3

Framing and Priming Effects

Exploring Challenges Connected to Cross-Level Approaches in Media Effects Research

Bertram T. Scheufele and Dietram A. Scheufele

Priming and Framing – Modern Paradigms of Media Effects

Media effects models have changed considerably over the course of the last century (for an overview, see McQuail, 2005). Magic bullet or hypodermic needle models, assuming that mass media have a strong and direct effect on a largely monolithic audience, dominated most of the early empirical studies in our field (Katz & Lazarsfeld, 1955).

In the late 1930s and 1940s, Lazarsfeld and his colleagues (e.g., Lazarsfeld, Berelson, & Gaudet, 1948) shifted the focus of media effects research toward a more refined understanding of (a) the audience, and (b) the communication process. Two concepts were particularly important in models of limited media effects: two-step flows of information, and reinforcement (Katz, 1957). The two-step flow idea refers to a process in which highly informed and socially-connected opinion leaders pass on relevant information to lay audiences (for an overview, see Weimann, 1994). Information, in other words, does not reach audiences directly, but is passed on through opinion leaders. Lazarsfeld et al.'s results also showed that most persuasive effects were reinforcement effects rather than attitude change, since lay audiences tended to selectively expose themselves to ideologically consistent messages.

In the mid-1970s, arguments against this limited effects model were raised simultaneously in Europe and the United States. In Germany, Noelle-Neumann offered her spiral of silence model as a new way of modeling strong media effects on people's perceptions of opinion climates and social norms, fueled by various social-psychological traits and processes (Noelle-Neumann, 1974). Her call for a “return to the concept of powerful mass media” (Noelle-Neumann, 1973) was echoed by Gerbner and colleagues who proposed a cultivation model of media effects (Gerbner & Gross, 1974). Cultivation attributes strong mainstreaming effects to commercial media systems, particularly television.

The 1980s and 1990s, finally, marked the transition to what is seen as the current paradigm of “negotiation” models of media effects (McQuail, 2005). The media effects theories subsumed under this umbrella are also often referred to as cognitive models of media effects (D. A. Scheufele, 2000a). They are distinguished from the strong effects paradigms of the 1970s in at least two respects.

First, many of the recent effects models have acknowledged the inherent contingent nature of media effects (Eveland, Jr., 1997), depending on a host of orientations (McLeod, Kosicki, & McLeod, 2009) that audiences use to make sense of and interpret communication content. This “negotiation” of meaning between communicated content and audience interpretation is also from where the paradigm derives its name.

Second, the focus of inquiry in the current approach to media effects has shifted away from Noelle-Neumann's or Gerbner's foci on the social-level outcomes of mass communication, to the interplay between message systems and audience characteristics at different levels of analysis (D. A. Scheufele, 1999). Since the 1980s, our field has therefore seen repeated calls for multi-level approaches to media effects (Pan & McLeod, 1991), bridging micro-, meso-, and macro-levels of analysis.

Some of the most dominant effects models subsumed under the “negotiation” paradigm of media effects are agenda setting, priming, and framing. Agenda setting usually refers to the idea that more frequent or more prominent media coverage also increases our perceptions of how important we think an issue is or how easily we can retrieve it from memory when needed (McCombs, 2004; McCombs & Shaw, 1972, 1993). Priming (at least in the context of this media effect paradigm) extends the agenda-setting model and argues that the issues that have been made most easily retrievable in voters' minds by mass media, will also be the ones that voters take into account when judging the performance or other traits of candidates (Iyengar & Kinder, 1987). Framing, finally, refers to the idea that the mode of presentation of a message (i.e., a message frame), can shape how audiences interpret this message by establishing implicit linkages to audience's pre-existing interpretive schema (D. A. Scheufele, 1999; Tewksbury & D. A. Scheufele, 2009). Before exploring the specific effects models in this paradigm and their differences in greater detail, however, it is necessary to first explore the multi-level assumptions underlying all of them.

A Multi-Level Matrix for Media Framing and Priming Effects

A model of multi-level framing and priming effects is most easily illustrated by structuring levels of analysis around relevant social collectives (B. Scheufele, 2011). Such classifications can follow a formal hierarchy where the level of analysis is determined by the size of the collectives that are being studied (e.g., Lazarsfeld & Menzel, 1961), Smaller collectives, in such a model, are referred to as the micro-level whereas the larger collectives would be referred to as the meso- or macro-levels. Multi-level analyses (e.g., Hayes, 2006; Hox, 2002) rely on similar classifications.

Multi-level classifications can also be made by tying levels to social entities. Here, the individual recipient, for instance, marks the micro-level, social groups or political organizations represent the meso-level, and the political system with its communication structures is located at the macro-level. Classifications tied to social entities have been introduced by scholars in political science (e.g., Eulau, 1977), sociology (e.g., Alexander, Giesen, Münch, & Smelser, 1987; Coleman, 1994) and communication (e.g., McLeod & Pan, 1989; Pan & McLeod, 1991).

Regardless of how distinctions are made between different levels of analysis, the question arises about how to conceptualize (communication) processes linking different levels of analysis. Coleman's (1994) seminal work – often referred to as the “Coleman bathtub” (e.g., Oakes & Kaufman, 2006, p. 11) – distinguishes three “propositions” which are explained by using the example of Weber's (1958) “The Protestant ethic and the spirit of capitalism.” (1) The first proposition links the macro- to the micro-level, and explains how the social situation and the individual's knowledge, motives or values interact by applying a bridge hypothesis. (2) The second proposition links the individual's values with his or her “economic behavior” (both at micro-level). Here, “[i]ndividuals with certain values (referred to in proposition 1) adopt certain kinds of orientations to economic behavior” (Coleman, 1994, p. 8). This is similar to Esser's (1999) rational-choice theory of social behavior, which assumes that subjects weigh different behavioral options and choose the one promising the best outcome. (3) The third proposition links the micro- back to the macro-level and explains how individual economic behavior leads to capitalism as a macro-phenomenon: “Certain orientations to economic behavior (referred to in proposition 2) on the part of individuals help bring about capitalist economic organizations in a society” (Coleman, 1994, p. 8). Esser (1999) calls this the logic of aggregation, which affords rules of aggregation or transformation in order to explain how social phenomena at a higher level emerge from individual behavior at the micro-level.

