Risk Talk and Affect Pulses in Social Media

Principal Investigator: 
Performance Period: 
July 2014 to June 2015
Commercialization Status: 
The primary focus of this research at the University of Kentucky has been to examine the perceptions of risk relayed through social media in response to terrorism signals, both domestically and abroad.  Historical research on risk perception has been conducted using surveys and experiments; this study focuses on public statements made via the social networking site, Twitter, that correspond with characteristics of risk perceptions that have been previously identified.  In particular, this work expands the theoretical work on public perceptions of risk to demonstrate that individuals convey risk perceptions through “risk talk,” or talk that relays subjective statements referencing observations that reflect both thoughts and feelings in real time.   By studying risk perceptions in relation to terrorism risk signals (that is, events that occur and are likely to stimulate public awareness of and talk about the risk), we are able to identify perceptions that are common to terrorism risks, in general, as well as perceptions that differ in relation to selected risk signals. For instance, controllability/preventability; global catastrophe; and expert knowledge dominate risk talk of terrorism, regardless of the risk signal with which it is associated. Some risk perceptions vary in accordance with the risk signal, including its objective characteristics.  Moderate rates of risk talk were found to correspond with different signal events including personal impact (Boston Marathon), increasing (Boston Marathon Manhunt and Capture), severity and observability (the killing of a British soldier), and personal impact, voluntariness, and knowledge of exposure (NSA surveillance). The identification of these public perceptions relayed in real time may enable risk communicators to develop messages that address key concerns about terrorism risk.   Results of this work have led to a new method for coding and analyzing risk talk on Twitter.  This method was first used to analyze terrorism risk signals; it is currently being applied to another important case that garnered significant public attention in 2014, the Ebola outbreak. Early analyses comparing these two threats, terrorism and public health outbreak, demonstrates some key differences among risk talk.  Where much of the chatter about terrorism that was not oriented toward perception focused on rhetoric by defining “what is terrorism and who is a terrorist,” chatter about Ebola includes a sort of gallows humor, possibly indicating a lack of personal efficacy among the public, or a frustration with the overwhelming public response.   The results of this research indicates both the ability to identify public risk perceptions through the analysis of risk talk on social media in response to risk signals as well as the value of examining multiple and varied hazard types to determine the variants that cross hazards and their resultant public risk perceptions.  It provides a theoretical contribution to research on risk perception, and has the potential to offer practical applications in future risk communication efforts.