Framing the Participatory Society: Measuring Discrepancies Between Interpretation Frames and Media Frames

Framing the Participatory Society: Measuring Discrepancies Between Interpretation Frames and... Abstract This research introduces a methodological framework for a valid and reliable analysis of both media and interpretation frames. It responds to recent calls in literature (1) to acknowledge the social–cultural context and power relations underlying framing, (2) to consider frames as a cluster of frame-elements, and (3) to identify interpretation frames and media frames using a similar analytical approach to ensure a valid comparison of the outcomes of both interrelated framing processes. The methodological framework was applied to three studies. Results show that the discrepancy between interpretation frames and media frames revolved around the distinction between bottom-up and top-down frames. The introduced method provides meaningful insights into the complete framing process of both the sender and receiver of communication messages. During the King’s speech held on September 17, 2013, the Dutch government coined the idea of the “participatory society” as a new form of society intended to replace the classical Dutch welfare state. In the participatory society, each citizen who is able to participate will be asked to take more responsibility, especially regarding community care and social security. This societal development received a lot of resistance from citizens and became topic of fierce political debate in the subsequent period. The resistance articulated by the people was inherently different from the way the government expressed its views on the participatory society. The government’s frame emphasized a certain definition of the problem that empowered citizens as being in charge of their own fate. It also emphasized the causes for a shift from the classical welfare state to the participatory society. By stating that Dutch citizens should contribute more to their society as a whole, a certain moral evaluation was emphasized. Finally, by arguing that citizens need to increase their participation in different fields of society, a recommendation on how to “treat” the participatory society was introduced. In other words, all four elements constituting a frame were present in the King’s speech (Entman, 1993). Framing research in communication science has mainly focused on the sender-side of communication, predominantly by using a quantitative or qualitative content analysis to distil frames out of media content (e.g., Semetko & Valkenburg, 2000; Tankard, 2001). The receiver-side of framing has mainly been analyzed as the effects of framed communication. Carragee and Roefs (2004) argue that this traditional approach neglects the origins of frames. In their opinion, framing research should be related to underlying sociocultural processes, as the distribution of power in society influences both the interpretation and construction of frames. In this reading, framing is defined as a process of social reality construction that involves both receivers and senders of news stories. As social reality is constructed within a specific meaning-loaden context, the interpreting receiver of frames should be assigned a central role in the analysis of frames (Gamson & Modigliani, 1989). Previous studies are inconclusive on how frames should be operationalized in a valid and reliable way, both on the sender- and receiver-side (e.g., Matthes & Kohring, 2008). Reviewing five prevalent approaches in framing research, Matthes and Kohring point to two major issues for reliability and validity. First, inductive interpretative framing approaches lack reliability. Predominantly, these approaches provide little transparency in how frames are extracted from the data. Furthermore, these approaches are frequently based on an identification of frames as single entities by coders. This implies that the identification of a set of frames in the text becomes an arbitrary practice, which harms the reliability and validity of the analysis (see also Tankard, 2001). Second, deductive, often systematic computer-assisted approaches, may lack validity. As frames are embedded in both media and public discourses, the extraction of frames cannot be reduced to a set of computer-coded variables. Furthermore, the deductive approach is limited, as it presupposes that all frames are known beforehand, which overlooks the identification of novel frames that are not anticipated by the researchers. So far, comprehensive methodological tools for a valid and reliable analysis of the complete interactive framing process are still lacking (e.g., Carragee & Roefs, 2004; Entman, 1993; Scheufele, 1999). By responding to conceptualizations and suggestions for future research by Matthes & Kohring (2008) and Van Gorp (2007), this article first of all proposes a standardized method for the analysis of sender-side media frames and receiver-side interpretation frames that strives toward validity and reliability. Second, this method is validated by analyzing how the Dutch participatory society is framed by Dutch news media and interpreted by Dutch citizens. The general aim of this study is therefore to explore the discrepancy between media frames and interpretation (audience) frames using a systematic method of framing analysis. To precisely study the discrepancy and overlap between media frames and interpretation frames, three studies were conducted. In the first study, interpretation frames were analyzed qualitatively. In the second study, frames operating in the media discourse were identified quantitatively. Based on qualitative insights, the third study analyzed interpretation frames by means of a quantitative cluster analysis. The Sender-Side of Framing: Media Frames and Their Operationalization The process of framing entails the reconstruction of social reality in meaningful patterns of interpretation resonating in cultural values shared by both sender and receiver (Gamson & Modigliani, 1989). These patterns of interpretation manifest themselves in a communication text as story lines and organizing ideas, which guide both the sender and receiver of communication in their process of making sense of (news) events (Huang, 2010). An important question for framing analysis is how frames can be identified validly in a text. Entman’s (1993) conceptualization of frames as consisting of four distinct elements provides more insight into how frames manifest themselves in communication texts. First, frames promote a particular problem definition by highlighting what the issue is about, what the costs and benefits of the issue might be, and which actors are involved in the issue. Second, the causal interpretation emphasizes the causes and consequences of the issue and identifies who is responsible for these causes and effects. Third, the moral evaluation of the problem evaluates the causes and effects, and connects these to moral values and norms familiar in the cultural context. Finally, the treatment recommendation suggests potential solutions to the problem. The operationalization of framing can be made specific by regarding the four elements and locations as a “frame package” embedded in a communication text (Gamson & Modigliani, 1989; Van Gorp, 2007). By reconstructing this package in a text, the frame can be identified in a systematic and objective way. The frame package consists of three different parts that together represent the frame. First, the manifest framing devices refer to arguments, visual images of the issue, examples of the issue, metaphors, and word choice. These framing devices can be coded in a text as manifest characteristics of the frame package (Van Gorp, 2007). The cement that holds the manifest framing devices together is the frame itself. It provides an overarching meaning structure to the manifest components of the frame. Second, at a more latent level, frames consist of reasoning devices (Gamson & Modigliani, 1989; Van Gorp, 2007). These reasoning devices reflect the four elements of the frame as conceptualized by Entman (1993). The reasoning devices in a text can be identified as the aggregate of implicit and explicit references that promote a particular problem definition, causal interpretation, moral evaluation, and/or treatment recommendation. Third, as frames are embedded in cultural contexts, they reflect broader cultural themes shared by the receiver and sender of frames (Van Gorp, 2007). This thematic component is more difficult to code in an objective way, as it requires the coder to be familiar with the underlying cultural values, norms, and stereotypes. Moreover, attaching cultural themes to a frame package is an arbitrary practice that largely depends on researchers’ interpretation. Here, it is important to note that talking about the media as sender thus not inherently implies that the agency of the individual journalist is unbounded. On the contrary, media framing is constructed within the cultural context in which it is embedded, with multiple institutional and noninstitutional actors trying to influence the content of coverage, especially when it comes to politically contested issues. Moreover, news values and journalistic practices constrain the individual journalist (see Vliegenthart & Van Zoonen, 2011). Reflecting on the above-presented operationalization of framing put forward in previous studies, the main focus of this research is to (1) systematically identify reasoning devices as frame-elements; to (2) reconstruct these elements into frame packages, both as media frames and interpretation frames; and to (3) attach meaningful themes as labels to these frame packages. The qualitative study, in particular, aims to provide an in-depth understanding of the cultural values underlying the frames. The measurement tool that is designed to identify frames in a systematic way is outlined in detail in the method section. The Receiver-Side of Framing: Interpretation Frames in the Public Discourse Social constructivism postulates that the media interpret and emphasize issues in a different way compared with the active meaning-constructing receivers of these media messages (Holstein & Gubrium, 1995; Neuman, Just & Crigler, 1992; Wicks, 2001). Previous research differs in the locus of power ascribed to senders (media) versus receivers. Some authors argue that receivers choose from a variety of persistent frames that are made available to them by the media (e.g., Reese, 2001). In this reading, the media are powerful in the sense that they control the social reality construction process of receivers by forcing them to choose from a limited number of alternative interpretations. A relevant question is whether framing is indeed as powerful as suggested in the “powerful media effects paradigm” (McQuail, 2010). Some authors challenge this paradigm by ascribing more agency to receivers, who construct their own interpretation of issues in interaction with media makers and other social actors, rather than being controlled by powerful media forces (e.g., Scheufele, 1999). This study ties in with this reading by emphasizing that both media senders and receivers are part of the interpretation process underlying framing, as both construct their vision on reality out of the same stock of social and cultural resources. The paradox in previous literature on the receiver-side of framing is that on one hand, framing is approached from a social-constructivist perspective as the result of cultural actors’ interpretations, while at the other hand, frames are viewed as external to the individual and should therefore be analyzed independently from the actors as part of cultural resources. This study takes on an intermediate perspective on this issue by acknowledging the power of the sender of frames to propose a preferred reading of the participatory society while at the same time assigning a central role to the agency of the audience. In this reading, receivers process information in a motivated way instead of simply accepting or rejecting media frames on the participatory society (Gamson & Modigliani, 1989). People are thus not only active in resisting frames that are counter to their attitudes and accepting frames that are consistent with their attitudes, they construct their own individual interpretation of issues as well. Therefore, traditional approaches to audience-side framing that measure people’s interpretation simply as the effects of framed communication are insufficient in capturing the complete meaning construction process of the audience. Bridging the Gap: Framing as Interactive Practice Between Sender and Receiver Entman (1993) argued that the dominant frame embedded in communication texts on a certain issue should be congruent with the most common interpretation frames among receivers. However, Neuman et al. (1992) found that media frames and audience frames are not by definition congruent, and alternative frames not present in media coverage may resonate in the public discourse, also because citizens have a plethora of alternative information sources at their disposal (Gamson and Modigliani, 1989). In other words, not all interpretation frames are “out there” to be analyzed directly from media content. Although the audience-side of framing has been analyzed in framing effect studies, the discrepancies and linkages between media frames and interpretation frames analyzed separately on both sides of the communication process have not received extensive attention (Scheufele, 1999). Building on Scheufele’s (1999) conceptualization of framing as a cyclical process between senders and receivers, the perspective on framing taken in this study acknowledges two divergent but interrelated processes of framing. First, news media frame issues in a certain way, selecting aspects of social reality, and reconstructing these aspects in a salient way (Entman, 1993; Goffman, 1974; Van Gorp, 2007). Second, citizens decode the framed communication text in a way that is meaningful within their own frame of reference. If the media frame is congruent with their own viewpoint, citizens are expected to accept the frame as embedded in the communication text (Entman, 1993). However, if citizens are involved with the issue and the media frame is dissonant with their viewpoint on the issue at stake, the media frame will be rejected or adjusted. In contrast to the way the power of frames is referred to in traditional literature on framing as media effects, the two extremes of acceptance and rejection should not be regarded as the single two possible receivers’ interpretations. As Hall (1973) postulated, a middle position is also feasible—in which the decoding stance is mixed regarding acceptance and rejection of the connoted meanings. This position partly accepts the legitimacy of the dominant media discourse, but adapts it according to the situational context. The threefold (accept/mixed/reject) conceptualization of decoding positions hypothesized by Hall (1973) is in line with the empirical findings of Neuman et al. (1992), Gamson & Modigliani (1989), and Huang (2010). Acceptance and rejection represent the two ends on a continuum of possible responses toward each of the four elements that constitute a frame. Against this backdrop, this study introduces and validates a methodological tool for measuring media frames and interpretation frames in a similar but independent way, making similarities and differences clear. The substantial research question we pose reads as follows: RQ: To what extent and how does the framing of the participatory society differ between interpretation frames of citizens and media frames of Dutch news media? All empirical studies in this article are concerned with answering (parts of) this research question. Study 1 aims to provide insights into the framing process from the perspective of the receiver by means of a qualitative framing analysis of interviews and focus groups. By means of a quantitative content analysis, Study 2 aims to provide insights into the media frames used to frame the participatory society in Dutch newspapers. Study 3 is devoted to validate the inductively gathered interpretation frames of Study 1 by means of a quantitative cluster analysis of interpretation frames. The framing analysis in all three studies is based on a similar measurement instrument developed for this study. The characteristics of this framing measurement instrument are outlined in greater detail in the following section. Methodological Framework Responding to Matthes and Kohring’s (2008) suggestion for future research to measure media frames and audience frames with a similar method that clusters manifest variables together as frames, we used the in-depth analysis of qualitative material in Study 1 to develop a framing measurement tool that was applied to quantitative cluster analyses of both media frames and audience frames. The five steps that were used to qualitatively develop the measurement tool are discussed below. The first step of analysis resembles open coding in the grounded theory approach (Bryman, 2012; Flick, 2009). Regarding the four frame-elements or reasoning devices problem definition, causal interpretation, moral evaluation, and/or treatment recommendations as sensitizing concepts, fragments in the transcripts that reflected these concepts were highlighted. Open codes were used to code the reasoning devices that were captured in a certain fragment. During the second step of analysis, the open codes were structured around the four frame-elements. This step resembles axial coding in the grounded theory approach, for which the “axes” were represented by the four frame-elements. By completing the codebook for each respondent individually, the step from open coding to axial coding resulted in the development of structured themes and categories derived from relatively unique open codes. The codebook was revised until theoretical saturation was achieved (Glaser & Strauss, 1967). More specifically, after new rounds of analysis of interview transcripts no longer resulted in the emergence of new questions (codes), the codebook included all aspects of which the reasoning devices (frame-elements) are constructed. During the third step of analysis, the final version of the codebook was