We are especially interested in what Coleman (1994) calls “Macro-to-Micro and Micro-to-Macro Transitions.” And, in fact, a very similar multi-level matrix based on the Coleman bathtub was provided by Pan and McLeod (1991) for mass communication. They speak of “cross-level linkages” in general, as well as of a “top-down” linkage in terms of the macro-micro-transition and a “bottom-up” linkage regarding the micro-macro-transition. B. Scheufele (2008) applied these ideas more specifically to media effects approaches. Using media coverage of violent antiwar demonstrations as one example, he refers to “bottom-up” media effects if mass media not only influence individuals (e.g., a single protester), but also shape the attributes of meso- or macro-units (e.g., the amount of violence of a demonstration). Conversely, he speaks of “top-down” feedback from a higher to a lower level of social aggregation if, for instance, the amount of violence at a demonstration provokes other protesters to attack the police.

These multi-level linkages are illustrated in Figure 3.1. In a first scenario, for instance, one could imagine two individuals (A and B) who are unfamiliar with antiwar demonstrations. Their willingness to participate (YA and YB) in a demonstration is low, since both the mass media (M) and their family (X1) or friends (X2) tell them that demonstrations are often interrupted by violent protesters. Three other subjects (C, D, and E) may be antiwar activists (X3) and therefore participate in a demonstration against, for instance, “Operation Iraqi Freedom” in 2003. The size of this demonstration (Z) – formally speaking, an attribute of the meso-unit “demonstration” – is therefore N = 3.

A second scenario assumes a different kind of media framing. Here, the mass media consistently frame “Operation Iraqi Freedom” as an unjustified delayed response to 9/11. Subjects A and B, in this scenario, may be mobilized to attend the rally partly based on these media reports, but also partly because their friends and family themselves follow the news and tell them that the invasion of Iraq was not justified. In this second scenario, therefore, media coverage contributes to the size of the demonstration “Operation Iraqi Freedom,” measured by the number of protesters which is now N = 5. This micro-meso transition can be considered a bottom-up media framing effect (up arrows in Figure 3.1). If the course of the demonstration is peaceful, A and B may participate in a future demonstration again. This meso-micro-transition can be considered a top-down feedback (down arrows in Figure 3.1) from the demonstration that A and B experienced (meso-level) to their willingness to participate in a subsequent demonstration (micro-level).

Figure 3.1 Micro-macro and macro-micro links. Adapted from B. Scheufele (2011). Jenseits des Rezipienten. Die Herausforderungen für die Medienwirkungsforschung durch die Mehr-Ebenen-Problematik am Beispiel von fünf Wirkungsbereichen. In T. Quandt & B. Scheufele (Eds.), Ebenen der Kommunikation. Mikro-Meso-Makro-Links in der Kommunikationswissenschaft (p. 118). Wiesbaden, Germany: VS Verlag für Sozialwissenschaften.

A third scenario replaces “subject” by “group” (each of them with N = 10,000). In other words, the antiwar demonstration has no longer five, but 50,000 participants and individual effects are now extrapolated to collectives. McQuail (2005) describes such “collective reaction” as follows: “Here some of the same effects are experienced simultaneously by many people in a shared situation or context, leading to joint action usually of an unregulated and non-institutional kind. Fear, anxiety and anger are the most potent reactions, which can lead to panic or civil disturbance” (pp. 468–469).

Based on this model, media effects theories, including priming and framing, need to be understood as processes that simultaneously influence individual, group and social outcomes. In the following sections, we therefore delineate each effects model at each unit of analysis, and then return to the model outlined in Figure 3.1 to discuss broader implications and future agendas for political communication research.

Framing and Priming at the Micro-Level

The various models of framing and priming effects have their explanatory foundations at the micro- or individual level. The broadest distinction at that level of analysis is the one between accessibility-based and applicability-based models of media effects (Price & Tewksbury, 1997).

Accessibility-based models include agenda setting and framing. Both effects models assume that the amount and prominence of media coverage of an issue can also increase how important audience members think these issues are or how easily they can retrieve them from memory (for an overview, see D. A. Scheufele, 2000a). Agenda setting and priming – in this context – have to be understood as two steps in a chronological series of effects. Agenda setting refers to the transfer of salience from mass media to audiences. Priming as a subsequent step, assumes that mass media can “prime” issues by making them more salient and – in turn – also activate related concepts in people's minds. This is usually referred to as spreading activation (Collins & Loftus, 1975). As a result, the salient issue and related concepts are more likely to be used for subsequent judgments about candidates or issues (Hastie & Park, 1986).

Applicability-based models are distinctly different. They do not rely on the idea of increased salience and ease of retrieval, but instead assume that the effect of messages depends on their applicability to pre-existing interpretive schema. The degree of applicability can increase or decrease, depending on how a message is framed. Kahneman's Nobel-prize winning work on prospect theory nicely illustrates this phenomenon. Respondents were shown an ambiguous stimulus in the form of a set of symbols that could be interpreted as numbers or letters. Respondents who saw the ambiguous stimulus embedded in a series of numbers interpreted it consistent with the surrounding frame, that is, as numbers, and respondents who saw it embedded in a set of letters interpreted it as letters (Kahneman, 2003).

