Facebook’s Data for Good Program Debuts 3 New Types of Maps to Help Fight Covid-19

Facebook’s Data for Good program made three new types of disease prevention maps available to researchers and nonprofits to help them arrive at better-informed forecasts and be more targeted with their protective measures in the battle against Covid-19. Facebook head of health Kang-Xing Jin and Data for Good policy lead Laura McGorman detailed the new offerings in a Newsroom post, writing, “Our disease prevention maps are aggregated sets of information that health researchers can use to better understand how population dynamics influence the spread of disease. Researchers and health experts around the world have advocated for more of this information to respond to the pandemic, so today, we’re sharing three new tools.” Co-location maps help illustrate the probability that people in one area will come into contact with people in another area, helping disease modelers predict how the coronavirus might spread. [kickout id="1"] Movement range trends provide a regional-level look at whether people are staying near their homes or venturing elsewhere, enabling researchers to gauge the effectiveness of preventive measures. [kickout id="2"] And the social connectedness index paints a picture of friendships across states and countries, aiding predictions on the spread of the disease, as well as determinations of which areas were hit the hardest. [kickout id="3"] The maps aggregate information from Facebook, but only at the city or country level, so that individuals cannot be identified via this data. In another Data for Good initiative, people in the U.S. began seeing links atop their News Feeds Monday to an optional, off-Facebook survey aimed at helping health researchers monitor and forecast the spread of Covid-19. The survey is being conducted by Carnegie Mellon University Delphi Research Center, and Jin and McGorman said CME Delphi Research will not share individual responses with Facebook, and the social network will not share information about participants with the researchers. They added, “We’ll share a random ID number that CMU will send back to us when someone completes the survey. Then we’ll share a single statistic known as a weight value that doesn’t identify you but helps correct for any sample bias.” [kickout id="4"] Facebook and its trusted organization partners established the Covid-19 mobility data network to provide real-time insights from the social network’s Data for Good tools. Those partners include universities (Harvard School of Public Health, National Tsing Hua University in Taiwan and University of Pavia in Italy), as well as nonprofits and institutions (the Bill & Melinda Gates Foundation, Direct Relief and the World Bank). Institute for Disease Modeling chair of applied math and senior research manager Daniel Klein wrote in the Newsroom post, “Covid-19 has inherent delays that challenge the pace at which we seek to evaluate policy impact toward a measured response. Mobility data from Facebook’s Data for Good program provides a near-real-time view of important correlates of disease transmission. This data, in combination with other sources, allows us to make better models to inform public health decisions.” Harvard T.H. Chan School of Public Health associate director of the Center for Communicable Disease Dynamics Caroline Buckee added, “Measuring the impact of social distancing policies is absolutely critical at this stage, and aggregated data of this kind provides insights that protect individual privacy but are actionable for policymakers and researchers building predictive models.”