. Understanding community health can aid in identifying potential factors that contribute to better health, as well as potential factors that contribute towards poorer health and need improvement.
In this dissertation, a method is developed to allow researchers to identify groups of regions that are similar in terms of health. The new proposed method to identify groups of regions is compared alongside a more general, commonly used approach to highlight the advantages of the new proposed method. The states of Iowa and Texas are presented as separate examples to demonstrate how each of the modeling approaches are used to identify groups of counties that are similar in terms of health, based on the type of health data used in the model.
Finally, the results of the new proposed modeling approach for Iowa and Texas are each compared to the associated analysis that is performed annually and made publicly available for reference by the County Health Rankings & Roadmaps program.
- Academic Unit
- Biostatistics
- Record Identifier
- 9984774548102771
Metrics
2 File views/ downloads
16 Record Views