Validating Social Vulnerability Index Models: A Multi-Faceted Comparison Analysis Across Scales, Indicators, Model Structures, and Expert Stakeholder Insights
Open Access DepositedProactive and equitable planning for natural hazards is essential, as these events can cause mass destruction and impact livelihoods, disproportionately burdening historically underserved and marginalized communities. To support equitable decision-making in hazard preparedness, organizations can utilize tools like a social vulnerability index (SVI), developed to identify vulnerable populations to ensure that those with inherent social inequities are considered in planning. However, SVI construction involves methodological choices that introduce epistemic uncertainty, potentially impacting stakeholder decisions. While significant efforts have been made to understand how construction processes affect index results through various validation approaches, further validation of SVIs is necessary to ensure that these indexes effectively capture vulnerability.This dissertation contributes to these efforts by exploring the validity of SVIs through internal and user validation approaches. The first two studies focus on the underexplored role of spatial scale—areal units and geographic boundaries—in SVI construction. Scale remains debated among researchers regarding its impact on index outcomes. Chapter Two examines how variations across scalar properties, along with indicator selection, affect SVI rankings for two indexes: the Centers for Disease Control Social Vulnerability Index and the University of South Carolina’s Hazards Vulnerability and Resilience Institute Social Vulnerability Index. These are compared across three model structures—hierarchical with z-score standardization, hierarchical with percentile ranking normalization, and inductive with z-score standardization—using an uncertainty and sensitivity analysis. This study finds scale selection a notable model stage in driving variability in SVI outcomes, particularly when considered alongside indicator selection. However, the impact of scale selection varies across model structures, with the inductive model presenting less robustness and greater sensitivity to changes in scale and indicator selection compared to hierarchical models. Chapter Three builds on these findings by examining how scalar components influence indicators within SVI model structures. While indicator selection is a recognized critical stage in SVI construction, less is known about how indicators differ across various scalar elements. This chapter investigates these interactions by analyzing their effects on vulnerability rankings and spatial patterns using geospatial and hotspot mapping and identifying how individual indicators across the indexes vary across scales in highly vulnerable areas. This study finds that scale and indicator selection do more than shift spatial patterns of vulnerability—they actively reshape the role of individual indicators in different contexts. These insights reinforce the importance of considering scale-indicator interactions when constructing SVIs. The last study, Chapter 4, integrates a user validation assessment exploring expert perspectives through a survey-based participatory geographic information systems (GIS) mapping approach. This chapter compares expert-delineated areas facing vulnerability and environmental justice (EJ) concerns to SVIs and EJ indexes (EJIs) through confusion matrixes and spatial assessments. This study explores how expert perspectives align or diverge from existing indexes and across respondents. Findings highlight the variability in experts' spatially delineated areas of vulnerability and injustices and discrepancies between experts' assessments and SVIs and EJIs spatial patterns. This study uncovers instances where these indexes fail to identify areas of concern, contributing to ongoing discussions aimed at improving index development to better reflect local realities.
- All rights reserved
Notice to Authors
If you are the author of this work and you have any questions about the information on this page, please use the Contact form to get in touch with us.