Building on channel complementarity theory and media-system dependency theory, this study explores the impact of conflict-oriented news coverage of health issues on information seeking online. Using Google search data as a measure of behavior, we demonstrate that controversial news coverage of the U.S. Preventive Services Task Force's November 2009 recommendations for changes in breast cancer screening guidelines strongly predicted the volume of same-day online searches for information about mammograms. We also found that this relationship did not exist 1 year prior to the coverage, during which mammography news coverage did not focus on the guideline controversy, suggesting that the controversy frame may have driven search behavior. We discuss the implications of these results for health communication scholars and practitioners.

Engaging social networks to encourage preventive health behavior offers a supplement to conventional mass media campaigns and yet we do not fully understand the conditions that facilitate or hamper such interpersonal diffusion. One set of factors that should affect the diffusion of health campaign information involves a person's community. Variables describing geographic communities should predict the likelihood of residents accepting campaign invitations to pass along information to friends, family, and others. We investigate two aspects of a community--the availability of community ties and residential stability--as potential influences on diffusion of publicly-funded breast cancer screening in the United States in 2008-2009. In a survey study of 1515 participants living in 91 zip codes across the State of Minnesota, USA, we focus on the extent to which women refer others when given the opportunity to nominate family, friends, and peers to receive free mammograms. We predicted nomination tendency for a particular zip code would be a function of available community ties, measured as religious congregation density in that zip code, and also expected the predictive power of available ties would be greatest in communities with relatively high residential stability (meaning lower turnover in home residence). Results support our hypotheses. Congregation density positively predicted nomination tendency both in bivariate analysis and in Tobit regression models, and was most predictive in zip codes above the median in residential stability. We conclude that having a local infrastructure of social ties available in a community predicts the diffusion of available health care services in that community.

Growing evidence suggests that basic exposure measures, such as recognition-based items, might not operate identically among older and younger adults. We present two studies relevant to this debate. Study 1 provides experimental confirmation of the recognition decline hypothesis, finding an interaction between age and exposure in predicting recognition memory for an advertisement related to global warming. Study 2 assesses television news project evaluation data to explore whether verbatim detail recognition difficulty explains Study 1 results. The two studies provide complementary evidence, not only illustrating recognition decline among the elderly but also providing careful control of exposure in Study 1 and the use of multiple messages, realistic viewing scenarios, and free recall data in Study 2. Taken together, the studies offer a cautionary tale for campaign evaluators. On a broader theoretical level, the results suggest a fruitful path for communication research focused on the nuanced and potentially critical moderating role of age.