Monday, 29 January 2024
Hall E (The Baltimore Convention Center)
When extreme weather — tornadoes, hurricanes, winter storms — is in the forecast, what information do members of the public want, and when do they want it? While recent work has explored these questions in a survey setting, this work aims to evaluate public information preferences in a social media setting. To this end, we collected and categorized public responses to posts from National Weather Service offices during 35 severe weather events between 2019 and 2021. Undergraduate research assistants from the Institute for Public Policy Research and Analysis Code Lab assisted by identifying relevant posts from the week leading up to, day of, and day following these events. Data was collected from Twitter and Facebook using both automated and manual collection methods. This work will overview the manual Facebook data collection process and initial results of public information preferences across severe weather events. As the nature of Twitter/X is rapidly evolving and its features are changing, it will be important for researchers to develop other ways to study how the NWS communicates on social media, the agency’s most direct way to communicate with the public. In addition to answering the research questions laid out above, this work aims to develop a systematic method for data collection from NWS Facebook accounts.

