Session: Big Data and Prevention Science: Using Data to Inform Decisions in Public Health (Society for Prevention Research 23rd Annual Meeting)

Big data is all the rage in marketing, information technology (IT), and medical decision-making, however, there has been less application of this framework to prevention science. The current presentation provides three examples of big data applications to public health to inform prevention science and practice. All three papers feature a university/practice partnership which was formed to share data and collaborate on analytics, and ultimately leveraging the resulting findings to inform decisions related to the prevention of behavior and/or mental health problems.  The first paper focuses on community-level data captured via geographic information systems (GIS) to examine the link between observational data and alcohol outlet zoning. This project had a particular focus on local establishments licensed to sell alcohol and their impact on youth risk behaviors. Data will be presented that detail the extent to which existing zoning and municipal laws were being enforced and the magnitude of association between local laws and ordinances and negative youth outcomes.  The second paper used data from 26,034 adolescents across 58 Maryland high schools who completed a school climate survey through a collaborative university/state/mental health providers partnership called the Maryland Safe and Supportive Schools Project. This paper has a particular focus on the amount of agreement between students at a school, which provides useful information to inform practitioners’ understanding of norms related to alcohol use.  A generalized variance approach was used to account for the variety of alcohol perceptions within schools. Using a formula derived from Simpson’s Diversity Index, youths’ responses were used to model the probability of two randomly selected students having different perceptions. Multilevel models were then fit to examine perceptions of alcohol norms. Implications of these findings for data-based decision-making are considered. The third paper uses data from the Maryland State Department of Education as a part of their federal Race to the Top Project, which required the integration of data sets to inform data-driven decision-making. A university/practice partnership created a dashboard focused on the use of archival data to inform decisions regarding risk factors for school dropout. The partnership collaborated on the analyses which resulted in cutpoints and the development of graphics to inform both the visualization and interpretation of data by school-based practitioners and state-level policy makers.  Together, these 3 studies highlight the potential utility of a big data framework for leveraging existing data to inform decision-making related to public health and prevention science, through a research/practice partnership.