Some researchers have treated applicability and accessibility models as equivalent (McCombs, 2004). We disagree with this view, and suggest that the two models are distinctly different. Priming (and agenda setting) focuses on what has been made salient by media coverage and how we use it for evaluations. The core process underlying priming and agenda setting, therefore is the ease with which a mediated issue can be retrieved from memory. Framing, in contrast, is largely unrelated to ease of retrieval of the mediated message. Instead, it focuses on how well the message resonates with pre-existing schema (for an overview, see B. T. Scheufele & D. A. Scheufele, 2010).

This distinction becomes even clearer when we think back to the characteristics of the negotiation paradigm of media effects outlined earlier. One of these characteristics was the interaction of audience characteristics with the features of the mediated message. Agenda setting, first formulated in the heyday of the strong effects paradigm of the 1970s (McCombs & Shaw, 1972), assumes that the transfer of salience from media to audiences is a universal phenomenon that occurs across all groups of society. In fact, most of the early agenda-setting research tested its hypotheses by comparing methodologically questionable rank-orders of the most salient issues in news media to issue rankings in national Gallup polls (Funkhouser, 1973a, 1973b).

Framing, in contrast, is firmly embedded in the negotiation paradigm. It assumes that framing effects are stronger for messages that better resonate with socially shared schema, or for groups who are more familiar with the schema that a particular frame is applicable to. “Three strikes and you're out” legislation in the United States is a good example. Without explaining the specifics of the legislation, it provides audiences with an intuitive understanding of the basic idea behind it. The “three strikes” frame is likely to be much more effective among US audiences for whom it resonates with an underlying shared schema. For audiences in Europe, this frame would be much less effective.

B. Scheufele (2004) picked up the argument of applicability and socially shared schema, but added two important other types of framing effects: The first one is called transformation effect: If media reports repeatedly frame an issue in a consistent fashion, audience schema can be modified step by step in terms of this media frame (see also B. T. Scheufele & D. A. Scheufele, 2010). A second, and partly related effect is called the establishing effect: If no socially shared audience schema is available, such a schema can be established by cumulative and consonant media framing. Such an effect should be more likely for issues or situations totally unfamiliar for most audience members. These two effects do not contradict the negotiation model of framing and priming effects, but rather complement and specify it. Finally, framing and priming can also result in shifts of opinions. Such attitudinal changes are quite different from the aforementioned persuasive media effects (Nelson, Oxley, & Clawson, 1997).

In short, agenda setting, priming and framing are all processes that can explain cognitive effects at the individual or micro-level. In particular, they provide insights into how media transfer salience to individual audience members (agenda setting) and potentially shape their evaluations of policies or political actors (priming). With a parallel logic, they can shape how issues are being discussed among audiences (frame setting) and how those frames affect outcomes, such as attributions of responsibility or other issue stances (individual-level framing effects) (D. A. Scheufele, 1999). More importantly, however, both accessibility and applicability models have implications well beyond the individual level.

Framing and Priming Effects at the Meso- and Macro-Levels

As outlined earlier, framing and priming models are largely based on theories adopted from cognitive psychology and try to explain how cognitions, emotions, opinions, and behavior of individuals are shaped by media framing and priming. But what about media effects on units at higher levels of analysis, such as public discourse, the political arena, or social movements? A number of scholars in sociology, for example, have suggested that attempts by non-governmental oranizations (NGOs) or other social movements to create and promote frames need to be examined using a collective identity approach (e.g. Gamson, 1992; Gamson & Modigliani, 1989). By the same token, however, it is also reasonable to ask whether predominant media framing or priming can promote or undermine a social movement's framing activities. This would be a media effect on a unit at the meso-level.

Sociological Approaches – Framing Beyond the Micro-Level

Traditional framing and priming effects approaches provide few insights into how media framing and priming alter the pattern of public discourse on a topic, how they support or undermine particular policy stances, or how they strengthen or weaken social movements more generally. A more exhaustive, multi-level model of framing or priming effects, therefore, needs to combine sociological traditions of framing research with other theories of media effects, such as the spiral of silence theory (Noelle-Neumann, 1974; D. Scheufele & Moy, 2000). We will discuss some of these connections in greater detail later on in this chapter.

Although the sociological tradition of framing research is largely agnostic of the historical backdrop of media effects research, two strands of this tradition are nonetheless relevant for a multi-level model of framing and priming effects. The first strand focuses mostly on framing in the context of social movements and mobilization (Snow & Benford, 1988, 1992; Snow, Rochford, Worden, & Benford, 1986). The second emphasizes the aspect of collective identity (e.g. Gamson, 1992; Gamson & Modigliani, 1989) (for an overview, see Opp, 2009).

According to mobilization-based approaches (e.g., Snow & Benford, 1988) social movements “frame, or assign meaning to and interpret relevant events and conditions in ways that are intended to mobilize potential adherents and constituents, to garner bystander support, and to demobilize antagonists” (p. 198). Anticipating similar distinctions in the communication discipline by Entman (1993), Snow and Benford (1988) highlighted problem definition (diagnostic framing) and treatment recommendation (prognostic framing) as important frame elements. But their conceptual framework also emphasized the motivational aspect of a movement's framing strategy: First, a movement has to define what the problem is (e.g., globalization) and attribute responsibility, i.e. who can be blamed for it or can be held responsible (e.g., capitalism). Second, the movement has to offer solutions or strategies to solve the problem (e.g., fair trade). Third, a social movement has to mobilize its existing members, but also call other people (e.g., through leaflets or the Internet) to support or join the movement.

Snow et al. (1986) distinguish four types of frame alignment necessary to gain support: Frame bridging means connecting frames and thereby bringing together people who share similar views or ideologies. With frame amplification the movement addresses core values of society (e.g., freedom, humanity) which can serve as a “springboard for mobilizing support” (Snow et al., 1986, p. 469). Frame extension and transformation are somewhat self-explanatory: A social movement sometimes has to extend or transform its framing strategy in order to attract a broader cross section of people.

