Building Theory From Media Ideology: Coding for Power in Journalistic Discourse

Abstract

Critical discourse analysis (CDA) is an increasingly important approach for critical-qualitative communication scholarship. This essay has three purposes: (1) to explain the history and applications of CDA, (2) to provide an empirical snapshot of how CDA has been used in journalism studies research, and (3) to provide a methodological intervention for improving CDA research in our field. An exploratory analysis of 17 Association for Education in Journalism and Mass Communication journal articles reveals that CDA research on journalism topics has focused on race and ethnicity, analyzed mainstream print news outlets, and applied CDA as a method in its own right. Authors have neither adequately defined discourse for readers nor have they sufficiently explained their coding procedures. Journalism scholars must improve the transparency of our coding methods, and we must examine ideological formations beyond dominant-hegemonic discourses. By analyzing mainstream journalism, alternative media, and online media side-by-side, CDA researchers can build stronger theory about ideology’s role in journalistic contexts.

Keywords

critical discourse analysis cultural studies ideology journalism qualitative research methods

Ideological analysis is a pressing concern for journalism scholars. As President Trump leads a crusade against “fake news,” and as accusations of sexual assault rock communications professions, researchers must tackle questions regarding the uses and abuses of power in media. Rarely has the need for critical-qualitative mass communication research been so apparent. Nevertheless, critical analysis is sometimes perceived today as it was an era ago—as “substantially opposed on questions of the methodology and objectives of social science, about what constitutes knowledge in the social sciences” (Ray, 1979, p. 149). Ideological research has been critiqued for lacking rigorous, empirical methods. Many humanistic mass communication scholars continue to view their work as belonging to a separate research paradigm (Jensen, 2013), drawing attention to the perceived disjoint in reliability and validity between qualitative and quantitative research (Golafshani, 2003). This is a problem within our field.

Some journalism scholars have intervened by adopting a robust qualitative research framework known as critical discourse analysis (CDA). CDA is an interdisciplinary approach of tacit theories and methods that grew out of traditions in rhetoric and cognitive science, sociolinguistics, and anthropology (Wodak & Meyer, 2009). Researchers who use CDA conceptualize language as “an integral element of the material social process” (Fairclough, 2001, p. 122), meaning they view communicative objects as outcomes and constituents of social practice (Fairclough & Wodak, 1997). CDA is a helpful approach for scholars interested in examining mass media’s roles in the social construction of reality. Its attention to power analysis, historicity, and cultural context also make it unique among qualitative analysis tools. However, CDA’s analytical strategies are as diverse as the fields it emerged from, and its foremost scholars do not always agree on its goals or outcomes. Critics of CDA have argued researchers’ methodological commitments have been poorly documented (Carvalho, 2008), and there are disagreements regarding whether CDA is a theory, method, or worldview. These concerns merit a revived discussion about CDA in communication disciplines.

In this essay, I review the primary tenets of CDA for an audience of journalism scholars, and I offer methodological interventions for CDA in our field. To begin, I highlight active scholarly debates about CDA as a conceptual roadmap for reading this essay. Then, I present an exploratory review of 17 CDA studies about journalistic discourse that were published in Association for Education in Journalism and Mass Communication (AEJMC) division and interest group journals. Through thematic analysis, I illustrate the strengths and weaknesses of CDA as it has been applied within journals in our discipline. Finally, in response to concerns that critical-qualitative findings lack generalizability, rely on negligible data, are based on subjective interpretation, and analyze limited discursive contexts (Sandelowski, 1986), I introduce a rigorous, comparative method for CDA coding. This method builds theory through ideological critique and helps journalism scholars empirically visualize oppression across ideologically diverse media.

