Characterizing Homeland Security Risk: A Principal Component Analysis of 10 Hazards

Abstract This research reduces the number of attributes to describe the varied risks in the homeland security domain using Principal Component Analysis (PCA). Reducing the dimensions of homeland security risks to a smaller, more manageable set of characteristics can enhance policy-making processes, especially given the broad spectrum of consequence and non-consequence attributes and the varying frequencies of such events. PCA was used to reduce the larger set of risk attributes to five – representing health, economic, societal, dread and the unknown – or two – representing consequence and perceptual characteristics. While the five-component approach describes the data more completely, the two-component approach also explains a large proportion of the variance of the dataset but does so with a lower cognitive load. Either approach can provide composite variables that describe the homeland security risks in a more efficient fashion without losing excessive information on the risks.