Abstract
Homelessness and housing research has a detection problem, in which understanding housing issues (e.g., substandard housing) is plagued by problems like undercounting and incomplete data. To partially compensate for these pitfalls, the current research engages in an exercise aiming to sharpen the statistical approach to homelessness and housing research. The macro-level sample included Florida’s 27 Continuums of Care geographies (composed of one or more Florida counties). The dependent variables were sheltered homelessness, energy-based substandard housing, and plumbing-based substandard housing. The independent variables were misdemeanor partner violence rates, felony partner violence rates, high school non-completion rates, urbanicity, and population burdens of racism and ethnocentrism. Radio diagrams were constructed to visualize three continuous statistical indicators from phi coefficients: effect size point estimate, p-value, and effect size confidence interval range. This strategy contextualizes findings without binary constraints on interpretation.
Recommended Citation
Montanez, Julio; Donley, Amy; and Reiss, Jacquelyn
(2026)
"Statistical Triangulation: Weighing Multiple Statistical Tools in Macro-Level Housing Research,"
Journal of Applied Disciplines: Vol. 4:
Iss.
1, Article 3.
Available at:
https://opus.govst.edu/jad/vol4/iss1/3