A dataset integrating water-related public health, social media, census, and administrative data in the United States

This project presents a dataset that is
assembled from multiple sources between 2013 and 2017, including contaminants
in drinking water, cancer incidence rates, public perception of the
relationship between water contaminants and cancer on Twitter, and census data
covering the population living in the United States. The units of analysis are 3,219 counties and
33,144 zip codes. The users of this
dataset can address model-driven questions regarding water contaminants and
cancer incidence rates in a geographic area, as well as how water contaminant
levels and cancer incidence rates are directly and indirectly influenced by local
population, economic, and social characteristics.