Discourse and Critical Discourse Analysis

Like other “emergent” and “iterative” approaches to qualitative journalism research, CDA lacks conceptual clarity. There is no single agreed-upon way of “doing” CDA. And discourse itself itself has seemingly infinite definitions. Discourse may refer to an instance of text or talk in its social context (van Dijk, 1989) or to a sociological phenomenon—the ideological process of constructing knowledge about a topic (Foucault, 1978). The term may also be “used abstractly to mean statements in general or to refer to a particular group or type of statements” (Philo, 2007, p. 176). Some scholars cite specific formal theories of discourse (see Wetherell, Taylor, & Yates, 2001 for examples), while others use broad definitions, arguing that “discourse means anything from a historical monument, a lieu de mémoire, a policy, a political strategy, narratives in a restricted or broad sense of the term, text, talk, a speech, topic-related conversations, to language per se” (Wodak & Meyer, 2009, pp. 2–3). In journalism studies, we analyze discourse on many levels, including headlines, graphic elements, and story framing (Philo, 2007; van Dijk, 1998). We must also be attentive to processes of production, asking ourselves “who has preferential access to journalists, who will be interviewed, quoted, and described in news reports, and whose opinions will thus be able to influence the public?” (van Dijk, 2013b, p. 68). In order to crystallize what our field means by “discourse,” journalism scholars must choose conceptual definitions that articulate our interests in media ideology as an outcome of newsroom dynamics and social processes. I appreciate Anabela Carvalho’s (2008) definition from her essay Media(ted) Discourse and Society. She writes that journalistic discourse is “media representation of social issues,” but also “the discursive construction of events, problems and positions by social actors” and “the discursive strategies that they employ” (p. 161). Journalism discourse is time-sensitive, mass disseminated social surveillance. It is imbued with power relations based on reportorial practices, interview sources, and production methods (Tuchman, 1978; van Dijk, 2013b).

Journalism CDA researchers should be politically aligned with the anticapitalist, antiimperialist logics that have historically guided critical theory and similar research perspectives in cultural studies. CDA traces its roots to Marxist thinkers and structuralists such as Ferdinand de Saussure. The political critiques made by Louis Althusser (1971) and Antonio Gramsci (1971) provide what Carvalho (2008) calls “an important backdrop” for the development of research on discourse, popularized by the Frankfurt School and the Birmingham Center for Cultural Studies. Today, Norman Fairclough, Teun van Dijk, and Ruth Wodak are the most recognizable CDA scholars in the sprawling field of discourse analysis. Each conceives discourse as being “linked to power and social interests” (Philo, 2007, p. 176). But their perspectives conflict in notable ways. Fairclough and Wodak’s approach to CDA examines the “semantic and grammatical features of texts,” while van Dijk has championed a binary thematic analysis tool called the ideological square (Philo, 2007, p. 178). Importantly, Fairclough considers CDA to be a method, whereas van Dijk considers CDA to be a worldview or a theoretical perspective for undertaking research. According to Wodak and her research partner Michael Meyer (2009), “proponents of CDA use discourse analysis to challenge what they regard as undesirable social and political practices” (p. 2), meaning CDA is necessarily linked with scholarly activism.

For Fairclough and Wodak (1997), discourses are grounded within contextual interactions between identity, production, cultural values, social relations, consciousness, and semiology, and they are always historically contingent. Discourse then is analogous to the circuit of culture (Du Gay et al., 1997), a metaphor popularized in Stuart Hall’s (1997) Representation. Fairclough and Wodak’s approach is unique among CDA research because it focuses primarily on communication in action—on discourse as a social process. In the third edition of his seminal text, Language and Power, Fairclough (2014) clarifies CDA’s goals, including that CDA is a “normative critique” (p. 49) of discourse, that CDA’s allegiances are transdisciplinary, that capitalism is the primary subject of critique, that discourse is socially motivated and interactive, that CDA’s outcome is social activism, and that analyses should be interdiscursive, thematic, and grammatical. Importantly, Fairclough provides what he calls “a guide and not a blueprint” (p. 129) for textual analysis in CDA, which focuses on a discourse’s vocabulary, grammar, and textual structures. However, critics have argued social action cannot be interpreted from text and talk without also observing media production and reception (Philo, 2007).

