Collaborative Research: Online Hazard Communication in the Terse Regime: Measurement, Modeling, and Dynamics

Overview: Social media use by emergency management organizations has surged in the past 5 years, reflecting both the increasing use of social media technologies by the general population and the potential for social media to facilitate key emergency response tasks. While "social media" encompasses a diverse array of technologies and practices, particular interest has converged on the practices and processes of short text message exchange - what we here call the "terse regime." Though social media are an increasingly important fixture of the emergency management landscape, the study of online communication in response to extreme events is still quite young and theory development has lagged. In this project we propose to develop novel measurement techniques and formal theoretical models for understanding the dynamics of online communicative behavior in the terse regime. The core questions to be addressed by this research include the following: 1. What governs the allocation of attention of the online public to specificc organizations and messages during disaster, and how does this affect retransmission of terse messages? 2. What governs the dynamics of organizations' online terse-regime communications, and how does their behavior evolve in response to hazard stimuli, public behavior, or their own interactions? 3. How does informal online response vary across time and space in terms of both message production and message transmission? 4. How can we better recognize, detect, and measure online communicative processes in response to hazard events? To answer these questions, we pursue several linked activities, including: collection of a systematic, baseline controlled longitudinal "backbone sample" of hazard-related communication from a prominent micro-blogging site over a three-year period, supplemented by demographic and other information on hazards, official warnings and alerts, and the impacted populations; collection of detailed data on online communication by and interactions among organizations involved in emergency response activities in a major US metropolitan area; development of agent-based and other theory-driven models for terse message transmission, organizational interaction and communication behavior, and allocation of attention to official communications online; and development of novel techniques for measuring the public's online response to hazard events and for distinguishing between responses generated by specific social processes or by particular sub-populations. Intellectual Merit: This research has the potential to fundamentally transform our understanding of and theoretical orientation to terse hazard communication processes at a particularly critical time in the disaster life cycle, and revolutionize our methods for the study of organizational and attentional dynamics online. It will also give us the unique opportunity to study the microdynamics of organizational behavior at an early period in the institutionalization of a new technology. Broader Impact: This research has the potential to transform our understanding of online hazard communication in the terse regime, leading to methods and models that are transferable to other terse messaging channels and that can lead to new technology design, messaging strategies, and communication policy. Our findings will provide the empirical basis necessary to inform future decision making on strategies to appropriately utilize social media for warnings and risk communication, to understand networked patterns of communication among the public, and to predict how organizations themselves will act in a distributed environment. Our findings will be communicated to the practitioner community not only by presentations at practitioner-oriented meetings and events, but also by the development (w/partners in San Francisco) of training materials for the use of social media by emergency management agencies.