Individual- and Community-Level Effects on Risk Perception in Cancer Cluster Investigations


*Address correspondence to Craig W. Trumbo, Department of Journalism and Technical Communication, 229C Clark Hall, Colorado State University, Fort Collins, CO 80523, USA; tel: 907-491-2077; fax: 907-491-2908;


This article focuses on the relative influence of individual versus community effects on risk perception. The study is grounded in literature examining how individuals manage information and make risk judgments in the context of suspected environmental cancer threats. We focus on three individual-level perspectives: the psychometric model of risk perception, an adaptation of the heuristic-systematic information processing model, and cancer anxiety. We also evaluate five sets of community-level variables that frame cancer cluster investigations: demographic, epidemiologic, sociologic, etiologic, and pathologic. Data were collected through a mail survey of 30 communities in which cancer cluster investigations were being conducted. Response rates averaged 43%, with 1,111 records in the final data set. Through multilevel modeling and other techniques, the results show that the individual-level model developed in previous work remains a robust description of risk perception in these cases. However, the analysis also shows that the community-level measures neither improve the individual-level model nor offer any substantial explanatory power of their own. We provisionally conclude that, within the context of cancer cluster investigations, risk perception is a phenomenon located in a common psychological dimension that is substantially independent of contextual influences. We also suggest that risk communication efforts in this specific context might successfully draw from a common approach informed more by individual than community factors.