translated into a framing matrix (see Van Gorp, 2007). Categories were related to one another during this step (Bryman, 2012). Specifically, the four reasoning devices or frame-elements were connected to form frame packages in a framing matrix. The columns of the matrix represented the four reasoning devices, in which the different questions for each reasoning device were synthesized to one underlying theme, the frame-element. The rows of the matrix represented individual participants. During the fourth step of analysis, the individual frame packages were compared for similarities and differences to extract frames from the individual packages. This step inherently contains a subjective element, as the researchers have to attach the frame packages to social–cultural values and meanings. However, by staying close to the original frame packages in the clustering process, the researchers’ bias was minimized. To enhance the richness of frames and to maintain the embeddedness of frames in their social–cultural context, frames were illustrated with in-verbatim quotes of participants during the fifth step of analysis. The above-mentioned procedures for the framing analysis formed the core of the framing measurement tool designed in this study. In the next step, the results from the qualitative analysis were used to develop statements reflecting opposing positions on the different frame-elements. These statements were measured on 7-point scales (see Appendix C1 in the supplemental material for statements). These items reflected the qualitatively identified variety of different positions on all frame-elements and provided the input for the cluster analysis of media frames and interpretation frames (Study 2 and Study 3). Our approach ensured that we did not “feed” the quantitative analyses with a limited selection of predeveloped frames. Instead, we reduced participant’s interpretations to opposing positions on statements measuring the four frame-elements. Participants’ qualitatively identified interpretations of the participatory society, for example, revolved around different top-down and bottom-up views on how to deal with the government’s request to increase participation, which were reduced to three statements measuring the treatment recommendation in the quantitative phase of the research. Study 1: Qualitative Inductive Analysis of Interpretation Frames Method Sample. First, as a prestudy, we conducted a qualitative content analysis on transcripts of individual interviews and focus group discussions, originally not collected for the purpose of this study (N = 40). This sample was used to inductively identify frame-elements, which informed the measurement of frames in subsequent studies. This pilot study aimed to map the development of the debate on the participatory society in public opinion by analyzing interviews conducted at various stages of the debate. The pilot study further aimed for maximum variation by also including transcripts from interviews on a broader topic: citizens’ general views on and satisfaction with society. The open-ended analyses of this rich material resulted in a variety of interpretations on all four frame-elements. The data collection was ended when the analysis of new material did not result in novel themes. This exploratory content analysis was used to develop an interview guide for primary data collection. For the data collection of this inductive analysis, participants were sampled based on purposive sampling (Bryman, 2012). The inclusion criteria aimed to reflect maximum variety regarding age, gender, region, and educational level. The composition of the focus groups was homogenous with respect to educational level to avoid the influence of unequal levels of knowledge and power relations on the experiences shared within groups (McLafferty, 2004). In total, 12 participants participated in individual interviews. The mean age of participants was 47 years old, ranging from 23 to 68 years old. Half of participants were female. The two focus groups consisted of seven participants each. The mean age of participants in the focus groups was 47 years old, ranging from 24 to 66 years old. In total, 8 of 14 participants in the focus groups were female. Data collection. The semi-structured interview guide is included in the supplemental material (Appendix A1). The interviews and focus group discussions were held in May 2014. All conversations were audio–video taped and transcribed in verbatim. The average duration of the individual interviews was 40 min. The focus group discussions lasted 90 min. Data analysis. The five-step framing analysis measurement tool outlined in the methodological framework was used for the qualitative analysis of interpretation frames. The framing matrices including the four frame-elements are available on demand. Intercoder agreement was assessed for a randomly selected sample of six interview transcripts and one focus group discussion (50%). Two coders independently coded this subsample using the framing matrix. Differences were discussed until agreement on all four frame-elements was reached. The initial differences between independent analyses were marginal, and did not have implications for the outcomes of the framing-analysis as reported in the results section. Results Study 1 Based on the input of the framing-matrix, five interpretation frames were identified in the in-depth interviews and focus groups. In Appendix A2 of the supplemental material, a detailed description of all frames with in-verbatim quotes is included. The dominant interpretation frame can be labelled as Participation is bad when forced upon us from above. In their problem definition, participants emphasized that participating in society is a positive principle, as it may enhance solidarity. In their causal interpretation, participants related the participatory society entirely to economic savings. Morally reasoned, participants emphasized their limited willingness and ability to participate more. Furthermore, participants believed that the participatory society has negative consequences for Dutch society. As treatment recommendation, participants suggested that the government should not force citizens to participate. By allowing them more freedom of choice, citizens will decide for themselves what they are willing and able to do besides their responsibilities in busy everyday life. The second frame that counters the dominant top-down interpretation frame touches on an exclusionist view of society. This frame can be labelled as “we” are more willing than “they” are. This frame emphasizes that, besides top-down initiated austerity measures, the decreasing solidarity in Dutch society has caused the shift toward the participatory society from bottom-up. A need for Dutch citizens to look out for each other is emphasized in the moral evaluation. This desire for citizens’ initiatives will however not be satisfied, as most “other” citizens are unwilling to do more for society. Distrust in “other” individualistic citizens is emphasized in the moral evaluation, which counters the empathy for “victims” foregrounded in the dominant frame. According to participants’ treatment recommendation, the social dilemma resulting from the participatory society can be solved by governmentally offered incentives to participate. The third interpretation frame highlights a more positive problem definition: participants emphasized they are both willing and able to improve the quality of their living environment from bottom-up. This frame can be labelled as participation is a virtue, not a burden. The moral evaluation of this frame suggests that a shift in mentality is needed; citizens will need to have more mutual respect and need to look out for each other more to make the participatory society work. Similar to the previously described frames, inequality in willingness and ability to participate were emphasized. In the treatment recommendation, however, willingness and ability were interpreted as two different things. A lack of ability needs to be overcome by limiting the responsibilities for citizens (top-down), whereas limited willingness needs to be overcome by a shift in mentality of Dutch citizenry (bottom-up). The fourth interpretation frame was only found in the focus group discussions. It emphasizes how the participatory society is a disguised austerity measure initiated by the government. This frame can be labelled as participation is nothing but disguised savings. Morally reasoned, doing more for each other in society is a positive thing. However, this should be motivated by a bottom-up emerging desire for solidarity, and should not result from austerity measures. The fifth frame, also found exclusively in the focus group discussions, introduces a completely bottom-up interpretation of the participatory society. This frame can be labelled as participation enhances solidarity. In the causal interpretation, the participatory society is seen as the result of decreasing solidarity, which incited citizens to act from a less individualistic perspective. To the extent that this frame represents the only “pure” bottom-up interpretation, it can be thought of as a counter frame of all previously mentioned frames that foregrounded the top-down initiated economical savings as main driving force behind the participatory society. Conclusion and Discussion Study 1 This study revealed in what ways the participatory society was interpreted from either “bottom-up” or “top-down” audience frames. The top-down interpretations articulated least trust in the participatory society, as it was forced on citizens by the government. Bottom-up frames, in contrast, emphasized how citizens felt they were willing to do more for society. Still, participation should be something initiated by citizens themselves, and the government should only play a facilitating role. To the extent that new rounds of data collection in the interviews and focus group did not reveal new themes in the interpretation frames, theoretical saturation of interpretation frames was achieved (Glaser & Strauss, 1967). As the variety in viewpoints on the participatory society has been mapped out exhaustively, we can move on to the quantitative phase of the framing analysis. As a limitation of this first qualitative study, it should be noted that generalizations beyond the participants might potentially be problematic. However, by including qualitative data collected in various periods of the debate, and by triangulating the issue-specific findings with those of a more general topic in the pilot study, we aimed for maximum variation in interpretations regarding the four frame-elements resonating in public opinion. In the process of selective coding, these interpretations were translated into more generic themes. These were formulated as manifest variables indicating latent frame-elements in the content analysis of Study 2. Study 2: Quantitative Analysis of Media Frames Method Data collection. A quantitative content analysis was conducted to identify media frames. For the collection of data, a coding protocol and a coding sheet were developed (see Appendix B1 in the supplemental material for codebook). The variables in the codebook were based on a quantification of the qualitative study’s in-depth insights on the different frame-elements. Regarding the problem definition, for example, the qualitative study revealed that participants distinguished between the impact of the participatory society for themselves personally and for society in general. Moreover, in expressing their (dis)approval of this societal development, participants distinguished between participation initiated by citizens and participation requested by the government. To ensure these stances were reflected in the codebook, we included these distinctions in the variables measuring frame-elements. An example of such a variable included in the codebook is The participatory society as introduced by the government is a negative societal development versus The participatory society as introduced by the government is a positive societal development. On a 7-point semantic differential scale, we scored the position of each article regarding those oppositional statements. The codebook was pretested on a small sample (N = 15) of articles from regional newspapers. Based on the results of the pretest, the codebook was adjusted until the coding of new data no longer resulted in changes to the codebook. In that sense, saturation of the categories in the codebook was achieved. Sample. All articles covering the Dutch participatory society published in all Dutch national newspapers between September 17, 2013 and February 28, 2014 were included in the sample. The starting point is chosen, as the word “participatory society” was picked up in the media discourse right after the King’s speech on the September 17. From this point on, the discussion developed in both the media and public discourse. The sample was a census, considering every unit in the population was selected in the specified time frame (Bryman, 2012; Riffe, Lacy & Fico, 2005). The unit of analysis was the newspaper article. The final sample consisted of 152 Articles. Articles were selected using the following inclusion criteria in LexisNexis: the article had to be published in a Dutch national newspaper between September 17, 2013 and February 28, 2014, and the article had to contain at least one of the key words “participatiesamenleving”, “participatiemaatschappij,” or “doe-democratie” (three different Dutch words describing the concept participatory society). After a thorough reread of all sampled articles, 29 articles were excluded from the sample for the reason their content did not reflect a discussion on the participatory society. Data analysis. First, intercoder reliability was assessed for a sample of 20 randomly selected articles coded by two independent coders. Krippendorff’s alpha was computed as index for intercoder reliability. As this content analysis concerns an exploratory study, a value of 0.60 was considered as an acceptable threshold value for the variables included in the framing analysis (e.g., Lombard, Snyder-Duch & Bracken, 2002). As none of the variables in the codebook had a reliability lower than 0.60, all variables were retained for the analysis. The value of Krippendorff’s alpha for all variables is included in the supplemental material (Appendix B1). A principal component analysis was conducted on all 16 statements of the codebook. The factor analysis (rotation varimax) yielded a solution of four components with an eigenvalue greater than one that together explained 65.0% of the variance (the rotated component matrix is included in Table B1 in the supplemental material). As a robustness check, we also estimated the dimensional structure with an alternative measurement model for ordinal data that takes the hierarchy of item difficulty into account: Rasch scaling. The scalability indices of both measurement models provide similar results. Table 1 Framing Matrix Used to Identify Media Frames Frame-element Problem definition Causal interpretation Moral evaluation Treatment recommendation Frames Assessment of citizens’ initiatives in terms of their capacity and willingness to participate Assessment of causes and consequences of governmental influence on the participatory society Solidarity and equality of consequences participatory society for Dutch citizens Should the government or its citizens participate more in society? 1. (N = 60): “Citizens are better able to initiate participation from bottom-up than the government. But both need to participate more in society” (M = 4.04, SD = 0.49): Some confidence in citizens’ own initiatives (development is not positive/not negative) (M = 3.45, SD = 0.60): a moderately negative assessment of governmental responsibilities and causal influence (M = 3.40, SD = 0.76): to some extent, consequences of participatory society are unequally distributed and solidarity decreasing (M = 3.77, SD = 0.52); both government and citizens need to participate more 2. (N = 55): “Trust in bottom-up citizens’ initiatives” (M = 5.04, SD = 0.86): confidence in citizens’ own initiatives (development is relatively positive) (M = 3.70, SD = 0.77): less trust in causes and consequences government. Bit negative assessment of government’s causal responsibility (M = 4.44, SD = 0.62): participatory has no unequal consequences and is not solidary decreasing (M = 2.85, SD = 0.82): citizens should participate more than the government 3. (N = 26): “Skeptical of citizens’ initiatives and the government’s influence on the unfair participatory society” (M = 3.72, SD = 0.47): least confident (less than average) in citizens’ own initiatives (M = 2.90, SD = 0.51): least positive assessment of governmental influence on causes and consequences of participatory society (M = 3.27, SD = 0.78): consequences of participatory society are unequally distributed and solidarity decreasing (M = 5.56, SD = 0.54): citizens should contribute less to society than the government 4. (N = 11): “Confidence in both citizens who have to participate more in society and the government, who has to stimulate the participatory society from top-down” (M = 5.93, SD = 0.60): most confidence in citizens’ own initiatives. Problem is not that big as citizens are positive toward participating (M = 5.09, SD = 0.43): most positive evaluation of governmental influence. Still, government caused participatory society but consequences are not that bad (M = 5.30, SD = 0.57): participatory society has equal consequences and is solidary increasing (M = 2.00, SD = 0.00): citizens should participate more than the government Frame-element Problem definition Causal interpretation Moral evaluation Treatment recommendation Frames Assessment of citizens’ initiatives in terms of their capacity and willingness to participate Assessment of causes and consequences of governmental influence on the participatory society Solidarity and equality of consequences participatory society for Dutch citizens Should the government or its citizens participate more in society? 