Surveillance of a Comparative Set of Homeland Security Risks

The survey was designed to measure public perceptions of homeland security risks, maintaining consistency with a prior nationally representative study (Lundberg and Willis, in press). Participants were introduced to ten homeland security hazards, spanning natural disasters, terrorist events, and major accidents, chosen for their relevance to DHS concerns and diversity in casualty rates, economic impact, and frequency. Each hazard was described in detail, including its effects and mitigation efforts, followed by a structured table outlining 17 risk attributes. Respondents rated their concern on a five-point Likert scale, indicated past exposure, and ranked all ten hazards by perceived risk. Data collection occurred through Amazon Mechanical Turk (MTurk) using Qualtrics, with a compensation model ensuring participant engagement. MTurk’s convenience sampling was validated against a nationally representative survey, showing strong correlations in absolute concern (r = 0.881, p < 0.01) and relative risk rankings (r = 0.758, p < 0.01), supporting the dataset's generalizability.