Modeling High-Stakes Consumer Decisions in Repeated Contexts : The Problem of Mammography Adherence Following False Alarm Test Results

We thank Emily Conant, M.D. for facilitating access to the mammography clinic of the Hospital at the University of Pennsylvania and for verifying the accuracy of our experimental materials. We thank for invaluable help in data collection. Abstract Consumers often have to decide whether to acquire information in stressful, high-stakes domains. We study women in mammography waiting rooms to test how the information-gathering experience influences intentions regarding their next scheduled mammogram. In particular, we investigate how a patient's " false alarm " result on one occasion (i.e., a mammogram's indication that cancer is present when a 'more accurate' follow-up test reveals it is not) affects willingness to get a regularly scheduled mammogram the next time it is due. We adapt the normative value of information framework by hypothesizing that stress may moderate reactions to a false positive test result. Study 1 demonstrates that a false alarm can delay planned testing adherence but only when that false alarm test result has no implications about the likelihood of future cancer. For patients receiving a false alarm result, information providing either problem-focused coping support by stressing the patient's control or emotion-focused coping support by stressing the frequency of false alarm results mitigates negative future adherence effects. These reactions to false alarm results are moderated by stress, a finding that is developed in Study 2 where we show that stress alters patients' assessments of the utilities associated with testing outcomes. Study 3 shows that, in the presence of a false alarm result, women with actual false alarm history who experience stress are more likely to delay their next mammogram whereas women without actual false alarm history who experience stress are less likely to delay their next mammogram. Stress 2 Consumer decisions vary considerably in the magnitude of their potential impact. For instance, the impact of choosing the wrong brand of a consumer package good is likely to be small, while the impact of failing to obtain information on cancer warning signs could literally be of life or death proportions. Clearly, consequential consumer decisions such as those regarding health care, housing, insurance or security involve a substantial proportion of the economy and are likely to occupy a significant proportion of consumers' overall decision efforts. However, despite calls to the contrary), the vast majority of marketing research in modeling consumer choice focuses on decisions with very limited consequences. Consequential decision-making can generate negative emotion or stress, …