1. (N = 60): “Citizens are better able to initiate participation from bottom-up than the government. But both need to participate more in society” (M = 4.04, SD = 0.49): Some confidence in citizens’ own initiatives (development is not positive/not negative) (M = 3.45, SD = 0.60): a moderately negative assessment of governmental responsibilities and causal influence (M = 3.40, SD = 0.76): to some extent, consequences of participatory society are unequally distributed and solidarity decreasing (M = 3.77, SD = 0.52); both government and citizens need to participate more 2. (N = 55): “Trust in bottom-up citizens’ initiatives” (M = 5.04, SD = 0.86): confidence in citizens’ own initiatives (development is relatively positive) (M = 3.70, SD = 0.77): less trust in causes and consequences government. Bit negative assessment of government’s causal responsibility (M = 4.44, SD = 0.62): participatory has no unequal consequences and is not solidary decreasing (M = 2.85, SD = 0.82): citizens should participate more than the government 3. (N = 26): “Skeptical of citizens’ initiatives and the government’s influence on the unfair participatory society” (M = 3.72, SD = 0.47): least confident (less than average) in citizens’ own initiatives (M = 2.90, SD = 0.51): least positive assessment of governmental influence on causes and consequences of participatory society (M = 3.27, SD = 0.78): consequences of participatory society are unequally distributed and solidarity decreasing (M = 5.56, SD = 0.54): citizens should contribute less to society than the government 4. (N = 11): “Confidence in both citizens who have to participate more in society and the government, who has to stimulate the participatory society from top-down” (M = 5.93, SD = 0.60): most confidence in citizens’ own initiatives. Problem is not that big as citizens are positive toward participating (M = 5.09, SD = 0.43): most positive evaluation of governmental influence. Still, government caused participatory society but consequences are not that bad (M = 5.30, SD = 0.57): participatory society has equal consequences and is solidary increasing (M = 2.00, SD = 0.00): citizens should participate more than the government Table 1 Framing Matrix Used to Identify Media Frames Frame-element Problem definition Causal interpretation Moral evaluation Treatment recommendation Frames Assessment of citizens’ initiatives in terms of their capacity and willingness to participate Assessment of causes and consequences of governmental influence on the participatory society Solidarity and equality of consequences participatory society for Dutch citizens Should the government or its citizens participate more in society? 1. (N = 60): “Citizens are better able to initiate participation from bottom-up than the government. But both need to participate more in society” (M = 4.04, SD = 0.49): Some confidence in citizens’ own initiatives (development is not positive/not negative) (M = 3.45, SD = 0.60): a moderately negative assessment of governmental responsibilities and causal influence (M = 3.40, SD = 0.76): to some extent, consequences of participatory society are unequally distributed and solidarity decreasing (M = 3.77, SD = 0.52); both government and citizens need to participate more 2. (N = 55): “Trust in bottom-up citizens’ initiatives” (M = 5.04, SD = 0.86): confidence in citizens’ own initiatives (development is relatively positive) (M = 3.70, SD = 0.77): less trust in causes and consequences government. Bit negative assessment of government’s causal responsibility (M = 4.44, SD = 0.62): participatory has no unequal consequences and is not solidary decreasing (M = 2.85, SD = 0.82): citizens should participate more than the government 3. (N = 26): “Skeptical of citizens’ initiatives and the government’s influence on the unfair participatory society” (M = 3.72, SD = 0.47): least confident (less than average) in citizens’ own initiatives (M = 2.90, SD = 0.51): least positive assessment of governmental influence on causes and consequences of participatory society (M = 3.27, SD = 0.78): consequences of participatory society are unequally distributed and solidarity decreasing (M = 5.56, SD = 0.54): citizens should contribute less to society than the government 4. (N = 11): “Confidence in both citizens who have to participate more in society and the government, who has to stimulate the participatory society from top-down” (M = 5.93, SD = 0.60): most confidence in citizens’ own initiatives. Problem is not that big as citizens are positive toward participating (M = 5.09, SD = 0.43): most positive evaluation of governmental influence. Still, government caused participatory society but consequences are not that bad (M = 5.30, SD = 0.57): participatory society has equal consequences and is solidary increasing (M = 2.00, SD = 0.00): citizens should participate more than the government Frame-element Problem definition Causal interpretation Moral evaluation Treatment recommendation Frames Assessment of citizens’ initiatives in terms of their capacity and willingness to participate Assessment of causes and consequences of governmental influence on the participatory society Solidarity and equality of consequences participatory society for Dutch citizens Should the government or its citizens participate more in society? 1. (N = 60): “Citizens are better able to initiate participation from bottom-up than the government. But both need to participate more in society” (M = 4.04, SD = 0.49): Some confidence in citizens’ own initiatives (development is not positive/not negative) (M = 3.45, SD = 0.60): a moderately negative assessment of governmental responsibilities and causal influence (M = 3.40, SD = 0.76): to some extent, consequences of participatory society are unequally distributed and solidarity decreasing (M = 3.77, SD = 0.52); both government and citizens need to participate more 2. (N = 55): “Trust in bottom-up citizens’ initiatives” (M = 5.04, SD = 0.86): confidence in citizens’ own initiatives (development is relatively positive) (M = 3.70, SD = 0.77): less trust in causes and consequences government. Bit negative assessment of government’s causal responsibility (M = 4.44, SD = 0.62): participatory has no unequal consequences and is not solidary decreasing (M = 2.85, SD = 0.82): citizens should participate more than the government 3. (N = 26): “Skeptical of citizens’ initiatives and the government’s influence on the unfair participatory society” (M = 3.72, SD = 0.47): least confident (less than average) in citizens’ own initiatives (M = 2.90, SD = 0.51): least positive assessment of governmental influence on causes and consequences of participatory society (M = 3.27, SD = 0.78): consequences of participatory society are unequally distributed and solidarity decreasing (M = 5.56, SD = 0.54): citizens should contribute less to society than the government 4. (N = 11): “Confidence in both citizens who have to participate more in society and the government, who has to stimulate the participatory society from top-down” (M = 5.93, SD = 0.60): most confidence in citizens’ own initiatives. Problem is not that big as citizens are positive toward participating (M = 5.09, SD = 0.43): most positive evaluation of governmental influence. Still, government caused participatory society but consequences are not that bad (M = 5.30, SD = 0.57): participatory society has equal consequences and is solidary increasing (M = 2.00, SD = 0.00): citizens should participate more than the government The four-factor solution resulted in scales that were interpretable in terms of meaningful frame-elements. The frame-elements are labelled as following: (1) assessment of citizens’ initiatives in terms of their capacity and willingness to participate (Cronbach’s alpha = 0.84); (2) assessment of causes and consequences of governmental influence on the participatory society (Cronbach’s alpha = 0.76); (3) moral evaluation in terms of solidarity and equality of consequences the participatory society has for Dutch citizens (Cronbach’s alpha = 0.69); and (4) treatment recommendation: should the government or citizens participate more in society? (Cronbach’s alpha = 0.81). After reducing the 16 items in the codebook to the four above-described frame-elements, the next step was to analyze how articles clustered on these frame-elements. To do so, a hierarchical cluster analysis using Ward’s method with Euclidian distances was conducted on the four computed scales (Burns & Burns, 2009). Based on the qualitative study, we expected to find a solution in the region of five clusters, interpretable as meaningful frames that differ substantially on the distinguished frame-elements. We first allowed the clustering procedure to estimate a solution between 2 and 12 clusters. The dendrogram and agglomeration schedule supported a four-cluster solution as the most optimal one: merging clusters to a lower number resulted in solutions with too distant clusters, while distinguishing between a higher number of clusters resulted in clusters that were too similar. For this reason, we rerun the cluster analysis with a fixed number of four clusters. This solution was saved as a new variable, of which the four categories were reinterpreted as media frames. Results Study 2: Media Frames on the Participatory Society To interpret the four clusters as meaningful media frames, a framing matrix consisting of the four scales was constructed for each frame separately (see Table 1 for framing matrix). Based on analyses of variance (one-way ANOVA’s), significant differences between frames on their scores on the four frame-elements were interpreted. All frames differed significantly on their assessment of the four frame-elements included in the matrix (see ANOVA Table B2 of the supplemental material). The first media frame was found in 39.5% of all articles (N = 60). This frame is best described as: “Citizens are better able to initiate participation from bottom-up than the government, who is forcing participation from top-down.” This frame emphasizes some trust in citizens to participate successfully in the participatory society from bottom-up. The causes and consequences of governmental influence are interpreted with less trust. Morally reasoned, this frame emphasizes that the consequences of the participatory society are unequally distributed in society because this development excludes citizens and decreases solidarity. As treatment recommendation, this frame suggests that both citizens and the government should participate more in society. The second media frame was found in 36.2% (N = 55) of all articles and is referred to as “Trust in bottom-up citizens’ initiatives.” This frame emphasizes significantly more trust in citizens’ initiatives than the first frame, but is just as skeptical of governmentally initiated influences. In the moral evaluation, inequality does not play a role. In the treatment recommendation, this frame clearly emphasizes that citizens need to contribute more to society than the government. The third media frame is less dominant than the first two frames. In total, 17.2% of the articles (N = 26) frames the participatory society from this “Sceptical of citizens’ initiatives and the government’s influence on the unfair participatory society” media-frame. This frame emphasizes least trust in both the government’s influence on the participatory society and citizen’s initiatives to participate more in society. Inequality of consequences and social exclusion are emphasized in the moral evaluation. As treatment recommendation, the government rather than citizens should participate more in society. The final frame found in the media discourse concerns a pure bottom-up frame that emphasizes trust in citizens’ ability and willingness to participate more in society. Furthermore, the government’s influence on causes and consequences of the participatory society is assessed with great confidence. This frame recommends that citizens should participate substantially more than the government. This frame was found in only 7.2% of all articles (N = 11) and challenges the position emphasized in the third frame. This frame can be described as: “Confidence in both citizens, who have to participate more in society, and the government, who has to stimulate the participatory society from above.” Conclusion and Discussion Study 2 The four frames found in this study can be reduced even further to two more generic frames resonating in the media discourse. First, a dominant bottom-up citizens’ initiative frame, in which citizens’ willingness and ability to participate is made salient and the freedom to participate and take matters into own hands are highlighted as virtue of modern citizenship, regardless of the economic recession. Second, the top-down “governmentally forced participatory society frame” functions as a counter frame. This frame defines the participatory society as an inevitable result from a government that is no longer able to carry the consequences of the welfare state. In general, the media discourse articulated a lack of trust in the government to stimulate participation. Study 3: Quantitative Analysis of Interpretation Frames Method Data collection. The framing analysis of this study is based on responses to a set of questions included in an opinion poll, presented as an online survey executed by a research organization. This question block consisted of 18 statements on the participatory society. The statements were formulated on a 7-point semantic differential scale. The statements were based on the qualitative results of Study 1 and aimed to capture the variety of perspectives found on all four frame-elements. In addition, more recent developments of the debate in public opinion after the earlier rounds of data collection were taken into account. To provide an example, the distinction between helping family versus helping strangers became a fierce topic of debate as identified in various offline and online media, which we incorporated in the measures of the different frame-elements. In the framing-matrix, all statements measuring the four frame-elements are included (Appendix C2 of the supplemental material). Sample. The sample of this study consisted of 999 respondents selected from the panel of the research organization. The sample is representative of the Dutch voting population in terms of gender, age, educational level, region, family size, social class, and voting behavior. All participants in the study were aged >18 years. The response rate of this panel is 70%. Respondents were invited to participate in the research via e-mail, letter, and/or telephone. The mean age of participants was 50 years (SD = 17.40), and 47.5% of the participants were female. Data analysis. We conducted a confirmatory factor analysis to assess whether the statements fitted a four-dimensional structure consisting of the four frame-elements. As indicated by the suboptimal model fit, we did not find support for this solution [χ2(118) = 754.50, p < .001, Confirmatory Fit Index (CFI) = 0.866, root mean square error of approximation = 0.074, 90% confidence interval (0.069; 0.079)]. We therefore did not reduce the variables to scales. Next, we identified a meaningful number of interpretation frames by conducting a hierarchical cluster analysis using Ward’s method with Euclidian distance (Burns & Burns, 2009; Matthes & Kohring, 2008). The framing analyses in Study 1 and Study 2 indicated that a number of four or five frames on a certain topic are an appropriate solution, which allows for the presence of both dominant and counter frames. Similar to Study 2, we allowed the clustering procedure to estimate a solution ranging between 2 and 12 clusters. The agglomeration schedule and dendrogram confirm the presence of five clusters with a meaningful overall distance between statements that could also be interpreted substantially. We compared the five-cluster solution to both higher and lower numbers of clusters. Merging clusters to a lower number resulted in solutions with too distant clusters, while distinguishing between a higher number of clusters resulted in clusters that were too similar. Based on this outcome, the cluster analysis was rerun with a fixed solution of five clusters, which resulted in a new variable with five categories. These categories grouped the participants into five different interpretation frames. By constructing a similar framing matrix as Study 1 and Study 2, these five clusters were interpreted as interpretation frames. More specifically, we compared the mean scores on all statements between clusters to systematically reconstruct the frames per frame-element. In one interpretation frame, for example, participants expressed a negative problem definition, and blamed the government in their causal interpretation (see Appendix C2 in the supplemental material for complete framing matrix). Results Study 3: Interpretation Frames in the Public Discourse As the identification of significant mean differences on statements is a necessary prerequisite to validate discriminant clusters, analyses of variance (one-way ANOVA) were conducted on the five identified clusters or “frames” (Burns & Burns, 2009). As depicted in ANOVA Table C1 of the supplemental material, all between-group means differed significantly (p < .05), which implies that the 18 statements reliably distinguished between five significantly different interpretation frames. Similarities between frames were interpreted as meaningful as well, as not all frames necessarily differ on all separate parts of all four frame-elements (e.g., Entman, 1993). Similar to Study 1 and Study 2, a framing matrix consisting of all four frame-elements was constructed (see supplemental material, Appendix C2). The mean values of the statements and the significant differences between these values across the frames guided the interpretation of the clusters into frames. The first frame is labelled as “Necessity is the mother of invention when it comes down to our readiness for participation.” This is the most dominant frame of reference carried by 31.0% of the participants (N = 309) and represents a balanced evaluation of the participatory society that is neither positive nor negative. Citizens who interpret the participatory society from this interpretation frame do not reject the participatory society as societal development. However, they do not experience a great willingness or ability to participate more in society if the government is