Project Details
During disasters, online activity increases, and the public frequently turns to social media sites such as Facebook and Twitter to relay and gather information. Given the growing importance of social media as a disaster communication tool, it is vital to understand how individuals use, behave, and interpret information on social media sites to better inform policy, guidance, and operations and to ensure that emergency managers, first responders, and policy makers can best optimize how they use these tools.
This research produced several key findings including:
The source from which participants received disaster information impacted the perceived credibility of the information. However, the sources of disaster information alone did not differentially influence participants’ reported likelihood of evacuating nor their reported likelihood of taking any other measured recommended actions.
After exposure to the hypothetical disaster information, across the board participants reported intentions to communicate predominately via interpersonal channels than through organizational media channels. Furthermore, demographics such as gender and age affected how participants responded to the hypothetical disaster. Finally, participants’ involvement with media coverage of the April 15, 2013 Boston Marathon Bombings alone was significantly related to all cognitive, emotional, and intended behavioral responses with the sole exception of “writing a blog post.”
An online field experiment with a nationally representative random sample of 2,015 U.S. residents tested the potential effects of disaster information form and source on three types of outcomes: (1) cognitive responses (e.g., perceived information credibility and attribution of responsibility to the government); (2) emotional responses (e.g., discrete emotions associated with a disaster, such as outrage, shock, and anxiety); and (3) intended behavioral responses (e.g., stated intentions to engage in information seeking, communicating, and taking protective actions related to a disaster). The report also examines potential differences in these responses based on demographic characteristics including gender, age, and race.
After random assignment to one of 12 conditions, participants read a brief narrative that asked them to imagine themselves traveling to San Francisco, checking into a hotel, going to sleep for the night, and waking up to find out about an unfolding disaster involving multiple coordinated terrorist attacks. They were then presented with initial disaster information, which was provided via social media (tweet or Facebook post) or traditional media (website post) from a local government source (San Francisco Mayor’s Office of Emergency Management), national government source (Federal Emergency Management Agency; FEMA), local news media source (San Francisco Chronicle), or national news media source (USA Today). After reading more in-depth disaster information from the same source, participants completed a questionnaire assessing potential cognitive, affective, and behavioral responses to the disaster information.