Collective-identity approaches (e.g., Gamson, 1992; Gamson & Modigliani, 1989) expand on mobilization-based approaches and commonly distinguish three components of “collective action frames:” The first component – injustice – refers to the feeling among movement participants that they are not being treated fairly. The agency component addresses people's awareness of being able to change things and influence (political) decisions through collective action. The identity component, finally, represents the in-group feeling which triggers collective action.

Using the example of nuclear energy, Gamson and Modigliani (1989) saw at least two roles for mass media in the context of social identity: On the one hand, mass media serve as an arena or playground for public discourse. On the other hand, they stimulate public discourse by providing interpretative packages and framing devices, such as metaphors, key words or exemplars.

Although these framing concepts have been applied to media framing (e.g., Eilders & Lüter, 2000), they were initially defined as attributes and tools of social movements, rather than media coverage or even media effects. The multi-level matrix presented in Figure 3.1 therefore takes these concepts a step further and connects them to media effects research. It outlines how the sociological and psychological approaches mentioned earlier can be integrated into a multi-level model of media framing and priming effects.

The target or dependent variable of such effects may be a social group, an organization or a political system. But according to D. A. Scheufele (1999) media frames themselves can be the dependent variable: Social groups try to promote their perspective and thereby influence how other NGOs or policy groups, for instance, think about an issue. These processes of “frame building” and “frame setting” (D. Scheufele, 1999) have been examined in several studies (e.g., Cohen & Wolfsfeld, 1993; Entman, 2004; McAdam, 1994; Miller & Riechert, 2001; Walgrave & Manssens, 2005).

One of the more exhaustive summaries of research on framing strategies and activities in terms of political deliberation was provided by Pan and Kosicki (2001). Their analysis also moves this line of research from the earlier, more conceptual work to the “real political consequences” (p. 51) of framing activities during the 1992 presidential race: “The candidate who pledged to keep healthcare reform on top of his domestic agenda was swept into the White House and the Democratic Party maintained a sufficient working majority in both House and Senate” (p. 51). This was not only a result of efficient agenda building (Cobb & Elder, 1971), but also a result of successfully framing the issue as a question connected to social security and therefore effectively shifting evaluations of the presidential candidates toward these issues.

How media framing and priming contribute to political outcomes (e.g., elections, political decisions) can be considered a bottom-up media effect (see Figure 3.1): Media framing (see M in Figure 3.1) influences voting behavior of individual members of the electorate at the micro-level (see, for example, Y_A in Figure 3.1) as described by Price and Tewksbury (1997). Furthermore, media framing has an impact on the election outcome at the macro-level (see Z in Figure 3.1). This follows a simple rule of aggregation: All voting decisions Y_N (micro-level) sum up to the election victory Z of a political party (macro-level). In other cases, however, modeling media framing or priming effects on meso- and macro-units can be quite challenging.

Cross-Sectional Challenges of Modeling Multi-Level Media Framing and Priming Effects

When modeling multi-level media framing and priming effects, we face at least four problems.

(1) We already outlined the problem of linking levels in the “Operation Iraqi Freedom” scenario. Media coverage of terrorism provides another good illustration (e.g., Brosius & Weimann, 1991; Nacos, 2002; Stohl & Stohl, 2007; Tsfati & Weimann, 2002;). Terrorists try to gain public awareness for what they assume to be social or political problems, and try to promote their ideologies via televised messages, attacks on civil and military targets, and so on.

The first two activities are similar to what Snow and Benford (1988) called diagnostic and motivational framing. The terrorist attacks on Spanish trains prior to the Spanish national elections in March 2004, for example, were covered extensively by Spanish and international media. In the last days of the election campaign, the attacks and media reports had indirectly supported the Socialist party, which was a strong opponent of the Spanish involvement in “Operation Iraqi Freedom.” Since the media had framed the late election phase around the larger theme of terrorism, the electoral victory of the Socialist Party – and subsequently the withdrawal of Spanish troops from Iraq – was partly a media impact on the macro-unit “political system,” and not just a simple individual-level media framing and priming effect on individual Spanish citizens as the micro-unit.

In addition, mass media themselves were a target of framing or priming. According to Nacos (2002) terrorists try “to win media attention and news coverage for their actions, their grievances, and their political ends” (p. 10). It is therefore difficult to separate media effects at various levels from other confounding factors. In other words, research on framing and priming effects on terrorism needs to explore a complex interplay of different units at the micro-, meso-, and macro-level. Among them are poverty or historically grounded hostilities, the ideologies of terrorists, mass media as the conveyor of terrorism-related messages, the Internet as a tool for recruitment and planning or attacks, but also as a communication platform for terrorists (Tsfati & Weimann, 2002), the activities of supporters, the fear of the public, and finally the decisions and debates in various national and international political arenas. Mass media play both a mediating and a shaping role in this context (e.g., Nacos, 2002). So, it is not only the media who prime politicians and the public by applying a certain frame – it is political actors themselves who frame and prime the issue according to their own preferences.

Similar problems can occur when we study media framing and priming effects in the context of social movements (e.g., Opp, 2009). Framing and priming effects operate from lower to higher levels of analysis and backwards through a complex set of cross-level interactions. This includes frame and agenda setting by mass media, citizens' beliefs and predispositions (micro-level), the dynamics of demonstrations or the structure of particular social movements (meso-level), as well as the political decisions or even economic structures of a country (macro-level). The 2007 G8 summit in Heiligendamm (Germany) is a good example. Mass media framed the demonstrations around instances of violent opposition, and therefore shaped how the public, the police, and the demonstrators understood the events surrounding the summit. Yet, media coverage of the protests (meso-level) may also have shaped people's willingness to participate in subsequent demonstrations (micro-level). We called this top-down feedback earlier (see down arrows in Figure 3.1). Some people, for instance, may be concerned about repeated reports about instances of violence and decide not to participate in future rallies. Others may feel provoked by media portrayals of violence and get even more violent, thereby reinforcing this “spiral of violence.” Again, it is difficult to separate the framing and priming effect of media coverage from the impact of other factors.