withdrawing from society. Furthermore, clear lines are drawn in terms of readiness for participation in society. Citizens feel responsible to help a neighbor in need, for example, by giving him or her a ride to the hospital. Helping a neighbor in need of care in less urgent, everyday situations, such as getting up in the morning, is too demanding for these citizens. The second frame, referred to as “Despite feeling responsible for others, we are unwilling and unable to participate more in society” is held by 28.0% of the participants (N = 280). Reasoned from this interpretation frame, the participatory society is a negative societal development, especially for weaker groups. Moreover, citizens lack the capacity and willingness to participate more. They do experience feelings of responsibility for others, but are only willing and able to act on these feelings in emergency situations. Health care and neighborhood security are tasks for the government, which should not be transferred to citizens’ responsibilities. If the government retreats, especially weaker groups in society will be harmed. This interpretation frame counters the generic “Bottom-up citizens’ initiative frame” found in Study 2 as media frame and in Study 1 as interpretation frame. The third frame is labelled as “In times of decreasing solidarity, citizens are responsible to take care of each other in times of need, which is especially important for weaker groups in society.” This interpretation frame, held by 28.0% of the respondents (N = 280) emphasizes an even stronger sense of responsibility (M = 5.80, SD = 1.15) than articulated in Frames 1 and 2. This frame is similar to the bottom-up frames found in Study 1 that emphasized citizens’ willingness to participate in society. This frame differs from the bottom-up frames in the qualitative study in the sense that citizens experience a willingness to participate, but at the same time emphasize they are not always able to do so in the causal interpretation and moral evaluation. As treatment recommendation, citizens emphasize that the government needs to stimulate participation. The preferred role of the government as safety net selected in the moral evaluation and treatment evaluation is congruent with the limits to participation highlighted in the bottom-up frames identified in the qualitative study. The fourth frame is labelled as “The participatory society is a negative societal development for all citizens who are willing to help others in need but lack the capacity and desire to participate more in society.” This frame is rare: only 8.0% (N = 76) of all participants interpreted the participatory society from this frame of reference. Reasoned from the moral evaluation of this frame, all groups in society will be affected equally by the participatory society (M = 5.83, SD = 1.15), which is in sharp contrast with all other frames that emphasize that some groups in society will be affected more than others. Furthermore, this interpretation frame highlights the willingness of citizens to incidentally take a neighbor to the hospital (M = 5.92, SD = 1.28). However, similar to Frames 1 and 2, this moral evaluation does not result in more willingness to help a neighbor with daily care of getting dressed: (M = 2.74, SD = 1.53). Hence, participation is reserved for emergency situations. The fifth frame is carried by 5.0% (N = 54) of the respondents and is referred to as “We feel responsible and are able to participate more in society, but weaker groups lack this capacity completely.” Despite the problem definition and causal interpretation emphasize the participatory society as a solidarity decreasing development, citizens are both able and willing to participate more in society. This frame counters the dominant frame “Despite feeling responsible for others, we are unwilling and unable to participate more in society.” This frame represents the only pure bottom-up interpretation, as it highlights both the ability and willingness to participate in different domains of society, not only in emergency situations. Despite the emphasis on willingness and ability to participate foregrounded in the causal and moral interpretations, citizens holding this frame are still negative about the participatory society as introduced by the government in their problem definition (M = 1.57, SD = 1.35), which underlines the moral evaluation that especially weaker groups are harmed by the participatory society. Conclusion and Discussion Study 3 The largest overall differences in the interpretation of statements across all four frame-elements were found between the second frame “Despite feeling responsible for others, we are unwilling and unable to participate more in society” and the fifth frame “We feel responsible and are able to participate more in society, but weaker groups lack this capacity completely.” The fifth frame can therefore be interpreted as the counter frame of the second frame. Interpreting the participatory society from the dominant second frame, a stronger inability and unwillingness to participate in society is experienced than in the fifth frame, which articulates a strong belief of citizens being able and willing to participate in society when needed. For example, respondents interpreting reality from the second frame believe they are unable to look out for the safety in their neighborhood if the police will contribute less to this task (M = 2.18, SD = 1.36). This experienced ability is significantly and substantially higher for the fifth frame (M = 5.83, SD = 1.55); Mdifference = 3.66, p < .001. At this stage, we can draw conclusions on the discrepancy between media frames and interpretation frames (see Figure 1). The largest difference concerns the amount of trust attached to citizens’ initiatives. Both the media and public discourse emphasized a lack of trust in the government who forced the scheme of participation on citizens. Still, by covering successful bottom-up citizens’ initiatives, the media inherently articulated more trust in citizens’ initiatives than citizens highlighted themselves. The media frames and interpretation frames also demonstrated overlap. The dominant media frame, which emphasized citizens’ willingness and ability to participate, resonates with bottom-up interpretation frames held by some citizens. However, this frame was far less frequently articulated by citizens than the media. We interpret this as support for framing within the negotiated media effects paradigm (e.g., McQuail, 2010, Scheufele, 1999; Scheufele & Tewksbury, 2007). The media may propose certain frames to receivers, and receivers still possess the agency to resist and adjust the meaning of these frames in line with their own frame of reference. Figure 1 View largeDownload slide Graphical depiction of the overlap and discrepancies between interpretation frames (Study 1 and Study 3) and media frames (Study 2). Arrows between boxes indicate overlap between frames, and arrows on the left and right side indicate overlap within the receiver- or sender-side. When boxes are not connected, discrepancies are highlighted. The strength of the overlap decreases from top to bottom. Figure 1 View largeDownload slide Graphical depiction of the overlap and discrepancies between interpretation frames (Study 1 and Study 3) and media frames (Study 2). Arrows between boxes indicate overlap between frames, and arrows on the left and right side indicate overlap within the receiver- or sender-side. When boxes are not connected, discrepancies are highlighted. The strength of the overlap decreases from top to bottom. Overall Discussion The aim of this research was to assess discrepancies between media frames and interpretation frames. The introduced method responded to recent calls in the literature, which suggest (1) to acknowledge the social–cultural context and power relations underlying the framing process (Carragee & Roefs, 2004) (2) to consider frames as a cluster of frame-elements and reasoning devices (e.g., Matthes & Kohring, 2008; Van Gorp, 2007), and (3) to identify interpretation frames and media frames using a similar analytical approach to ensure a valid comparison of both interrelated framing processes (Matthes & Kohring, 2008). In three studies, this research aimed to validate the introduced methodological framework departing from a social-constructivist viewpoint. The overarching research question that focused on the discrepancies between media frames and interpretation frames was: To what extent and how does the framing of the participatory society differ between interpretation frames of citizens and media frames of Dutch news media? Our research shows that media discourse and public discourse both emphasized a clear distinction between bottom-up frames and top-down frames on the participatory society but differed in their evaluation of these top-down and bottom-up frames and their consequences for Dutch society. The interpretation frames identified in Study 1 and Study 3 emphasize a lack of trust in citizens’ bottom-up initiatives, whereas the media frames identified in Study 2 predominantly emphasize the success stories of bottom-up initiated examples of the participatory society. Concerning the top-down frames, both media and citizens emphasized a lack of trust in the feasibility of the forcing schemes promoting the participatory society from a governmental perspective (Scott, 1998). Furthermore, in both the media and public discourse, the capacity of Dutch society to carry the consequences of the participatory society was questioned. Interpretation frames and media frames both emphasized that especially weaker groups in society will be affected by the consequences of the participatory society. Owing to the exploratory nature of the studies incorporated in this research, some limitations regarding the design of the studies and the operationalization of the research materials need to be highlighted. First, the second study has exclusively focused on media frames present in Dutch newspapers. Dutch citizens draw on an extensive stock of sociocultural resources, and newspapers only represent one of these resources among many competing ones. Future research should include more of these media-based sociocultural resources to arrive at a more complete comparison between media frames and interpretation frames. The second limitation concerns the operationalization of frame-elements incorporated in the content analysis and subsequent cluster analyses. Some variables link up to more than one single frame-element, which may affect the identification of frames as clusters of frame-elements. However, the development of variables was informed by the in-depth insights of the qualitative study. Moreover, we ensured that the variables were attached to the most suitable frame-element. In any case, the study has both clear methodological and theoretical implications. Methodologically, the study demonstrates the importance to consider media framing and public framing as related, yet conceptually distinct phenomena. Instead of assuming that frames present in media coverage will also be the ones that are present among members of the public, it is worth to analyze both realms separately and explicitly question similarities and differences in those frames. Moreover, it demonstrates how qualitative and quantitative methods can be complementary in one study. Theoretically, the findings offer support against the idea of strong framing effects. Although the media may put forward a certain set of frames to understand political issues, members of the audience will only partially adopt this set of frames: based on, for example, different types of information and personal experiences, they will resist, adjust, and supplement the stock of frames put forward by the media. This implies that a situation of perfect alignment between media framing and audience framing is highly unlikely. This finding fits well with the negotiated media effects paradigm (Gamson & Modigliani, 1989). Overall, a structural and detailed approach as presented in this study can unravel differences and similarities in the presence of frames and offers a starting point for understanding media effects for a plethora of issues in a wide variety of contexts. Michael Hameleers is PhD candidate in Political Communication at the University of Amsterdam (ASCoR). The main topic of his dissertation research is populism in the media. His research interests include framing, mediated populism, and the role of social identity in media effects. Rens Vliegenthart is professor for Media and Society in the Department of Communication Science at ASCoR at the University of Amsterdam. His research interests include media–politics interactions, media-effects, and comparative politics. Acknowledgments The findings presented in this article resulted from an intensive cooperation between the University of Amsterdam (ASCoR), Veldkamp, and TNS NIPO. The different parties had no financial interests in the outcomes of this study. The mutual interests of the involved parties were exclusively based on the development of a methodological tool for framing analysis applicable in both scientific and market research. Supplementary Data Supplementary Data are available at IJPOR online. References Burns R. , Burns R. ( 2009 ). Chapter 23: cluster analysis. In Business research methods and statistics using SPSS (pp. 552 – 567 ). California : Sage Publications . Bryman A. ( 2012 ). Social research methods . Oxford : Oxford University Press . Carragee K. M. , Roefs W. ( 2004 ). The neglect of power in recent framing research . Journal of Communication , 54 , 214 – 233 . doi: 10.1111/j.1460-2466.2004.tb02625.x Google Scholar CrossRef Search ADS Entman R. M. ( 1993 ). Framing: Toward clarification of a fractured paradigm . Journal of Communication , 43 , 51 – 59 . doi: 10.1111/j.1460-2466.1993.tb01304.x Google Scholar CrossRef Search ADS Flick U. ( 2009 ). An introduction to qualitative research ( 4th ed .). London : Sage Publications . Gamson W. A. , Modigliani A. ( 1989 ). Media discourse and public opinion on nuclear power: A constructionist approach. American Journal of Sociology , 95 ( 1 ), 1 – 37 . doi: 10.1086/229213 Google Scholar CrossRef Search ADS Goffman E. ( 1974 ). Frame analysis: An essay on the organization of experience . Cambridge : Harvard University Press . Glaser B. G. , Strauss A. L. ( 1967 ). The discovery of grounded theory: Strategies for qualitative research . Chicago : Aldine . Hall S. ( 1973 ). Encoding/decoding. In Hall S. , Hobson D. , Lowe A. , Tillis P. (Eds.), Culture, media, language (pp. 128 – 138 ). London : Hutchinson . Holstein J. A. , Gubrium J. F. ( 1995 ). The active interview. Thousand Oaks : Sage . Google Scholar CrossRef Search ADS Huang H. ( 2010 ). Frame-rich, frame-poor: An investigation of the contingent effects of media frame diversity and individual differences on audience frame diversity . International Journal of Public Opinion Research , 22 , 47 – 73 . doi: 10.1093/ijpor/edp024 Google Scholar CrossRef Search ADS Lombard M. , Snyder-Duch J. , Bracken C. C. ( 2002 ). Content analysis in mass communication: Assessment and reporting of intercoder reliability . Human Communication Research , 28 , 587 – 604 . doi: 10.1111/j.1468-2958.2002.tb00826.x Google Scholar CrossRef Search ADS Matthes J. , Kohring M. ( 2008 ). The content analysis of media frames: Toward improving reliability and validity . Journal of Communication , 58 , 258 – 279 . doi: 10.1111/j.1460-2466.2008.00384.x Google Scholar CrossRef Search ADS McLafferty I. ( 2004 ). Focus group interviews as a data collecting strategy . Journal of Advanced Nursing , 48 , 187 – 194 . doi: 10.1111/j.1365-2648.2004.03186.x Google Scholar CrossRef Search ADS PubMed McQuail D. ( 2010 ). McQuail’s mass communication theory ( 6th ed .). London : Sage . Neuman W. R. , Just M. R. , Crigler A. A. ( 1992 ). Common knowledge. News and the construction of political meaning . Chicago : The University of Chicago Press . Reese S. D. ( 2001 ). Introduction. In Reese S. D. , Gandy O. H. , Grant A. E. (Eds.), Framing public life: Perspectives on media and our understanding of the social world (pp. 1 – 31 ). Mahwah, NJ : Erlbaum . Riffe D. , Lacy S. , Fico F. ( 2005 ) Analyzing media messages; using quantitative content analysis in research ( second edition). New Jersey : Mahwah Scheufele D. A. ( 1999 ). Framing as a theory of media effects . Journal of Communication , 49 , 103 – 122 . doi: 10.1111/j.1460-2466.1999.tb02784.x Google Scholar CrossRef Search ADS Scheufele D. A. , Tewksbury D. ( 2007 ). Framing, agenda setting, and priming: The evolution of three media effects models . Journal of Communication , 57 , 9 – 20 . doi: 10.1111/j.0021-9916.2007.00326.x Scott J. C. ( 1998 ). Seeing like a state. How certain schemes to improve the human condition have failed . New Haven and London : Yale University Press . Semetko H. A. , Valkenburg P. M. ( 2000 ). Framing European politics: A content analysis of press and television news . Journal of Communication , 50 , 93 – 109 . doi: 10.1111/j.1460-2466.2000.tb02843.x Google Scholar CrossRef Search ADS Tankard J. W. ( 2001 ). The empirical approach to the study of media framing. In Reese S. D. , Gandy O. H. , Grant A. E. (Eds.), Framing public life: Perspectives on media and our understanding of the social world (pp. 95 – 106 ). Mahwah, NJ : Erlbaum . Van Gorp B. ( 2007 ). The constructionist approach to framing: bringing culture back in . Journal of Communication , 57 , 60 – 78 . doi: 10.1111/j.0021-9916.2007.00329.x Vliegenthart R. , Van Zoonen L. ( 2011 ). Power to the frame: Bringing sociology back to frame analysis . European Journal of Communication , 26 , 101 – 115 . doi: 10.1177/0267323111404838 Google Scholar CrossRef Search ADS Wicks R. H. ( 2001 ). Understanding audiences: Learning to use the media constructively . Mahwah, NJ : Erlbaum © The Author 2016. Published by Oxford University Press on behalf of The World Association for Public Opinion Research. All rights reserved. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png International Journal of Public Opinion Research Oxford University Press

Framing the Participatory Society: Measuring Discrepancies Between Interpretation Frames and Media Frames

Loading next page...
