Predicting radon testing among university employees.

To determine covariates of radon testing behavior, we surveyed by mail a random sample of all Boston University employees (N = 915) six to nine months after they had been informed of the availability of radon testing services through the University's medical center. The response rate was 58%. Analysis suggests blue collar workers were underrepresented within the response rate. Slightly more than half of the respondents (51%) were men. The majority (69%) were under the age of 45. Twenty-seven percent of the respondents (N = 143) had tested their homes for radon. Bivariate analysis revealed important differences between radon testers and nontesters. Testers were 12 times more likely to be home owners than renters (p = 0.00), and were more knowledgeable about radon's characteristics and testing procedures (p = 0.00). Testers were more likely to view radon as a serious problem (p = 0.00), to consider radon testing efficacious (p = 0.00), and to consider themselves susceptible to exposure (p = 0.00). Testers were also less likely to perceive barriers to radon testing. We used logistic regression to compare the usefulness of the Health Belief Model and the Diffusion of Innovations Model in predicting radon testing. We concluded that the knowledge deficits and barriers to radon testing identified in this study should be targeted in radon educational interventions.