(2) The second challenge relates to group contexts. When we study micro-level framing effects we can easily measure cognitions, emotions, opinions or reactions of an individual recipient. But with groups it is fairly difficult to isolate the specific role that media play in impacting group-level variables. Social protest is collective action either in the “smaller form” of a single demonstration or in the “larger form” of a social movement. Thinking back to Figure 3.1, the size of a protest or the violence of a demonstration or movement is therefore an attribute of a meso- or macro-level unit (Z). However, it is virtually impossible to establish if an individual participant's behavior (Y) is influenced by media coverage (M), and in which direction. More importantly, direct media effects may not even be a prerequisite for change, since demonstrators may be exposed to them indirectly through interpersonal channels (e.g., D. Scheufele, 2000b).

These group level confounds are further complicated by what is called reciprocal effects (Kepplinger, 2007; Lang & Lang, 1953). These can occur if a demonstration becomes more violent, for example, due to television teams being on-site. Such reciprocal effects make it even more difficult to isolate the specific impacts of initial media framing and priming. In sum, media framing and priming effects beyond the micro-level require significant conceptual thinking about the social dynamics surrounding them.

(3) The third problem concerns the mechanism of framing orpriming effects. This problem relates to the rules of aggregation or transformation mentioned earlier. According to Schulz (2009), models of learning, attitude change or information processing are not necessarily applicable to media effects on meso- or macro-level units. Moreover, other types of media effects have to be taken into account. As Halloran, Elliott, and Murdock (1970) outlined, media often create narratives around expectations about violence occurring, even before demonstrations take place. According to Merton (1996, pp. 41–50), such media framing can lead to a self-fulfilling prophecy: If media frame a demonstration in terms of violence before it actually occurs, they also influence interpretations and expectations among other social groups, such as the police. The police may deploy additional units to the demonstration, based on these media reports, which – in turn – can provoke demonstrators and trigger attacks on the police. These attacks trigger stricter enforcement by the police, which will further provoke protesters. The amount of violence at the demonstration (meso-level unit), therefore, is – at least in part – a result of the initial media framing, but also of various subsequent group dynamics, which are themselves triggered by media framing. Such a “spiral of violence” resembles the spiral of silence as described by Noelle-Neumann (1974): People monitor the opinion climate, and speak out or remain silent if they assume themselves to be of the opinion majority or minority. This stimulates a process in which the changing perception of the media climate reinforces the willingness to speak out, and so on. Schulz (2009) draws similar parallels to what he and others call the “mediatization” of the political system (see also, Kepplinger, 2002).

(4) The fourth challenge refers to the target of framing and priming effects. The individual recipient is – in many ways – an easy target (or unit of analysis) for empirical research. Yet, social movements, organizations, or even the political system are in many ways more relevant dependent variables. Unfortunately, shifting the dependent variable to group-level entities also requires an examination of the structure of these entities. For instance, if we explore media effects on an only loosely structured group, the dynamics of in- and out-group behavior (e.g., Tajfel, 1982) may explain how media framing and priming effects on each group member (micro-level) also affects the collective action of the whole group (meso-level). This could be called a follow-up effect. Organizations, however, are characterized by very different structures (e.g., Scott & Davis, 2007; Taylor, 1993). Media coverage is not perceived by an organization as a whole, but by its individual members. For instance, if the media frame an issue in such a way that opposes a particular political party's goals, the party leaders will likely distribute talking points and other framing documents that help other party members and campaign activists to counterframe these media reports. Here, media framing leads to a top-down feedback within the organization.

But the target of coverage could also be the party members who may read reports about their party leader being involved in a political scandal. Following the third-person effect hypothesis (e.g., Davison, 1983), these party members may fear that the electorate could vote for an opposition candidate in upcoming elections because of what they assume to be strong effects of the media coverage surrounding the scandal. Therefore, party members may communicate with other party members in order to oust their leader and avoid further political damage. This example proves that the group-level outcomes of individual-level framing effects are much more complicated than modeled previously, and other types of media effects have to be considered when we want to explore media framing and priming effects beyond the micro-level.

Longitudinal Challenges of Modeling Multi-Level Media Framing and Priming Effects

Finally, priming and framing research faces at least three problems connected to longitudinal considerations.

(1) The first problem refers to reciprocal bottom-up and top-down transitions. It can be illustrated by referring – once again – to the spiral of silence theory. Here, “the perception-to-expression-to-reinforcement process constitutes a major cross-level mechanism with bidirectional flow of causal forces” (Pan & McLeod, 1991, p. 158). Based on a spiral of silence model, the position seen by the public as winning or in the majority will – over time – establish itself as public opinion. Although only a small percentage of the electorate may in fact change their individual opinion (micro-level) based on public perceptions of the predominant opinion climate, this small group may change the outcome of a closely-contested political race (macro-level). The final target of media effects in the spiral of silence model, therefore, is located at the macro-level, that is, election outcomes or social norms. Again, not only media coverage, but several other factors contribute to this effect at the macro-level.

Similar cross-level dynamics have been discussed using the example of the anti-globalization protest at the G8 summit in Heiligendamm. Here, the framing strategies of the anti-globalization movement, the media reports, the overall opinion climate, the individual willingness to demonstrate, and finally the internal dynamics of a demonstration produce a complex cross-level interplay of factors. Yet, this cross-level structure is not static, as illustrated somewhat simplistically in Figure 3.1, but dynamic in the sense of alternating or simultaneous bottom-up and top-down transitions.