While Fairclough and Wodak represent the linguistic tradition in CDA research, van Dijk’s work has focused on elements of racism reproduced in news content, and unlike Fairclough and Wodak, van Dijk considers CDA to be a theoretical perspective rather than a research approach or a method proper. van Dijk (2013a) points out that CDA may draw on diverse methods that are “grammatical (phonological, morphological, syntactic), semantic, pragmatic, interactional, rhetorical, stylistic, narrative or genre analyses, among others, on the one hand, and through experiments, ethnography, interviewing, life stories, focus groups, participant observation, and so on, on the other hand”. Scholars who are new to CDA tend to mistake CDA as a method in its own right (van Dijk, 2013a). Instead, CDA should be understood as a descriptive and argumentative process adherent to an underlying critical-theoretical lens—one which takes into account the historical and institutional processes that allow for discourse about a subject to develop. The CDA scholar should “do critical discourse analysis by formulating critical goals, and then explain by what specific explicit methods you want to realize it” (van Dijk, 2013a) . In journalism studies, we need to chart our coding methods for analyzing power in media discourse rather than merely citing CDA in our methods sections.

van Dijk offers one method for analysis that he labels the ideological square. The ideological square identifies discursive formations of Othering, for example, us-versus-them discourses. According to van Dijk (1998), ideological formations will: “1. Emphasise our good properties/actions, 2. Emphasise their bad properties/actions, 3. Mitigate our bad properties/actions, 4. Mitigate their good properties/actions” (p. 33). The ideological square is a helpful tool for visualizing Orientalism and other forms of racism, but its closed coding approach may prevent more nuanced ideological critiques. For example, the ideological square ignores “the time plane in discourse analysis of journalistic texts” (Carvalho, 2008, p. 163), and it may overlook how Othering develops in media longitudinally. The closed coding approach also limits the identification of ideology in media. While the ideological square would appropriately describe and explain Yellow Peril, for instance, would it identify a model minority stereotype? We should keep in mind news discourses are “always-already” ideological (Althusser, 1971), even if they do not explicitly condemn a specific social group.

In the following section, I tighten this essay’s focus to paint a picture of how CDA has been used in contemporary journalism scholarship. Leading academics disagree about whether CDA is a method, theory, or worldview, and they have different ways they approach data and analyze discourse. This is well documented in the CDA literature, but to date little research has systematically examined how CDA has been applied within the mass communication discipline, nor within journalism studies specifically. After outlining how CDA research has been articulated in AEJMC division and interest group journals, I will offer two models for a CDA coding method that journalism studies researchers can apply to their own data sets (see Figures 1 and 2).

Figure 1.

Critical discourse analysis coding for media discourse.

Figure 2.

Transideological coding for critical discourse analysis.

CDA in AEJMC Journals

CDA is emerging as an important area for journalism studies research. In this section, I present the results of an exploratory review of 17 journalism-focused CDA studies published in AEJMC division and interest group journals during the last two decades. I build evidence showing that journalism scholars have cited CDA when they use ideological coding, but that our methods do not always line up with CDA as defined by its leading practitioners. I outline trends in the topics of analysis, objects of analysis, and the ideological positioning of the studies’ research objects. I also describe the researchers’ theoretical and methodological perspectives. This review shows that CDA research about journalistic discourse accepted to AEJMC journals tends to focus on race and ethnicity, analyze mainstream print news outlets, and cite CDA as a method in its own right. Authors using CDA in journalism research do not often define discourse for readers, nor do they adequately explain their coding procedures. I offer this exploratory review in order to capture CDA as a sometimes-misunderstood, but extremely valuable approach for mass communication scholarship.

Sampling

In order to conduct the exploratory review, I evaluated each of the AEJMC division and interest group journals for articles matching keyword searches for “critical discourse analysis.” A full list of citations to the articles can be found in the Appendix section. I selected a census of studies whose objects of analysis included journalistic discourse. In total, 17 studies published between 1996 and 2016 matched the selection criteria. My search parameters returned no examples of journalism-specific CDA research published in AEJMC journals before 1996—six years after Teun van Dijk (1990) launched his flagship journal, Discourse & Society.

CDA appears to be an increasingly popular lens for journalism research. Four (24%) CDA studies in the sample were published between 1996 and 2006, and the remaining 13 (76%) ran between 2007 and 2016. However, there is a dearth of CDA research in journalism compared with more popular text-based research methods such as quantitative content analysis.¹ Despite CDA’s increasing traction, it remains a marginalized research approach in our field, as does qualitative work more generally. Of the 17 articles, 7 (41%) were published in the Journal of Communication Inquiry and 4 (24%) were published in Journalism and Mass Communication Quarterly, pointing to the critical-theoretical orientation of CDA scholarship and its fit within theoretically rigorous journals. Two studies (12%) were published in Mass Communication & Society, and one study each (6%) was published in Electronic News, the Journal of Magazine Media, the Journal of Media and Religion, and Visual Communication Quarterly.

