This article examines the heuristic value of a model of risk communication outlined by Peter Sandman. It tests and expands the proposed constructs and seeks to establish a measurement model. Results in the first laboratory study indicated congruence between the model and data, while the second demonstrated the capacity for manipulation of the constructs. A field study then extended the measurement model in both scope and usefulness by demonstrating

its utility in an applied setting. Descriptive analyses indicate differences in perceptions of risk on the basis of sex and race. Implications for the use of the model are discussed.