(2) The aspect of follow-up effects of media framing and priming can be illustrated by using the example of the Spanish national elections in 2004 again. After the terrorist attacks, mass media, politicians, and other actors framed the final stage of the election campaign largely in terms of terrorism. Voters following these media reports were therefore also more likely to judge the parties in terms of their position towards the Spanish involvement in Iraq – with the Socialist Party being a strong opponent of this commitment. Such media framing and priming effects operate at the level of the individual voter (micro-level). Some voters were primed by the media while others already favored the Socialist Party for other reasons (see subjects A and B in Figure 3.1). The tally of all voters' decisions (see Y_A and Y_B in Figure 3.1) produced the electoral victory (see Z in Figure 3.1) of the Socialist Party. But the election outcome is not the only follow-up media effect. Another one is the Socialist Party's withdrawing of Spanish troops from Iraq prior to the election.

Another example is the evolution of a single demonstration into a more long-term climate of anti-globalization protests. If mass media create an a priori narrative of anti-globalization demonstrations around violent protests, as outlined earlier, they may actively contribute to such violence at the specific upcoming demonstration through a self-fulfilling prophecy. In addition, such framing – as well as the violence actually occurring – can shape the overall climate of protest in the long run. Both, protesters and the police will prepare better for each forthcoming summit. The police will send more police units and the protesters will be organized better themselves. The demonstrators will act more violently and the police will enforce laws more strictly. Whether such outcomes can still be perceived as a media effect may differ from case to case. The multi-level blueprint in Figure 3.1 at least gives a formal answer: Changes at higher levels can be called a follow-up media effect if a significant amount of variance at the higher level can be traced back to media framing and priming. Of course, other factors contribute to this as well, for instance, the financial resources or the framing activities of a social movement. Media may support or undermine social movements by framing and priming. But it is difficult to measure the empirical mechanisms behind these dynamics empirically. Thus, most aspects of cross-level dynamics almost inevitably remain an empirical blackbox, which even methodically sophisticated studies of long-term framing processes are unable to open (e.g., Ferree, Gamson, Gerhards, & Rucht, 2002).

(3) A final challenge is related to the dynamic structure of the target of framing and priming effects on units at higher levels. According to the spiral of silence theory, each person remaining silent will increase perceptions about opposing views gaining strength. As a result, opinion climates are changing almost continuously. Framing and priming are similar. If media frame a demonstration in terms of violence, they not only influence those people who are willing to protest in the first place, but also those who may have considered demonstrating but are now discouraged by the prospect of violence. As a result, fewer people may attend the rally and those actually protesting may appear even more violent. Here, media framing not only influences individuals (micro-level), but also has an indirect impact on higher levels of social aggregation. For instance, media reports during a G8 summit have an impact on the demonstration at this summit. But the amount of protesters or people's readiness to act violently may already be a result of earlier media reports, which offered narratives of violent protests at this summit (see also Kepplinger, 2007).

Conclusions

This chapter explored priming and framing as media effects theories beyond the micro-level approaches that have dominated much of the work in political communication. We first positioned this line of research in current models of media effects, which have shifted away from Noelle-Neumann's and Gerbner's foci on the social-level outcomes of mass communication. Instead, we urge researchers to fully explore negotiation models of media effects, that is, models that hypothesize interactions among micro-, meso-, and macro-levels of analysis, regardless of many of the resulting complexities outlined in this chapter. This would also allow us to more explicitly connect the psychological tradition of framing and priming effects research, on the one hand, and sociological approaches to framing, on the other hand, at a conceptual level.

Second, as our overview shows, agenda setting, framing, and priming effects cannot be fully understood unless they are linked successfully to other theories or models of media effects. The spiral of silence theory or the third-person effect hypothesis have already been incorporated in our considerations about framing and priming effects at the meso- and macro-level, and future research should more systematically explore connections between these different effects models and their outcomes of different levels of analysis.

REFERENCES

Alexander, J. C., Giesen, B., Münch, R., & Smelser, N. J. (Eds.). (1987). The micro-macro link. Berkeley, CA: University of California Press.

Brosius, H.-B., & Weimann, G (1991). The contagiousness of mass-mediated terrorism. European Journal of Communication, 63, 63–75.

Cobb, R. W., & Elder, C. (1971). The politics of agenda-building: An alternative perspective for modern democratic theory. Journal of Politics, 33, 892–915.

Cohen, A. A., & Wolfsfeld, G. (Eds.). (1993). Framing the Intifada: People and media. Norwood, NJ: Ablex.

Coleman, J. S. (1994). Foundations of social theory (paperback ed.). Cambridge, MA: Harvard University Press.

Collins, A. M., & Loftus, E. F. (1975). A spreading-activation theory of semantic processing. Psychological Review, 82, 407–428.

Davison, W. P. (1983). The third-person effect in communication. Public Opinion Quarterly, 47, 1–15.

Entman, R. M. (1993). Framing: Towards clarification of a fractured paradigm. Journal of Communication, 43, 51–58.

Entman, R. M. (2004). Projections of power. Framing news, public opinion, and US foreign policy. Chicago, IL: Chicago University Press.

Eilders, C., & Lüter, A. (2000). Germany at war: Competing framing strategies in German public discourse. European Journal of Communication, 15, 415–428.

Esser, H. (1999). Soziologie. Spezielle Grundlagen. Band 1: Situationslogik und Handeln (Sociology: Special foundations. Vol. 1: Logic and action). Frankfurt am Main, Germany/New York, NY: Campus.

