On Frames, Framing, and the Probability of Framing Effects: The Metacommunicative Role of the Omnipresent Terrorist Actor

In media depictions of terrorist actors and events, a select few organizations appear to be omnipresent. These dominant terrorist actors (i.e., al Qaeda and ISIS) are frequently used to provide a familiar frame of reference for understanding non-dominant actors (e.g., the New IRA and al Shabaab). In this dissertation, I attempt to answer the question “to what effect,” that is, what are the (potentially unintended) effects of the dominant actor framing device on beliefs about the framed organizations? In answering this question, I also seek to provide a framework for better understanding the framing phenomenon and framing effects writ large. I begin by providing a general overview of what is meant by frames and framing, outline the process wherein frames guide construction and interpretation of discourse products and provide evidence for the use of dominant actors as framing devices in media depictions of non-dominant actors. Following from this, I provide the outlines of a general model of information processing and belief formation, which is used to inform design of a probabilistic framing process model. Taken together, these models are used to craft a set of predictions about for whom the dominant actor framing device should have what effect on beliefs about the threat of non-dominant actors to the U.S.Using an online based survey experiment with a quota-based sample of 2,316 adults living in the U.S., I present individuals with a news article depicting the actions of a non-dominant actor (either al Shabaab or the New IRA), manipulated so that 1/3 link the non-dominant actor to ISIS, 1/3 link the non-dominant actor to al Qaeda, and 1/3 do not make explicit reference to any other organizations. I find that the single strongest predictor of beliefs about the threat of non-dominant actors to the U.S. is individuals prior perceptions of the threat from terrorism to the U.S., and I find a negative relation between beliefs about the threat of non-dominant actors and the