Evaluating water quality using social media and federal agency data

United States Environmental Protection Agency (USEPA) drinking water violation report is currently one of the most reliable measures of evaluating United States drinking water quality. While states continuously strive to comply with federal water quality standards making this documentation continuously relevant, consumers are likely to perceive water quality through sensory aesthetics or physical and virtual social networks. This research quantifies the relationship between consumer perceptions and government-reported drinking water quality to provide insights to state water managers and policymakers. We evaluated consumer perceptions of tap water using weekly social media data. The online search returned 898,709 mentions and 799,035 posts. Net sentiment, measured as the number of negative posts minus the number of positive posts divided by the number of posts expressing sentiment, was determined and ranged from -100 to 100. Net sentiment was uncorrelated with USEPA weekly water quality violations for most states. Net sentiment was correlated with violations related to arsenic standards (-0.223) and a total number of violations (-0.220) for Washington. For California, net sentiment was correlated with violations related to disinfectants and other organic compounds (-0.295). In many cases, water violations in one city became national news, which eclipsed local water issues circulating on social media.