Spielman et al Reanalysis

Spielman et al ¹ attempted to verify the validity of the Social Vulnerability Index(SoVI) developed by Cutter et al². Spielman et al. reproduce the main pieces of the original study which involves running a principal component analysis on potential demographic variables. They then select only 11 normalized variables which are added together at equal weights to produce the SoVI. Spielman et al change the study by recalculating the SoVI index at the extent of the FEMA subregions and then individual primary states within those subregions in addition to a nationwide SoVI calculation.
My interpretation of this reproduction is that Spielman et al were attempting to determine if there is a difference between the ranking of counties by SoVI between the national, regional, and state-based model. We made several modifications to this reproduction during our analysis. First, we visualized the nationwide SoVI model using an interactive leaflet map. This allowed us to see the general spatial trends of the social vulnerability model across the US. A more significant modification that we tested was calculating the percentage of variance in SoVI explained by each component of the model. This calculation was not done by in our original reproduction and allowed us to get a sense of which factors in the model were having the largest impact on the model itself. We then weighted the model factors by their percent variance explained such that the factors that explained the least variance had very little impact on the final model output and the factors that explained the most variance had the largest impact on the final model. Through comparison with the original Spielman et al results it is unclear what exactly the impact of weighting by percent variance explained is. This is because the factors that explain very little variance may already have a very slight impact on the model which would cause their relatively reduced weight to have less of an impact.

Link to full report

Link to full repository

1. Spielman, S. E., Tuccillo, J., Folch, D. C., Schweikert, A., Davies, R., Wood, N., & Tate, E. (2020). Evaluating social vulnerability indicators: criteria and their application to the Social Vulnerability Index. Natural Hazards, 100(1), 417–436. https://doi.org/10.1007/s11069-019-03820-z ↩

2. Cutter, S. L., et al. (2003). Social vulnerability to environmental hazards. Social Science Quarterly, pp. 242–261, https://doi.org/10.1111/1540-6237.8402002. ↩