Eulau, H. (1977). Multilevel methods in comparative politics. American Behavioral Scientist, 21, 39–62.

Eveland, W. P., Jr. (1997). Interactions and nonlinearity in mass communication: Connecting theory and methodology. Journalism and Mass Communication Quarterly, 74(2), 400–416.

Ferree, M. M., Gamson, W. A., Gerhards, J., & Rucht, D. (2002). Shaping abortion discourse: Democracy and the public sphere in Germany and the United States. Cambridge, UK: Cambridge University Press.

Funkhouser, G. R. (1973a). The issues of the sixties: An exploratory study in the dynamics of public opinion. Public Opinion Quarterly, 37(1), 62–75.

Funkhouser, G. R. (1973b). Trends in media coverage of the issues of the sixties. Journalism Quarterly, 50, 533–538.

Gamson, W.A. (1992). Talking politics. New York, NY: Cambridge University Press.

Gamson, W. A., & Modigliani, A. (1989). Media discourse and public opinion on nuclear power: A constructionist approach. American Journal of Sociology, 95, 1–37.

Gerbner, G., & Gross, L. (1974). System of cultural indicators. Public Opinion Quarterly, 38, 460–461.

Halloran, J. D., Elliott, P., & Murdock, G. (1970). Demonstrations and communication: A case study. Harmondsworth, UK: Penguin.

Hastie, R., & Park, B. (1986). The relationship between memory and judgment depends on whether the task is memory-based or on-line. Psychological Review, 93, 258–268.

Hayes, A. F. (2006). A primer on multilevel modeling. Human Communication Research, 32, 385–410.

Hox, J. J. (2002). Multilevel analysis: Techniques and applications. Mahwah, NJ: Lawrence Erlbaum Associates.

Iyengar, S., & Kinder, D. R. (1987). News that matters: Television and American opinion. Chicago, IL: University of Chicago Press.

Kahneman, D. (2003). Maps of bounded rationality: A perspective on intuitive judgment and choice. In T. Frängsmyr (Ed.), Les Prix Nobel: The Nobel Prizes 2002 (pp. 449–489). Stockholm, Sweden: Nobel Foundation.

Katz, E. (1957). The two-step flow of communication: An up-to-date report on a hypothesis. Public Opinion Quarterly, 21(1), 61–78.

Katz, E., & Lazarsfeld, P. F. (1955). Personal influence. New York, NY: Free Press.

Kepplinger, H. M. (2002). Mediatization of politics: Theory and data. Journal of Communication, 52, 972–986.

Kepplinger, H. M. (2007): Reciprocal effects: Toward a theory of mass media effects on decision makers. The Harvard International Journal of Press/Politics, 12, 3–23.

Lang, K., & Lang, G. E. (1953). The unique perspective of television and its effect: A pilot study. American Sociological Review, 18, 3–12.

Lazarsfeld, P. M., Berelson, B. R., & Gaudet, H. (1948). The people's choice: How the voter makes up his mind in a presidential campaign (2nd ed.). New York, NY: Duell, Sloan & Pearce.

Lazarsfeld, P. F., & Menzel, H. (1961). On the relation between individual and collective properties. In A. Etzioni (Ed.), Complex organizations: A sociological reader (pp. 422–440). New York, NY: Holt, Rinehart & Winston.

McAdam, D. (1994). Taktiken von Protestbewegungen: Das Framing der amerikanischen Bürgerrechtsbewegungen (Tactics of protest movements: The framing of American civil rights movements). In F. Neidhardt (Ed.), Öffentlichkeit, öffentliche Meinung, soziale Bewegungen (Public, public opinion, social movement) (Kölner Zeitschrift für Soziologie und Sozialpsychologie; Sonderheft 34) (pp. 393–412). Opladen, Germany: Westdeutscher Verlag.

McCombs, M. E. (2004). Setting the agenda: The mass media and public opinion. Oxford, UK: Blackwell.

McCombs, M. E., & Shaw, D. L. (1972). The agenda-setting function of mass media. Public Opinion Quarterly, 36(2), 176–187.

McCombs, M. E., & Shaw, D. L. (1993). The evolution of agenda-setting research: 25 years in the marketplace of ideas. Journal of Communication, 43(2), 58–67.

McLeod, D. M., Kosicki, G. M., & McLeod, J. M. (2009). Resurveying the boundaries of political communication effects. In J. Bryant & D. Zillmann (Eds.), Media effects: Advances in theory and research (3rd ed., pp. 228–251). Mahwah, NJ: Lawrence Erlbaum Associates.

McLeod, J. M., & Pan, Z. (1989). Getting levels across and crossing levels. Comments on Reeves. American Behavioral Scientist, 33, 199–202.

McQuail, D. (2005). Mass communication theory (5th ed.). Thousand Oaks, CA: Sage.

Merton, R. K. (1996). Social theory and social structure (P. Sztompka, Ed.). Chicago, IL: University of Chicago Press.

Miller, M. M., & Riechert, B. P. (2001). The spiral of opportunity and frame resonance: Mapping the issue cycle in news and public discourse. In D. Reese, O. H. Gandy, & A. E. Grant (Eds.), Framing public life. Perspectives on media and our understanding of the social world (pp. 107–121). Mahwah, NJ: Lawrence Erlbaum Associates.

Nacos, B. L. (2002). Mass mediated terrorism: The central role of media in terrorism and counter-terrorism. Lanham, MD: Rowman & Littlefield.

Nelson, T. E., Oxley, Z. M., & Clawson, R. A. (1997). Toward a psychology of framing effects. Political Behavior, 19, 221–246.

Noelle-Neumann, E. (1973). Return to the concept of powerful mass media. Studies in Broadcasting, 9, 67–112.

Noelle-Neumann, E. (1974). The spiral of silence: A theory of public opinion. Journal of Communication, 24(2), 43–51.