 
/lp/ou_press/framing-the-participatory-society-measuring-discrepancies-between-bhG22LQZWN
Publisher
Oxford University Press
Copyright
© The Author 2016. Published by Oxford University Press on behalf of The World Association for Public Opinion Research. All rights reserved.
ISSN
0954-2892
eISSN
1471-6909
D.O.I.
10.1093/ijpor/edw032
Publisher site
See Article on Publisher Site

Abstract

Abstract This research introduces a methodological framework for a valid and reliable analysis of both media and interpretation frames. It responds to recent calls in literature (1) to acknowledge the social–cultural context and power relations underlying framing, (2) to consider frames as a cluster of frame-elements, and (3) to identify interpretation frames and media frames using a similar analytical approach to ensure a valid comparison of the outcomes of both interrelated framing processes. The methodological framework was applied to three studies. Results show that the discrepancy between interpretation frames and media frames revolved around the distinction between bottom-up and top-down frames. The introduced method provides meaningful insights into the complete framing process of both the sender and receiver of communication messages. During the King’s speech held on September 17, 2013, the Dutch government coined the idea of the “participatory society” as a new form of society intended to replace the classical Dutch welfare state. In the participatory society, each citizen who is able to participate will be asked to take more responsibility, especially regarding community care and social security. This societal development received a lot of resistance from citizens and became topic of fierce political debate in the subsequent period. The resistance articulated by the people was inherently different from the way the government expressed its views on the participatory society. The government’s frame emphasized a certain definition of the problem that empowered citizens as being in charge of their own fate. It also emphasized the causes for a shift from the classical welfare state to the participatory society. By stating that Dutch citizens should contribute more to their society as a whole, a certain moral evaluation was emphasized. Finally, by arguing that citizens need to increase their participation in different fields of society, a recommendation on how to “treat” the participatory society was introduced. In other words, all four elements constituting a frame were present in the King’s speech (Entman, 1993). Framing research in communication science has mainly focused on the sender-side of communication, predominantly by using a quantitative or qualitative content analysis to distil frames out of media content (e.g., Semetko & Valkenburg, 2000; Tankard, 2001). The receiver-side of framing has mainly been analyzed as the effects of framed communication. Carragee and Roefs (2004) argue that this traditional approach neglects the origins of frames. In their opinion, framing research should be related to underlying sociocultural processes, as the distribution of power in society influences both the interpretation and construction of frames. In this reading, framing is defined as a process of social reality construction that involves both receivers and senders of news stories. As social reality is constructed within a specific meaning-loaden context, the interpreting receiver of frames should be assigned a central role in the analysis of frames (Gamson & Modigliani, 1989). Previous studies are inconclusive on how frames should be operationalized in a valid and reliable way, both on the sender- and receiver-side (e.g., Matthes & Kohring, 2008). Reviewing five prevalent approaches in framing research, Matthes and Kohring point to two major issues for reliability and validity. First, inductive interpretative framing approaches lack reliability. Predominantly, these approaches provide little transparency in how frames are extracted from the data. Furthermore, these approaches are frequently based on an identification of frames as single entities by coders. This implies that the identification of a set of frames in the text becomes an arbitrary practice, which harms the reliability and validity of the analysis (see also Tankard, 2001). Second, deductive, often systematic computer-assisted approaches, may lack validity. As frames are embedded in both media and public discourses, the extraction of frames cannot be reduced to a set of computer-coded variables. Furthermore, the deductive approach is limited, as it presupposes that all frames are known beforehand, which overlooks the identification of novel frames that are not anticipated by the researchers. So far, comprehensive methodological tools for a valid and reliable analysis of the complete interactive framing process are still lacking (e.g., Carragee & Roefs, 2004; Entman, 1993; Scheufele, 1999). By responding to conceptualizations and suggestions for future research by Matthes & Kohring (2008) and Van Gorp (2007), this article first of all proposes a standardized method for the analysis of sender-side media frames and receiver-side interpretation frames that strives toward validity and reliability. Second, this method is validated by analyzing how the Dutch participatory society is framed by Dutch news media and interpreted by Dutch citizens. The general aim of this study is therefore to explore the discrepancy between media frames and interpretation (audience) frames using a systematic method of framing analysis. To precisely study the discrepancy and overlap between media frames and interpretation frames, three studies were conducted. In the first study, interpretation frames were analyzed qualitatively. In the second study, frames operating in the media discourse were identified quantitatively. Based on qualitative insights, the third study analyzed interpretation frames by means of a quantitative cluster analysis. The Sender-Side of Framing: Media Frames and Their Operationalization The process of framing entails the reconstruction of social reality in meaningful patterns of interpretation resonating in cultural values shared by both sender and receiver (Gamson & Modigliani, 1989). These patterns of interpretation manifest themselves in a communication text as story lines and organizing ideas, which guide both the sender and receiver of communication in their process of making sense of (news) events (Huang, 2010). An important question for framing analysis is how frames can be identified validly in a text. Entman’s (1993) conceptualization of frames as consisting of four distinct elements provides more insight into how frames manifest themselves in communication texts. First, frames promote a particular problem definition by highlighting what the issue is about, what the costs and benefits of the issue might be, and which actors are involved in the issue. Second, the causal interpretation emphasizes the causes and consequences of the issue and identifies who is responsible for these causes and effects. Third, the moral evaluation of the problem evaluates the causes and effects, and connects these to moral values and norms familiar in the cultural context. Finally, the treatment recommendation suggests potential solutions to the problem. The operationalization of framing can be made specific by regarding the four elements and locations as a “frame package” embedded in a communication text (Gamson & Modigliani, 1989; Van Gorp, 2007). By reconstructing this package in a text, the frame can be identified in a systematic and objective way. The frame package consists of three different parts that together represent the frame. First, the manifest framing devices refer to arguments, visual images of the issue, examples of the issue, metaphors, and word choice. These framing devices can be coded in a text as manifest characteristics of the frame package (Van Gorp, 2007). The cement that holds the manifest framing devices together is the frame itself. It provides an overarching meaning structure to the manifest components of the frame. Second, at a more latent level, frames consist of reasoning devices (Gamson & Modigliani, 1989; Van Gorp, 2007). These reasoning devices reflect the four elements of the frame as conceptualized by Entman (1993). The reasoning devices in a text can be identified as the aggregate of implicit and explicit references that promote a particular problem definition, causal interpretation, moral evaluation, and/or treatment recommendation. Third, as frames are embedded in cultural contexts, they reflect broader cultural themes shared by the receiver and sender of frames (Van Gorp, 2007). This thematic component is more difficult to code in an objective way, as it requires the coder to be familiar with the underlying cultural values, norms, and stereotypes. Moreover, attaching cultural themes to a frame package is an arbitrary practice that largely depends on researchers’ interpretation. Here, it is important to note that talking about the media as sender thus not inherently implies that the agency of the individual journalist is unbounded. On the contrary, media framing is constructed within the cultural context in which it is embedded, with multiple institutional and noninstitutional actors trying to influence the content of coverage, especially when it comes to politically contested issues. Moreover, news values and journalistic practices constrain the individual journalist (see Vliegenthart & Van Zoonen, 2011). Reflecting on the above-presented operationalization of framing put forward in previous studies, the main focus of this research is to (1) systematically identify reasoning devices as frame-elements; to (2) reconstruct these elements into frame packages, both as media frames and interpretation frames; and to (3) attach meaningful themes as labels to these frame packages. The qualitative study, in particular, aims to provide an in-depth understanding of the cultural values underlying the frames. The measurement tool that is designed to identify frames in a systematic way is outlined in detail in the method section. The Receiver-Side of Framing: Interpretation Frames in the Public Discourse Social constructivism postulates that the media interpret and emphasize issues in a different way compared with the active meaning-constructing receivers of these media messages (Holstein & Gubrium, 1995; Neuman, Just & Crigler, 1992; Wicks, 2001). Previous research differs in the locus of power ascribed to senders (media) versus receivers. Some authors argue that receivers choose from a variety of persistent frames that are made available to them by the media (e.g., Reese, 2001). In this reading, the media are powerful in the sense that they control the social reality construction process of receivers by forcing them to choose from a limited number of alternative interpretations. A relevant question is whether framing is indeed as powerful as suggested in the “powerful media effects paradigm” (McQuail, 2010). Some authors challenge this paradigm by ascribing more agency to receivers, who construct their own interpretation of issues in interaction with media makers and other social actors, rather than being controlled by powerful media forces (e.g., Scheufele, 1999). This study ties in with this reading by emphasizing that both media senders and receivers are part of the interpretation process underlying framing, as both construct their vision on reality out of the same stock of social and cultural resources. The paradox in previous literature on the receiver-side of framing is that on one hand, framing is approached from a social-constructivist perspective as the result of cultural actors’ interpretations, while at the other hand, frames are viewed as external to the individual and should therefore be analyzed independently from the actors as part of cultural resources. This study takes on an intermediate perspective on this issue by acknowledging the power of the sender of frames to propose a preferred reading of the participatory society while at the same time assigning a central role to the agency of the audience. In this reading, receivers process information in a motivated way instead of simply accepting or rejecting media frames on the participatory society (Gamson & Modigliani, 1989). People are thus not only active in resisting frames that are counter to their attitudes and accepting frames that are consistent with their attitudes, they construct their own individual interpretation of issues as well. Therefore, traditional approaches to audience-side framing that measure people’s interpretation simply as the effects of framed communication are insufficient in capturing the complete meaning construction process of the audience. Bridging the Gap: Framing as Interactive Practice Between Sender and Receiver Entman (1993) argued that the dominant frame embedded in communication texts on a certain issue should be congruent with the most common interpretation frames among receivers. However, Neuman et al. (1992) found that media frames and audience frames are not by definition congruent, and alternative frames not present in media coverage may resonate in the public discourse, also because citizens have a plethora of alternative information sources at their disposal (Gamson and Modigliani, 1989). In other words, not all interpretation frames are “out there” to be analyzed directly from media content. Although the audience-side of framing has been analyzed in framing effect studies, the discrepancies and linkages between media frames and interpretation frames analyzed separately on both sides of the communication process have not received extensive attention (Scheufele, 1999). Building on Scheufele’s (1999) conceptualization of framing as a cyclical process between senders and receivers, the perspective on framing taken in this study acknowledges two divergent but interrelated processes of framing. First, news media frame issues in a certain way, selecting aspects of social reality, and reconstructing these aspects in a salient way (Entman, 1993; Goffman, 1974; Van Gorp, 2007). Second, citizens decode the framed communication text in a way that is meaningful within their own frame of reference. If the media frame is congruent with their own viewpoint, citizens are expected to accept the frame as embedded in the communication text (Entman, 1993). However, if citizens are involved with the issue and the media frame is dissonant with their viewpoint on the issue at stake, the media frame will be rejected or adjusted. In contrast to the way the power of frames is referred to in traditional literature on framing as media effects, the two extremes of acceptance and rejection should not be regarded as the single two possible receivers’ interpretations. As Hall (1973) postulated, a middle position is also feasible—in which the decoding stance is mixed regarding acceptance and rejection of the connoted meanings. This position partly accepts the legitimacy of the dominant media discourse, but adapts it according to the situational context. The threefold (accept/mixed/reject) conceptualization of decoding positions hypothesized by Hall (1973) is in line with the empirical findings of Neuman et al. (1992), Gamson & Modigliani (1989), and Huang (2010). Acceptance and rejection represent the two ends on a continuum of possible responses toward each of the four elements that constitute a frame. Against this backdrop, this study introduces and validates a methodological tool for measuring media frames and interpretation frames in a similar but independent way, making similarities and differences clear. The substantial research question we pose reads as follows: RQ: To what extent and how does the framing of the participatory society differ between interpretation frames of citizens and media frames of Dutch news media? All empirical studies in this article are concerned with answering (parts of) this research question. Study 1 aims to provide insights into the framing process from the perspective of the receiver by means of a qualitative framing analysis of interviews and focus groups. By means of a quantitative content analysis, Study 2 aims to provide insights into the media frames used to frame the participatory society in Dutch newspapers. Study 3 is devoted to validate the inductively gathered interpretation frames of Study 1 by means of a quantitative cluster analysis of interpretation frames. The framing analysis in all three studies is based on a similar measurement instrument developed for this study. The characteristics of this framing measurement instrument are outlined in greater detail in the following section. Methodological Framework Responding to Matthes and Kohring’s (2008) suggestion for future research to measure media frames and audience frames with a similar method that clusters manifest variables together as frames, we used the in-depth analysis of qualitative material in Study 1 to develop a framing measurement tool that was applied to quantitative cluster analyses of both media frames and audience frames. The five steps that were used to qualitatively develop the measurement tool are discussed below. The first step of analysis resembles open coding in the grounded theory approach (Bryman, 2012; Flick, 2009). Regarding the four frame-elements or reasoning devices problem definition, causal interpretation, moral evaluation, and/or treatment recommendations as sensitizing concepts, fragments in the transcripts that reflected these concepts were highlighted. Open codes were used to code the reasoning devices that were captured in a certain fragment. During the second step of analysis, the open codes were structured around the four frame-elements. This step resembles axial coding in the grounded theory approach, for which the “axes” were represented by the four frame-elements. By completing the codebook for each respondent individually, the step from open coding to axial coding resulted in the development of structured themes and categories derived from relatively unique open codes. The codebook was revised until theoretical saturation was achieved (Glaser & Strauss, 1967). More specifically, after new rounds of analysis of interview transcripts no longer resulted in the emergence of new questions (codes), the codebook included