Date of Award

Spring 5-3-2024

Level of Access Assigned by Author

Open-Access Thesis

Degree Name

Master of Science (MS)

Department

Spatial Information Science and Engineering

Advisor

Matthew P. Dube

Second Committee Member

Kristen Gleason

Third Committee Member

Sarah Walton

Abstract

This thesis explores rural housing insecurity through Swope and Hernandez’s (2019) 4 C’s of housing insecurity in rural areas. Little attention has been paid to rural areas in the conversation on housing insecurity and houselessness (Gleason et al., 2021). To facilitate further discussion on this understudied issue, this exploratory study used unsupervised machine learning to group census tracts into risk levels across seven axes of data from the American Community Survey. The axes were based on housing insecurity factors found in the literature. K-medoids clustering is used to group census tracts into high, medium, and low risk of housing insecurity for each axes. Multinomial logistic regression was used to determine variation between U.S. states based on how well state risk levels could be predicted with the national dataset. Furthermore, spatial autocorrelation analysis was employed to gauge the extent of spatial clustering within the identified risk levels and housing insecurity factors. The results indicate that many rural census tracts have a medium risk of housing insecurity, and the risk levels are hard to predict. The spatial autocorrelation results show that the housing insecurity variables were not as highly spatially clustered as expected.

Download

Share