When it comes to improving public health, social media can provide a valuable source of data and insights, but in order for it to become a valuable diagnostic tool, we must first develop effective methods for collecting and interpreting that data.
Ann Marie White, Ed.D., a mental health researcher at the University of Rochester, and Melanie Funchess, the director of the Mental Health Association of Rochester, teamed up a decade ago to bring violence prevention research and community action together. Their current work, which focuses on using Twitter data to prevent violence and suicide, is grounded in their previous research identifying and using informal community “helping networks” to enhance mental wellness in local communities.
Researchers use Twitter data to prevent violence and suicide through community-based helping networks.
Their recent behavioral and social sciences research talk, “To Tweet or Not to Tweet: Community-Based Participatory Research Approaches to Advance Wellness and Violence Prevention via Social Media,” highlighted their progress in developing community-based helping networks that take advantage of social media tools to improve public health in their neighborhoods.
Natural Helpers and Informal Helping Networks
When White and Funchess began their work on informal helping networks, they already knew that rates of suicide and violence are affected by the interaction of many risk and resilience factors among individuals, within relationships, and throughout communities. However, according to the World Health Organization’s Commission on Social Determinants of Health and the Centers for Disease Control and Prevention, community support, connectedness, and cohesion provide robust protection against many types of violence. So, White and Funchess decided to focus their research efforts on those aspects.
They started by conducting research in a particular neighborhood that, according to statistics, had high levels of crime among the low-income residents. However, as they spoke to the residents, White and Funchess found that they talked not about crime, but instead about how neighbors speak to neighbors, take care of their houses and those of others, and tend to gardens. The discrepancy between the community’s description in academic reports and by its own residents was striking.
Community support, connectedness, and cohesion provide robust protection against many types of violence.
White and Funchess found that certain people in the neighborhood played an informal role in creating and maintaining this sense of community. They characterized these “natural helpers” as having their finger on the pulse of their neighborhood by actively listening, building rapport, sharing life experiences, caring for the neighborhood, and connecting with and being respected by young people.
Although residents saw natural helpers as influential members of their community, the reach of natural helpers is limited to their immediate communities; furthermore, they may lack the resources, knowledge, or support they need to effectively address all of the needs of their community. Therefore, White and Funchess set up a natural helper learning collaborative, which hosted workshops on topics of interest to natural helpers, such as neighborhood beautification, defusing conflict, and building respectful relationships with culturally diverse neighbors. The collaborative also allowed the exchange of ideas and the building of social networks among natural helpers. White and Funchess are preparing a paper that puts this research into the context of violence prevention.
Identifying Natural Helping Networks in Social Media
The initial phase of the collaborative quickly became self-sustaining within the local communities. As it did, the collaborative’s participants decided to extend its networks into social media, given its growing prevalence. However, they soon realized that they needed help determining how best to do that.
To help them find the best approach, White and Funchess captured over 7 million tweets from 85,000 users; they are currently in the process of analyzing them to identify the social-media equivalents of informal community helping networks and natural helpers. White and Funchess explained that compared to average Twitter exchanges, these digital networks are characterized by more expressions of gratefulness, more prosocial and positive emotional exchanges, fewer mentions of work, and a greater focus on mental wellness.
White and Funchess’ research is an example of community-based participatory research. White and Funchess realized early on the importance of including community members’ input in study design. For example, if the researchers were trying to capture tweets that expressed thanks, they might write code to capture phrases like “thanks” and “thank you.” However, without input from the community, they might have missed related phrases like “props” and “big up.”
Because the tools that are developed from this research are ultimately meant to aid communities in creating mental wellness for themselves, the research questions are also directly informed by needs from the public. For example, one community member thought it would be helpful to have a way to identify drunk tweeting, and as a result, the research team is working on developing a tool for that.
White and Funchess hope that big data analysis can be automated to identify and characterize helping climates and larger social factors that influence health. They also want to extend their work into other social media platforms and automate searches for other aspects of social determinants of health.
ADDITIONAL RESOURCES
Ann Marie White, Ed.D. University of Rochester faculty page
Mental Health Association of Rochester
Photo Credit: Shutterstock/Oleksiy Mark