all aspects of which the reasoning devices (frame-elements) are constructed. During the third step of analysis, the final version of the codebook was translated into a framing matrix (see Van Gorp, 2007). Categories were related to one another during this step (Bryman, 2012). Specifically, the four reasoning devices or frame-elements were connected to form frame packages in a framing matrix. The columns of the matrix represented the four reasoning devices, in which the different questions for each reasoning device were synthesized to one underlying theme, the frame-element. The rows of the matrix represented individual participants. During the fourth step of analysis, the individual frame packages were compared for similarities and differences to extract frames from the individual packages. This step inherently contains a subjective element, as the researchers have to attach the frame packages to social–cultural values and meanings. However, by staying close to the original frame packages in the clustering process, the researchers’ bias was minimized. To enhance the richness of frames and to maintain the embeddedness of frames in their social–cultural context, frames were illustrated with in-verbatim quotes of participants during the fifth step of analysis. The above-mentioned procedures for the framing analysis formed the core of the framing measurement tool designed in this study. In the next step, the results from the qualitative analysis were used to develop statements reflecting opposing positions on the different frame-elements. These statements were measured on 7-point scales (see Appendix C1 in the supplemental material for statements). These items reflected the qualitatively identified variety of different positions on all frame-elements and provided the input for the cluster analysis of media frames and interpretation frames (Study 2 and Study 3). Our approach ensured that we did not “feed” the quantitative analyses with a limited selection of predeveloped frames. Instead, we reduced participant’s interpretations to opposing positions on statements measuring the four frame-elements. Participants’ qualitatively identified interpretations of the participatory society, for example, revolved around different top-down and bottom-up views on how to deal with the government’s request to increase participation, which were reduced to three statements measuring the treatment recommendation in the quantitative phase of the research. Study 1: Qualitative Inductive Analysis of Interpretation Frames Method Sample. First, as a prestudy, we conducted a qualitative content analysis on transcripts of individual interviews and focus group discussions, originally not collected for the purpose of this study (N = 40). This sample was used to inductively identify frame-elements, which informed the measurement of frames in subsequent studies. This pilot study aimed to map the development of the debate on the participatory society in public opinion by analyzing interviews conducted at various stages of the debate. The pilot study further aimed for maximum variation by also including transcripts from interviews on a broader topic: citizens’ general views on and satisfaction with society. The open-ended analyses of this rich material resulted in a variety of interpretations on all four frame-elements. The data collection was ended when the analysis of new material did not result in novel themes. This exploratory content analysis was used to develop an interview guide for primary data collection. For the data collection of this inductive analysis, participants were sampled based on purposive sampling (Bryman, 2012). The inclusion criteria aimed to reflect maximum variety regarding age, gender, region, and educational level. The composition of the focus groups was homogenous with respect to educational level to avoid the influence of unequal levels of knowledge and power relations on the experiences shared within groups (McLafferty, 2004). In total, 12 participants participated in individual interviews. The mean age of participants was 47 years old, ranging from 23 to 68 years old. Half of participants were female. The two focus groups consisted of seven participants each. The mean age of participants in the focus groups was 47 years old, ranging from 24 to 66 years old. In total, 8 of 14 participants in the focus groups were female. Data collection. The semi-structured interview guide is included in the supplemental material (Appendix A1). The interviews and focus group discussions were held in May 2014. All conversations were audio–video taped and transcribed in verbatim. The average duration of the individual interviews was 40 min. The focus group discussions lasted 90 min. Data analysis. The five-step framing analysis measurement tool outlined in the methodological framework was used for the qualitative analysis of interpretation frames. The framing matrices including the four frame-elements are available on demand. Intercoder agreement was assessed for a randomly selected sample of six interview transcripts and one focus group discussion (50%). Two coders independently coded this subsample using the framing matrix. Differences were discussed until agreement on all four frame-elements was reached. The initial differences between independent analyses were marginal, and did not have implications for the outcomes of the framing-analysis as reported in the results section. Results Study 1 Based on the input of the framing-matrix, five interpretation frames were identified in the in-depth interviews and focus groups. In Appendix A2 of the supplemental material, a detailed description of all frames with in-verbatim quotes is included. The dominant interpretation frame can be labelled as Participation is bad when forced upon us from above. In their problem definition, participants emphasized that participating in society is a positive principle, as it may enhance solidarity. In their causal interpretation, participants related the participatory society entirely to economic savings. Morally reasoned, participants emphasized their limited willingness and ability to participate more. Furthermore, participants believed that the participatory society has negative consequences for Dutch society. As treatment recommendation, participants suggested that the government should not force citizens to participate. By allowing them more freedom of choice, citizens will decide for themselves what they are willing and able to do besides their responsibilities in busy everyday life. The second frame that counters the dominant top-down interpretation frame touches on an exclusionist view of society. This frame can be labelled as “we” are more willing than “they” are. This frame emphasizes that, besides top-down initiated austerity measures, the decreasing solidarity in Dutch society has caused the shift toward the participatory society from bottom-up. A need for Dutch citizens to look out for each other is emphasized in the moral evaluation. This desire for citizens’ initiatives will however not be satisfied, as most “other” citizens are unwilling to do more for society. Distrust in “other” individualistic citizens is emphasized in the moral evaluation, which counters the empathy for “victims” foregrounded in the dominant frame. According to participants’ treatment recommendation, the social dilemma resulting from the participatory society can be solved by governmentally offered incentives to participate. The third interpretation frame highlights a more positive problem definition: participants emphasized they are both willing and able to improve the quality of their living environment from bottom-up. This frame can be labelled as participation is a virtue, not a burden. The moral evaluation of this frame suggests that a shift in mentality is needed; citizens will need to have more mutual respect and need to look out for each other more to make the participatory society work. Similar to the previously described frames, inequality in willingness and ability to participate were emphasized. In the treatment recommendation, however, willingness and ability were interpreted as two different things. A lack of ability needs to be overcome by limiting the responsibilities for citizens (top-down), whereas limited willingness needs to be overcome by a shift in mentality of Dutch citizenry (bottom-up). The fourth interpretation frame was only found in the focus group discussions. It emphasizes how the participatory society is a disguised austerity measure initiated by the government. This frame can be labelled as participation is nothing but disguised savings. Morally reasoned, doing more for each other in society is a positive thing. However, this should be motivated by a bottom-up emerging desire for solidarity, and should not result from austerity measures. The fifth frame, also found exclusively in the focus group discussions, introduces a completely bottom-up interpretation of the participatory society. This frame can be labelled as participation enhances solidarity. In the causal interpretation, the participatory society is seen as the result of decreasing solidarity, which incited citizens to act from a less individualistic perspective. To the extent that this frame represents the only “pure” bottom-up interpretation, it can be thought of as a counter frame of all previously mentioned frames that foregrounded the top-down initiated economical savings as main driving force behind the participatory society. Conclusion and Discussion Study 1 This study revealed in what ways the participatory society was interpreted from either “bottom-up” or “top-down” audience frames. The top-down interpretations articulated least trust in the participatory society, as it was forced on citizens by the government. Bottom-up frames, in contrast, emphasized how citizens felt they were willing to do more for society. Still, participation should be something initiated by citizens themselves, and the government should only play a facilitating role. To the extent that new rounds of data collection in the interviews and focus group did not reveal new themes in the interpretation frames, theoretical saturation of interpretation frames was achieved (Glaser & Strauss, 1967). As the variety in viewpoints on the participatory society has been mapped out exhaustively, we can move on to the quantitative phase of the framing analysis. As a limitation of this first qualitative study, it should be noted that generalizations beyond the participants might potentially be problematic. However, by including qualitative data collected in various periods of the debate, and by triangulating the issue-specific findings with those of a more general topic in the pilot study, we aimed for maximum variation in interpretations regarding the four frame-elements resonating in public opinion. In the process of selective coding, these interpretations were translated into more generic themes. These were formulated as manifest variables indicating latent frame-elements in the content analysis of Study 2. Study 2: Quantitative Analysis of Media Frames Method Data collection. A quantitative content analysis was conducted to identify media frames. For the collection of data, a coding protocol and a coding sheet were developed (see Appendix B1 in the supplemental material for codebook). The variables in the codebook were based on a quantification of the qualitative study’s in-depth insights on the different frame-elements. Regarding the problem definition, for example, the qualitative study revealed that participants distinguished between the impact of the participatory society for themselves personally and for society in general. Moreover, in expressing their (dis)approval of this societal development, participants distinguished between participation initiated by citizens and participation requested by the government. To ensure these stances were reflected in the codebook, we included these distinctions in the variables measuring frame-elements. An example of such a variable included in the codebook is The participatory society as introduced by the government is a negative societal development versus The participatory society as introduced by the government is a positive societal development. On a 7-point semantic differential scale, we scored the position of each article regarding those oppositional statements. The codebook was pretested on a small sample (N = 15) of articles from regional newspapers. Based on the results of the pretest, the codebook was adjusted until the coding of new data no longer resulted in changes to the codebook. In that sense, saturation of the categories in the codebook was achieved. Sample. All articles covering the Dutch participatory society published in all Dutch national newspapers between September 17, 2013 and February 28, 2014 were included in the sample. The starting point is chosen, as the word “participatory society” was picked up in the media discourse right after the King’s speech on the September 17. From this point on, the discussion developed in both the media and public discourse. The sample was a census, considering every unit in the population was selected in the specified time frame (Bryman, 2012; Riffe, Lacy & Fico, 2005). The unit of analysis was the newspaper article. The final sample consisted of 152 Articles. Articles were selected using the following inclusion criteria in LexisNexis: the article had to be published in a Dutch national newspaper between September 17, 2013 and February 28, 2014, and the article had to contain at least one of the key words “participatiesamenleving”, “participatiemaatschappij,” or “doe-democratie” (three different Dutch words describing the concept participatory society). After a thorough reread of all sampled articles, 29 articles were excluded from the sample for the reason their content did not reflect a discussion on the participatory society. Data analysis. First, intercoder reliability was assessed for a sample of 20 randomly selected articles coded by two independent coders. Krippendorff’s alpha was computed as index for intercoder reliability. As this content analysis concerns an exploratory study, a value of 0.60 was considered as an acceptable threshold value for the variables included in the framing analysis (e.g., Lombard, Snyder-Duch & Bracken, 2002). As none of the variables in the codebook had a reliability lower than 0.60, all variables were retained for the analysis. The value of Krippendorff’s alpha for all variables is included in the supplemental material (Appendix B1). A principal component analysis was conducted on all 16 statements of the codebook. The factor analysis (rotation varimax) yielded a solution of four components with an eigenvalue greater than one that together explained 65.0% of the variance (the rotated component matrix is included in Table B1 in the supplemental material). As a robustness check, we also estimated the dimensional structure with an alternative measurement model for ordinal data that takes the hierarchy of item difficulty into account: Rasch scaling. The scalability indices of both measurement models provide similar results. Table 1 Framing Matrix Used to Identify Media Frames Frame-element Problem definition Causal interpretation Moral evaluation Treatment recommendation Frames Assessment of citizens’ initiatives in terms of their capacity and willingness to participate Assessment of causes and consequences of governmental influence on the participatory society Solidarity and equality of consequences participatory society for Dutch citizens Should the government or its citizens participate more in society? 1. (N = 60): “Citizens are better able to initiate participation from bottom-up than the government. But both need to participate more in society” (M = 4.04, SD = 0.49): Some confidence in citizens’ own initiatives (development is not positive/not negative) (M = 3.45, SD = 0.60): a moderately negative assessment of governmental responsibilities and causal influence (M = 3.40, SD = 0.76): to some extent, consequences of participatory society are unequally distributed and solidarity decreasing (M = 3.77, SD = 0.52); both government and citizens need to participate more 2. (N = 55): “Trust in bottom-up citizens’ initiatives” (M = 5.04, SD = 0.86): confidence in citizens’ own initiatives (development is relatively positive) (M = 3.70, SD = 0.77): less trust in causes and consequences government. Bit negative assessment of government’s causal responsibility (M = 4.44, SD = 0.62): participatory has no unequal consequences and is not solidary decreasing (M = 2.85, SD = 0.82): citizens should participate more than the government 3. (N = 26): “Skeptical of citizens’ initiatives and the government’s influence on the unfair participatory society” (M = 3.72, SD = 0.47): least confident (less than average) in citizens’ own initiatives (M = 2.90, SD = 0.51): least positive assessment of governmental influence on causes and consequences of participatory society (M = 3.27, SD = 0.78): consequences of participatory society are unequally distributed and solidarity decreasing (M = 5.56, SD = 0.54): citizens should contribute less to society than the government 4. (N = 11): “Confidence in both citizens who have to participate more in society and the government, who has to stimulate the participatory society from top-down” (M = 5.93, SD = 0.60): most confidence in citizens’ own initiatives. Problem is not that big as citizens are positive toward participating (M = 5.09, SD = 0.43): most positive evaluation of governmental influence. Still, government caused participatory society but consequences are not that bad (M = 5.30, SD = 0.57): participatory society has equal consequences and is solidary increasing (M = 2.00, SD = 0.00): citizens should participate more than the government Frame-element Problem definition Causal interpretation Moral evaluation Treatment recommendation Frames Assessment of citizens’ initiatives in terms of their capacity and willingness to participate Assessment of causes and consequences of governmental influence on the participatory society Solidarity and equality of consequences participatory society for Dutch citizens Should the government or its citizens participate more in society? 