Investigating Heat Risk Messaging Using Social Media Studies and a Survey Experiment

Extreme heat causes hundreds of deaths each year in the United States even though cost-effective protective measures are available. Heat warning messages sent by government agencies have the potential to reduce the negative impacts by motivating people to take protective actions. To help reach the potential, this dissertation examined the content of warning messages and public responses to warning messages in the US. This research analyzed three kinds of data: 1) heat warning messages posted on Twitter, 2) public comments on heat warning messages posted on Facebook, and 3) experimental results collected using an online survey. Results show that, for heat warning messages posted on Twitter, most messages mentioned temperatures and/or Heat Index. Half of messages mentioned heat-safety tips. Less than one-third of messages mentioned heat-health impacts and people’s vulnerability (who is at risk and/or which behavior is at risk). For these four types of mentions, heat warning messages that mentioned more types were retweeted more frequently. In addition, compared to listing specific vulnerable subgroups such as older adults, a statement that “anyone can be at risk” appears to be more effective in making heat warning messages personally relevant to the public. The research also shows that Facebook comments on heat warning messages can suggest people’s needs for risk messaging. The findings can inform researchers and practitioners of how to better communicate risks in the context of extreme heat and other natural hazards