Oakes, J. M., & Kaufman, J. S. (2006). Introduction: Advancing methods in social epidemiology. In J. M. Oakes & J. S. Kaufman (Eds.), Methods in social epidemiology (pp. 3120). San Francisco, CA: Jossey-Bass/Wiley.

Opp, K.-D. (2009). Political protest and social movements. A multidisciplinary introduction, critique, and synthesis. Milton Park, UK: Routledge.

Pan, Z., & Kosicki, G. M. (2001). Framing as strategic action in public deliberation. In D. Reese, O. H. Gandy, & A. E. Grant (Eds.), Framing public life. Perspectives on media and our understanding of the social world (pp. 35–65). Mahwah, NJ: Lawrence Erlbaum Associates.

Pan, Z., & McLeod, J. M. (1991). Multilevel analysis in mass communication research. Communication Research, 18, 140–173.

Price, V., & Tewksbury, D. (1997). News values and public opinion: A theoretical account of media priming and framing. In G. A. Barett & F. J. Boster (Eds.), Progress in communication sciences: Advances in persuasion (Vol. 13, pp. 173–212). Greenwich, CT: Ablex.

Scheufele, B. (2004). Framing effects research: A theoretical and methodical critique. Communications – European Journal of Communication Research, 29, 401–428.

Scheufele, B. (2008). Das Erklärungsdilemma der Medienwirkungsforschung. Eine Logik zur theoretischen und methodischen Modellierung von Medienwirkungen auf die Meso- und Makro-Ebene (The explanation dilemma in media effects research. A logic for theoretical and methodical modeling of media effects on meso- and macro-level units). Publizistik, 53, 339–361.

Scheufele, B. (2011). Jenseits des Rezipienten. Die Herausforderungen für die Medienwirkungsforschung durch die Mehr-Ebenen-Problematik am Beispiel von fünf Wirkungsbereichen. In T. Quandt & B. Scheufele (Eds.), Ebenen der Kommunikation. Mikro-Meso-Makro-Links in der Kommunikationswissenschaft (Levels of communication: Micro-meso-macro-links in communication science) (pp. 113–142). Wiesbaden, Germany: VS Verlag für Sozialwissenschaften.

Scheufele, B. T., & Scheufele, D. A. (2010). Of spreading activation, applicability, and schemata: Conceptual distinctions and their operational implications for measuring frames and framing effects. In P. D'Angelo & J. A. Kuypers (Eds.), Doing news framing analysis: Empirical and theoretical perspectives (pp. 110–134). New York, NY: Routledge.

Scheufele, D. A. (1999). Framing as a theory of media effects. Journal of Communication, 49(1), 103–122.

Scheufele, D. A. (2000a). Agenda-setting, priming, and framing revisited: Another look at cognitive effects of political communication. Mass Communication & Society, 3(2/3), 297–316.

Scheufele, D. A. (2000b). Talk or conversation? Dimensions of interpersonal discussion and their implications for participatory democracy. Journalism & Mass Communication Quarterly, 77, 727–743.

Scheufele, D. A., & Moy, P. (2000). Twenty-five years of the spiral of silence: A conceptual review and empirical outlook. International Journal of Public Opinion Research, 12, 3–28.

Schulz, W. (2009). Politischer Medieneinfluss: Die Metamorphosen des Wirkungskonzepts. In B. Pfetsch & F. Marcinkowski (Eds.), Politik in der Mediendemokratie (Politics in the media democracy) (Politische Vierteljahresschrift, Sonderheft 42) (pp. 103–125). Wiesbaden, Germany: VS Verlag für Sozialwissenschaften.

Scott, W. R., & Davis, G. F. (2007). Organizations and organizing: Rational, natural, and open perspectives. Upper Saddle River, NJ: Prentice Hall.

Snow, D. A., & Benford, R. D. (1988). Ideology, frame resonance, and participant mobilization. In B. Klandermans, H. Kriesi & S. Tarrow (Eds.), International social movement research. Vol. 1: From structure to action: Comparing social movement research across cultures (pp. 197–217). Greenwich, CT: JAI Press.

Snow, D. A., & Benford, R. D. (1992). Master frames and cycles of protest. In: A. D. Morris & C. McClurg Mueller (Eds.), Frontiers in social movement theory (pp. 133–155). New Haven, CT: Yale University Press.

Snow, D. A., Rochford, E. B., Worden, S. K., & Benford, R. D. (1986). Frame alignment processes, micromobilization, and movement participation. American Sociological Review, 51, 464–481.

Stohl, C., & Stohl, M. (2007). Networks of terror: Theoretical assumptions and pragmatic consequences. Communication Theory, 17, 93–124.

Tajfel, H. (Ed.) (1982). Social identity and intergroup relations, Cambridge, UK: Cambridge University Press.

Taylor, J. R. (1993). Rethinking the theory of organizational communication: How to read an organization. Norwood, NJ: Ablex.

Tewksbury, D., & Scheufele, D. A. (2009). News framing theory and research. In J. Bryant & M. B. Oliver (Eds.), Media effects: Advances in theory and research (3rd ed., pp. 17–33). Hillsdale, NJ: Lawrence Erlbaum Associates.

Tsfati, Y., & Weimann, G. (2002). www.terrorism.com: Terror on the internet. Studies in Conflict and Terrorism, 25, 317–332.

Walgrave, S,. & Manssens, J. (2005). Mobilizing the White March: Media frames as alternatives to movement organizations. In H. Johnston & J. A. Noakes (Eds.), Frames of protest: Social movements and the framing perspective (pp. 113–140). Lanham, MD: Rowman & Littlefield.

Weimann, G. (1994). The influentials: People who influence people. Albany, NY: State University of New York Press.