1. (N = 60): “Citizens are better able to initiate participation from bottom-up than the government. But both need to participate more in society” (M = 4.04, SD = 0.49): Some confidence in citizens’ own initiatives (development is not positive/not negative) (M = 3.45, SD = 0.60): a moderately negative assessment of governmental responsibilities and causal influence (M = 3.40, SD = 0.76): to some extent, consequences of participatory society are unequally distributed and solidarity decreasing (M = 3.77, SD = 0.52); both government and citizens need to participate more 2. (N = 55): “Trust in bottom-up citizens’ initiatives” (M = 5.04, SD = 0.86): confidence in citizens’ own initiatives (development is relatively positive) (M = 3.70, SD = 0.77): less trust in causes and consequences government. Bit negative assessment of government’s causal responsibility (M = 4.44, SD = 0.62): participatory has no unequal consequences and is not solidary decreasing (M = 2.85, SD = 0.82): citizens should participate more than the government 3. (N = 26): “Skeptical of citizens’ initiatives and the government’s influence on the unfair participatory society” (M = 3.72, SD = 0.47): least confident (less than average) in citizens’ own initiatives (M = 2.90, SD = 0.51): least positive assessment of governmental influence on causes and consequences of participatory society (M = 3.27, SD = 0.78): consequences of participatory society are unequally distributed and solidarity decreasing (M = 5.56, SD = 0.54): citizens should contribute less to society than the government 4. (N = 11): “Confidence in both citizens who have to participate more in society and the government, who has to stimulate the participatory society from top-down” (M = 5.93, SD = 0.60): most confidence in citizens’ own initiatives. Problem is not that big as citizens are positive toward participating (M = 5.09, SD = 0.43): most positive evaluation of governmental influence. Still, government caused participatory society but consequences are not that bad (M = 5.30, SD = 0.57): participatory society has equal consequences and is solidary increasing (M = 2.00, SD = 0.00): citizens should participate more than the government Table 1 Framing Matrix Used to Identify Media Frames Frame-element Problem definition Causal interpretation Moral evaluation Treatment recommendation Frames Assessment of citizens’ initiatives in terms of their capacity and willingness to participate Assessment of causes and consequences of governmental influence on the participatory society Solidarity and equality of consequences participatory society for Dutch citizens Should the government or its citizens participate more in society? 1. (N = 60): “Citizens are better able to initiate participation from bottom-up than the government. But both need to participate more in society” (M = 4.04, SD = 0.49): Some confidence in citizens’ own initiatives (development is not positive/not negative) (M = 3.45, SD = 0.60): a moderately negative assessment of governmental responsibilities and causal influence (M = 3.40, SD = 0.76): to some extent, consequences of participatory society are unequally distributed and solidarity decreasing (M = 3.77, SD = 0.52); both government and citizens need to participate more 2. (N = 55): “Trust in bottom-up citizens’ initiatives” (M = 5.04, SD = 0.86): confidence in citizens’ own initiatives (development is relatively positive) (M = 3.70, SD = 0.77): less trust in causes and consequences government. Bit negative assessment of government’s causal responsibility (M = 4.44, SD = 0.62): participatory has no unequal consequences and is not solidary decreasing (M = 2.85, SD = 0.82): citizens should participate more than the government 3. (N = 26): “Skeptical of citizens’ initiatives and the government’s influence on the unfair participatory society” (M = 3.72, SD = 0.47): least confident (less than average) in citizens’ own initiatives (M = 2.90, SD = 0.51): least positive assessment of governmental influence on causes and consequences of participatory society (M = 3.27, SD = 0.78): consequences of participatory society are unequally distributed and solidarity decreasing (M = 5.56, SD = 0.54): citizens should contribute less to society than the government 4. (N = 11): “Confidence in both citizens who have to participate more in society and the government, who has to stimulate the participatory society from top-down” (M = 5.93, SD = 0.60): most confidence in citizens’ own initiatives. Problem is not that big as citizens are positive toward participating (M = 5.09, SD = 0.43): most positive evaluation of governmental influence. Still, government caused participatory society but consequences are not that bad (M = 5.30, SD = 0.57): participatory society has equal consequences and is solidary increasing (M = 2.00, SD = 0.00): citizens should participate more than the government Frame-element Problem definition Causal interpretation Moral evaluation Treatment recommendation Frames Assessment of citizens’ initiatives in terms of their capacity and willingness to participate Assessment of causes and consequences of governmental influence on the participatory society Solidarity and equality of consequences participatory society for Dutch citizens Should the government or its citizens participate more in society? 1. (N = 60): “Citizens are better able to initiate participation from bottom-up than the government. But both need to participate more in society” (M = 4.04, SD = 0.49): Some confidence in citizens’ own initiatives (development is not positive/not negative) (M = 3.45, SD = 0.60): a moderately negative assessment of governmental responsibilities and causal influence (M = 3.40, SD = 0.76): to some extent, consequences of participatory society are unequally distributed and solidarity decreasing (M = 3.77, SD = 0.52); both government and citizens need to participate more 2. (N = 55): “Trust in bottom-up citizens’ initiatives” (M = 5.04, SD = 0.86): confidence in citizens’ own initiatives (development is relatively positive) (M = 3.70, SD = 0.77): less trust in causes and consequences government. Bit negative assessment of government’s causal responsibility (M = 4.44, SD = 0.62): participatory has no unequal consequences and is not solidary decreasing (M = 2.85, SD = 0.82): citizens should participate more than the government 3. (N = 26): “Skeptical of citizens’ initiatives and the government’s influence on the unfair participatory society” (M = 3.72, SD = 0.47): least confident (less than average) in citizens’ own initiatives (M = 2.90, SD = 0.51): least positive assessment of governmental influence on causes and consequences of participatory society (M = 3.27, SD = 0.78): consequences of participatory society are unequally distributed and solidarity decreasing (M = 5.56, SD = 0.54): citizens should contribute less to society than the government 4. (N = 11): “Confidence in both citizens who have to participate more in society and the government, who has to stimulate the participatory society from top-down” (M = 5.93, SD = 0.60): most confidence in citizens’ own initiatives. Problem is not that big as citizens are positive toward participating (M = 5.09, SD = 0.43): most positive evaluation of governmental influence. Still, government caused participatory society but consequences are not that bad (M = 5.30, SD = 0.57): participatory society has equal consequences and is solidary increasing (M = 2.00, SD = 0.00): citizens should participate more than the government The four-factor solution resulted in scales that were interpretable in terms of meaningful frame-elements. The frame-elements are labelled as following: (1) assessment of citizens’ initiatives in terms of their capacity and willingness to participate (Cronbach’s alpha = 0.84); (2) assessment of causes and consequences of governmental influence on the participatory society (Cronbach’s alpha = 0.76); (3) moral evaluation in terms of solidarity and equality of consequences the participatory society has for Dutch citizens (Cronbach’s alpha = 0.69); and (4) treatment recommendation: should the government or citizens participate more in society? (Cronbach’s alpha = 0.81). After reducing the 16 items in the codebook to the four above-described frame-elements, the next step was to analyze how articles clustered on these frame-elements. To do so, a hierarchical cluster analysis using Ward’s method with Euclidian distances was conducted on the four computed scales (Burns & Burns, 2009). Based on the qualitative study, we expected to find a solution in the region of five clusters, interpretable as meaningful frames that differ substantially on the distinguished frame-elements. We first allowed the clustering procedure to estimate a solution between 2 and 12 clusters. The dendrogram and agglomeration schedule supported a four-cluster solution as the most optimal one: merging clusters to a lower number resulted in solutions with too distant clusters, while distinguishing between a higher number of clusters resulted in clusters that were too similar. For this reason, we rerun the cluster analysis with a fixed number of four clusters. This solution was saved as a new variable, of which the four categories were reinterpreted as media frames. Results Study 2: Media Frames on the Participatory Society To interpret the four clusters as meaningful media frames, a framing matrix consisting of the four scales was constructed for each frame separately (see Table 1 for framing matrix). Based on analyses of variance (one-way ANOVA’s), significant differences between frames on their scores on the four frame-elements were interpreted. All frames differed significantly on their assessment of the four frame-elements included in the matrix (see ANOVA Table B2 of the supplemental material). The first media frame was found in 39.5% of all articles (N = 60). This frame is best described as: “Citizens are better able to initiate participation from bottom-up than the government, who is forcing participation from top-down.” This frame emphasizes some trust in citizens to participate successfully in the participatory society from bottom-up. The causes and consequences of governmental influence are interpreted with less trust. Morally reasoned, this frame emphasizes that the consequences of the participatory society are unequally distributed in society because this development excludes citizens and decreases solidarity. As treatment recommendation, this frame suggests that both citizens and the government should participate more in society. The second media frame was found in 36.2% (N = 55) of all articles and is referred to as “Trust in bottom-up citizens’ initiatives.” This frame emphasizes significantly more trust in citizens’ initiatives than the first frame, but is just as skeptical of governmentally initiated influences. In the moral evaluation, inequality does not play a role. In the treatment recommendation, this frame clearly emphasizes that citizens need to contribute more to society than the government. The third media frame is less dominant than the first two frames. In total, 17.2% of the articles (N = 26) frames the participatory society from this “Sceptical of citizens’ initiatives and the government’s influence on the unfair participatory society” media-frame. This frame emphasizes least trust in both the government’s influence on the participatory society and citizen’s initiatives to participate more in society. Inequality of consequences and social exclusion are emphasized in the moral evaluation. As treatment recommendation, the government rather than citizens should participate more in society. The final frame found in the media discourse concerns a pure bottom-up frame that emphasizes trust in citizens’ ability and willingness to participate more in society. Furthermore, the government’s influence on causes and consequences of the participatory society is assessed with great confidence. This frame recommends that citizens should participate substantially more than the government. This frame was found in only 7.2% of all articles (N = 11) and challenges the position emphasized in the third frame. This frame can be described as: “Confidence in both citizens, who have to participate more in society, and the government, who has to stimulate the participatory society from above.” Conclusion and Discussion Study 2 The four frames found in this study can be reduced even further to two more generic frames resonating in the media discourse. First, a dominant bottom-up citizens’ initiative frame, in which citizens’ willingness and ability to participate is made salient and the freedom to participate and take matters into own hands are highlighted as virtue of modern citizenship, regardless of the economic recession. Second, the top-down “governmentally forced participatory society frame” functions as a counter frame. This frame defines the participatory society as an inevitable result from a government that is no longer able to carry the consequences of the welfare state. In general, the media discourse articulated a lack of trust in the government to stimulate participation. Study 3: Quantitative Analysis of Interpretation Frames Method Data collection. The framing analysis of this study is based on responses to a set of questions included in an opinion poll, presented as an online survey executed by a research organization. This question block consisted of 18 statements on the participatory society. The statements were formulated on a 7-point semantic differential scale. The statements were based on the qualitative results of Study 1 and aimed to capture the variety of perspectives found on all four frame-elements. In addition, more recent developments of the debate in public opinion after the earlier rounds of data collection were taken into account. To provide an example, the distinction between helping family versus helping strangers became a fierce topic of debate as identified in various offline and online media, which we incorporated in the measures of the different frame-elements. In the framing-matrix, all statements measuring the four frame-elements are included (Appendix C2 of the supplemental material). Sample. The sample of this study consisted of 999 respondents selected from the panel of the research organization. The sample is representative of the Dutch voting population in terms of gender, age, educational level, region, family size, social class, and voting behavior. All participants in the study were aged >18 years. The response rate of this panel is 70%. Respondents were invited to participate in the research via e-mail, letter, and/or telephone. The mean age of participants was 50 years (SD = 17.40), and 47.5% of the participants were female. Data analysis. We conducted a confirmatory factor analysis to assess whether the statements fitted a four-dimensional structure consisting of the four frame-elements. As indicated by the suboptimal model fit, we did not find support for this solution [χ2(118) = 754.50, p < .001, Confirmatory Fit Index (CFI) = 0.866, root mean square error of approximation = 0.074, 90% confidence interval (0.069; 0.079)]. We therefore did not reduce the variables to scales. Next, we identified a meaningful number of interpretation frames by conducting a hierarchical cluster analysis using Ward’s method with Euclidian distance (Burns & Burns, 2009; Matthes & Kohring, 2008). The framing analyses in Study 1 and Study 2 indicated that a number of four or five frames on a certain topic are an appropriate solution, which allows for the presence of both dominant and counter frames. Similar to Study 2, we allowed the clustering procedure to estimate a solution ranging between 2 and 12 clusters. The agglomeration schedule and dendrogram confirm the presence of five clusters with a meaningful overall distance between statements that could also be interpreted substantially. We compared the five-cluster solution to both higher and lower numbers of clusters. Merging clusters to a lower number resulted in solutions with too distant clusters, while distinguishing between a higher number of clusters resulted in clusters that were too similar. Based on this outcome, the cluster analysis was rerun with a fixed solution of five clusters, which resulted in a new variable with five categories. These categories grouped the participants into five different interpretation frames. By constructing a similar framing matrix as Study 1 and Study 2, these five clusters were interpreted as interpretation frames. More specifically, we compared the mean scores on all statements between clusters to systematically reconstruct the frames per frame-element. In one interpretation frame, for example, participants expressed a negative problem definition, and blamed the government in their causal interpretation (see Appendix C2 in the supplemental material for complete framing matrix). Results Study 3: Interpretation Frames in the Public Discourse As the identification of significant mean differences on statements is a necessary prerequisite to validate discriminant clusters, analyses of variance (one-way ANOVA) were conducted on the five identified clusters or “frames” (Burns & Burns, 2009). As depicted in ANOVA Table C1 of the supplemental material, all between-group means differed significantly (p < .05), which implies that the 18 statements reliably distinguished between five significantly different interpretation frames. Similarities between frames were interpreted as meaningful as well, as not all frames necessarily differ on all separate parts of all four frame-elements (e.g., Entman, 1993). Similar to Study 1 and Study 2, a framing matrix consisting of all four frame-elements was constructed (see supplemental material, Appendix C2). The mean values of the statements and the significant differences between these values across the frames guided the interpretation of the clusters into frames. The first frame is labelled as “Necessity is the mother of invention when it comes down to our readiness for participation.” This is the most dominant frame of reference carried by 31.0% of the participants (N = 309) and represents a balanced evaluation of the participatory society that is neither positive nor negative. Citizens who interpret the participatory society from this interpretation frame do not reject the participatory society as societal development. However, they do not experience a great willingness or ability to participate more in society if the government is withdrawing from society. Furthermore, clear lines are drawn in terms of readiness for participation in society. Citizens feel responsible to help a neighbor in need, for example, by giving him or her a ride to the hospital. Helping a neighbor in need of care in less urgent, everyday situations, such as getting up in the morning, is too demanding for these citizens. The second frame, referred to as “Despite feeling responsible for others, we are unwilling and unable to participate more in society” is held by 28.0% of the participants (N = 280). Reasoned from this interpretation frame, the participatory society is a negative societal development, especially for weaker groups. Moreover, citizens lack the capacity and willingness to participate more. They do experience feelings of responsibility for others, but are only willing and able to act on these feelings in emergency situations. Health care and neighborhood security are tasks for the government, which should not be transferred to citizens’ responsibilities. If the government retreats, especially weaker groups in society will be harmed. This interpretation frame counters the generic “Bottom-up citizens’ initiative frame” found in Study 2 as media frame and in Study 1 as interpretation frame. The third frame is labelled as “In times of decreasing solidarity, citizens are responsible to take care of each other in times of need, which is especially important for weaker groups in society.” This interpretation frame, held by 28.0% of the respondents (N = 280) emphasizes an even stronger sense of responsibility (M = 5.80, SD = 1.15) than articulated in Frames 1 and 2. This frame is similar to the bottom-up frames found in Study 1 that emphasized citizens’ willingness to participate in society. This frame differs from the bottom-up frames in the qualitative study in the sense that citizens experience a willingness to participate, but at the same time emphasize they are not always able to do so in the causal interpretation and moral evaluation. As treatment recommendation, citizens emphasize that the government needs to stimulate participation. The preferred role of the government as safety net selected in the moral evaluation and treatment evaluation is congruent with the limits to participation highlighted in the bottom-up frames identified in the qualitative study. The fourth frame is labelled as “The participatory society is a negative societal development for all citizens who are willing to help others in need but lack the capacity and desire to participate more in society.” This frame is rare: only 8.0% (N = 76) of all participants interpreted the participatory society from this frame of reference. Reasoned from the moral evaluation of this frame, all groups in society will be affected equally by the participatory society (M = 5.83, SD = 1.15), which is in sharp contrast with all other frames that emphasize that some groups in society will be affected more than others. Furthermore, this interpretation frame highlights the willingness of citizens to incidentally take a neighbor to the hospital (M = 5.92, SD = 1.28). However, similar to Frames 1 and 2, this moral evaluation does not result in more willingness to help a neighbor with daily care of getting dressed: (M = 2.74, SD = 1.53). Hence, participation is reserved for emergency situations. The fifth frame is carried by 5.0% (N = 54) of the respondents and is referred to as “We feel responsible and are able to participate more in society, but weaker groups lack this capacity completely.” Despite the problem definition and causal interpretation emphasize the participatory society as a solidarity decreasing development, citizens are both able and willing to participate more in society. This frame counters the dominant frame “Despite feeling responsible for others, we are unwilling and unable to participate more in society.” This frame represents the only pure bottom-up interpretation, as it highlights both the ability and willingness to participate in different domains of society, not only in emergency situations. Despite the emphasis on willingness and ability to participate foregrounded in the causal and moral interpretations, citizens holding this frame are still negative about the participatory society as introduced by the government in their problem definition (M = 1.57, SD = 1.35), which underlines the moral evaluation that especially weaker groups are harmed by the participatory society. Conclusion and Discussion Study 3 The largest overall differences in the interpretation of statements across all four frame-elements were found between the second frame “Despite feeling responsible for others, we are unwilling and unable to participate more in society” and the fifth frame “We feel responsible and are able to participate more in society, but weaker groups lack this capacity completely.” The fifth frame can therefore be interpreted as the counter frame of the second frame. Interpreting the participatory society from the dominant second frame, a stronger inability and unwillingness to participate in society is experienced than in the fifth frame, which articulates a strong belief of citizens being able and willing to participate in society when needed. For example, respondents interpreting reality from the second frame believe they are unable to look out for the safety in their neighborhood if the police will contribute less to this task (M = 2.18, SD = 1.36). This experienced ability is significantly and substantially higher for the fifth frame (M = 5.83, SD = 1.55); Mdifference = 3.66, p < .001. At this stage, we can draw conclusions on the discrepancy between media frames and interpretation frames (see Figure 1). The largest difference concerns the amount of trust attached to citizens’ initiatives. Both the media and public discourse emphasized a lack of trust in the government who forced the scheme of participation on citizens. Still, by covering successful bottom-up citizens’ initiatives, the media inherently articulated more trust in citizens’ initiatives than citizens highlighted themselves. The media frames and interpretation frames also demonstrated overlap. The dominant media frame, which emphasized citizens’ willingness and ability to participate, resonates with bottom-up interpretation frames held by some citizens. However, this frame was far less frequently articulated by citizens than the media. We interpret this as support for framing within the negotiated media effects paradigm (e.g., McQuail, 2010, Scheufele, 1999; Scheufele & Tewksbury, 2007). The media may propose certain frames to receivers, and receivers still possess the agency to resist and adjust the meaning of these frames in line with their own frame of reference. Figure 1 View largeDownload slide Graphical depiction of the overlap and discrepancies between interpretation frames (Study 1 and Study 3) and media frames (Study 2). Arrows between boxes indicate overlap between frames, and arrows on the left and right side indicate overlap within the receiver- or sender-side. When boxes are not connected, discrepancies are highlighted. The strength of the overlap decreases from top to bottom. Figure 1 View largeDownload slide Graphical depiction of the overlap and discrepancies between interpretation frames (Study 1 and Study 3) and media frames (Study 2). Arrows between boxes indicate overlap between frames, and arrows on the left and right side indicate overlap within the receiver- or sender-side. When boxes are not connected, discrepancies are highlighted. The strength of the overlap decreases from top to bottom. Overall Discussion The aim of this research was to assess discrepancies between media frames and interpretation frames. The introduced method responded to recent calls in the literature, which suggest (1) to acknowledge the social–cultural context and power relations underlying the framing process (Carragee & Roefs, 2004) (2) to consider frames as a cluster of frame-elements and reasoning devices (e.g., Matthes & Kohring, 2008; Van Gorp, 2007), and (3) to identify interpretation frames and media frames using a similar analytical approach to ensure a valid comparison of both interrelated framing processes (Matthes & Kohring, 2008). In three studies, this research aimed to validate the introduced methodological framework departing from a social-constructivist viewpoint. The overarching research question that focused on the discrepancies between media frames and interpretation frames was: To what extent and how does the framing of the participatory society differ between interpretation frames of citizens and media frames of Dutch news media? Our research shows that media discourse and public discourse both emphasized a clear distinction between bottom-up frames and top-down frames on the participatory society but differed in their evaluation of these top-down and bottom-up frames and their consequences for Dutch society. The interpretation frames identified in Study 1 and Study 3 emphasize a lack of trust in citizens’ bottom-up initiatives, whereas the media frames identified in Study 2 predominantly emphasize the success stories of bottom-up initiated examples of the participatory society. Concerning the top-down frames, both media and citizens emphasized a lack of trust in the feasibility of the forcing schemes promoting the participatory society from a governmental perspective (Scott, 1998). Furthermore, in both the media and public discourse, the capacity of Dutch society to carry the consequences of the participatory society was questioned. Interpretation frames and media frames both emphasized that especially weaker groups in society will be affected by the consequences of the participatory society. Owing to the exploratory nature of the studies incorporated in this research, some limitations regarding the design of the studies and the operationalization of the research materials need to be highlighted. First, the second study has exclusively focused on media frames present in Dutch newspapers. Dutch citizens draw on an extensive stock of sociocultural resources, and newspapers only represent one of these resources among many competing ones. Future research should include more of these media-based sociocultural resources to arrive at a more complete comparison between media frames and interpretation frames. The second limitation concerns the operationalization of frame-elements incorporated in the content analysis and subsequent cluster analyses. Some variables link up to more than one single frame-element, which may affect the identification of frames as clusters of frame-elements. However, the development of variables was informed by the in-depth insights of the qualitative study. Moreover, we ensured that the variables were attached to the most suitable frame-element. In any case, the study has both clear methodological and theoretical implications. Methodologically, the study demonstrates the importance to consider media framing and public framing as related, yet conceptually distinct phenomena. Instead of assuming that frames present in media coverage will also be the ones that are present among members of the public, it is worth to analyze both realms separately and explicitly question similarities and differences in those frames. Moreover, it demonstrates how qualitative and quantitative methods can be complementary in one study. Theoretically, the findings offer support against the idea of strong framing effects. Although the media may put forward a certain set of frames to understand political issues, members of the audience will only partially adopt this set of frames: based on, for example, different types of information and personal experiences, they will resist, adjust, and supplement the stock of frames put forward by the media. This implies that a situation of perfect alignment between media framing and audience framing is highly unlikely. This finding fits well with the negotiated media effects paradigm (Gamson & Modigliani, 1989). Overall, a structural and detailed approach as presented in this study can unravel differences and similarities in the presence of frames and offers a starting point for understanding media effects for a plethora of issues in a wide variety of contexts. Michael Hameleers is PhD candidate in Political Communication at the University of Amsterdam (ASCoR). The main topic of his dissertation research is populism in the media. His research interests include framing, mediated populism, and the role of social identity in media effects. Rens Vliegenthart is professor for Media and Society in the Department of Communication Science at ASCoR at the University of Amsterdam. His research interests include media–politics interactions, media-effects, and comparative politics. Acknowledgments The findings presented in this article resulted from an intensive cooperation between the University of Amsterdam (ASCoR), Veldkamp, and TNS NIPO. The different parties had no financial interests in the outcomes of this study. The mutual interests of the involved parties were exclusively based on the development of a methodological tool for framing analysis applicable in both scientific and market research. Supplementary Data Supplementary Data are available at IJPOR online. References Burns R. , Burns R. ( 2009 ). Chapter 23: cluster analysis. In Business research methods and statistics using SPSS (pp. 552 – 567 ). California : Sage Publications . Bryman A. ( 2012 ). Social research methods . Oxford : Oxford University Press . Carragee K. M. , Roefs W. ( 2004 ). The neglect of power in recent framing research . Journal of Communication , 54 , 214 – 233 . doi: 10.1111/j.1460-2466.2004.tb02625.x Google Scholar CrossRef Search ADS Entman R. M. ( 1993 ). Framing: Toward clarification of a fractured paradigm . Journal of Communication , 43 , 51 – 59 . doi: 10.1111/j.1460-2466.1993.tb01304.x Google Scholar CrossRef Search ADS Flick U. ( 2009 ). An introduction to qualitative research ( 4th ed .). London : Sage Publications . Gamson W. A. , Modigliani A. ( 1989 ). Media discourse and public opinion on nuclear power: A constructionist approach. American Journal of Sociology , 95 ( 1 ), 1 – 37 . doi: 10.1086/229213 Google Scholar CrossRef Search ADS Goffman E. ( 1974 ). Frame analysis: An essay on the organization of experience . Cambridge : Harvard University Press . Glaser B. G. , Strauss A. L. ( 1967 ). The discovery of grounded theory: Strategies for qualitative research . Chicago : Aldine . Hall S. ( 1973 ). Encoding/decoding. In Hall S. , Hobson D. , Lowe A. , Tillis P. (Eds.), Culture, media, language (pp. 128 – 138 ). London : Hutchinson . Holstein J. A. , Gubrium J. F. ( 1995 ). The active interview. Thousand Oaks : Sage . Google Scholar CrossRef Search ADS Huang H. ( 2010 ). Frame-rich, frame-poor: An investigation of the contingent effects of media frame diversity and individual differences on audience frame diversity . International Journal of Public Opinion Research , 22 , 47 – 73 . doi: 10.1093/ijpor/edp024 Google Scholar CrossRef Search ADS Lombard M. , Snyder-Duch J. , Bracken C. C. ( 2002 ). Content analysis in mass communication: Assessment and reporting of intercoder reliability . Human Communication Research , 28 , 587 – 604 . doi: 10.1111/j.1468-2958.2002.tb00826.x Google Scholar CrossRef Search ADS Matthes J. , Kohring M. ( 2008 ). The content analysis of media frames: Toward improving reliability and validity . Journal of Communication , 58 , 258 – 279 . doi: 10.1111/j.1460-2466.2008.00384.x Google Scholar CrossRef Search ADS McLafferty I. ( 2004 ). Focus group interviews as a data collecting strategy . Journal of Advanced Nursing , 48 , 187 – 194 . doi: 10.1111/j.1365-2648.2004.03186.x Google Scholar CrossRef Search ADS PubMed McQuail D. ( 2010 ). McQuail’s mass communication theory ( 6th ed .). London : Sage . Neuman W. R. , Just M. R. , Crigler A. A. ( 1992 ). Common knowledge. News and the construction of political meaning . Chicago : The University of Chicago Press . Reese S. D. ( 2001 ). Introduction. In Reese S. D. , Gandy O. H. , Grant A. E. (Eds.), Framing public life: Perspectives on media and our understanding of the social world (pp. 1 – 31 ). Mahwah, NJ : Erlbaum . Riffe D. , Lacy S. , Fico F. ( 2005 ) Analyzing media messages; using quantitative content analysis in research ( second edition). New Jersey : Mahwah Scheufele D. A. ( 1999 ). Framing as a theory of media effects . Journal of Communication , 49 , 103 – 122 . doi: 10.1111/j.1460-2466.1999.tb02784.x Google Scholar CrossRef Search ADS Scheufele D. A. , Tewksbury D. ( 2007 ). Framing, agenda setting, and priming: The evolution of three media effects models . Journal of Communication , 57 , 9 – 20 . doi: 10.1111/j.0021-9916.2007.00326.x Scott J. C. ( 1998 ). Seeing like a state. How certain schemes to improve the human condition have failed . New Haven and London : Yale University Press . Semetko H. A. , Valkenburg P. M. ( 2000 ). Framing European politics: A content analysis of press and television news . Journal of Communication , 50 , 93 – 109 . doi: 10.1111/j.1460-2466.2000.tb02843.x Google Scholar CrossRef Search ADS Tankard J. W. ( 2001 ). The empirical approach to the study of media framing. In Reese S. D. , Gandy O. H. , Grant A. E. (Eds.), Framing public life: Perspectives on media and our understanding of the social world (pp. 95 – 106 ). Mahwah, NJ : Erlbaum . Van Gorp B. ( 2007 ). The constructionist approach to framing: bringing culture back in . Journal of Communication , 57 , 60 – 78 . doi: 10.1111/j.0021-9916.2007.00329.x Vliegenthart R. , Van Zoonen L. ( 2011 ). Power to the frame: Bringing sociology back to frame analysis . European Journal of Communication , 26 , 101 – 115 . doi: 10.1177/0267323111404838 Google Scholar CrossRef Search ADS Wicks R. H. ( 2001 ). Understanding audiences: Learning to use the media constructively . Mahwah, NJ : Erlbaum © The Author 2016. Published by Oxford University Press on behalf of The World Association for Public Opinion Research. All rights reserved. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

International Journal of Public Opinion ResearchOxford University Press

Published: Dec 23, 2016

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off