This essay explores how CDA has been implemented in analyses of journalistic discourse within our field. No CDA studies that met the selection criteria were published in Journal of Advertising Education, Communication Methods & Measures, International Communication Research Journal, Communication Law & Policy, Journal of Media Ethics, Journal of Public Relations, Journal of Public Relations Education, Community Journalism, or Teaching Journalism and Mass Communication. Although some of those journals included CDAs, they did not focus on journalism specifically.

Coding

I coded the sample using first- and second-cycle coding—a grounded, open coding approach outlined by Johnny Saldaña (2015) in his book The Coding Manual for Qualitative Researchers. I began by precoding each article for its descriptive elements: The journal it was published in, author names, and the year it was published. As I precoded the sample, I started the first cycle of analysis (Saldaña, 2015), during which I read each study in full and began taking notes on initial themes I observed, called holistic codes. While conducting first-cycle coding, I analyzed each of the selected articles for eight elements: (1) the study’s topic(s) of analysis, (2) the study’s object(s) of analysis, (3) the study’s geographic focus, (4) the ideological positioning of the research objects (mainstream news or alternative media), (5) whether the researchers considered CDA a theory or method, (6) whether the researchers used CDA alone or mixed it with additional methods, (7) whether and how the researchers defined discourse, and (8) whether and how the researchers explained their coding procedures. I then conducted the second cycle of coding, focused on “classifying, prioritizing, integrating, synthesizing, abstracting, conceptualizing, and theory-building” (Saldaña, 2015). During the second cycle, I collapsed the initial codes into themes and reflected on what those themes mean in their research context.

Research topics

Journalism CDA research published in AEJMC journals focused on race and ethnicity as well as other representational problems involving marginalized groups. Eight (47%) of the 17 articles focused on racial issues. Studies with titles such as “‘Fatwa on the Bunny:’ News Language and the Creation of Meaning About the Middle East” and “Evolution of News Frames During the 2011 Egyptian Revolution: Critical Discourse Analysis of Fox News’s and CNN’s Framing of Protesters, Mubarak, and the Muslim Brotherhood” analyzed Orientalist discourses related to Islam, terrorism, and the Arab Spring. Others, such as “‘More Trouble than the Good Lord Ever Intended’: Representations of Interracial Marriage in U.S. News-Oriented Magazines” and “Teaching Journalism for a Better Community: A Deweyan Approach” addressed issues of blackness and racial representation in domestic media. It seems that our field understands CDA to be a useful tool for analyzing depictions of race and ethnicity in news coverage, as did van Dijk in his application of the ideological square.

However, hegemony was also analyzed in news coverage from linguistic rather than representational perspectives. Three studies (18%) focused on linguistic patterns of social domination—the “semantic and grammatical features of texts” used in Fairclough and Wodak’s approach to CDA (Philo, 2007). These analyses examined the supremacy of monolingual ideology in the U.S. press (Demont-Heinrich, 2007) and rhetorical strategies used in journalistic discourse. “Looking Back at Obama’s Campaign in 2008: ‘True Blue Populist’ and Social Production of Empty Signifiers in Political Reporting” examined the use of populist rhetoric in U.S. newspapers during Obama’s presidency, and “Journalistic Authority: Textual Strategies of Legitimation” analyzed adverb use as a type of reportorial bias.

The remaining six studies (35%) were a grab bag of representational analyses, analyzing topics including gender, military occupation, health care, religion, hate speech, and crime. Like Wodak and Meyer (2009) found in their analyses, our field uses “notions such as racist discourse, gendered discourse, discourses on un/employment, media discourse, populist discourse, discourses of the past, and many more” (p. 3). Like other CDA researchers, journalism scholars need to be transparent about our objects of analysis in our research projects.

Objects of analysis

In 11 of the manuscripts (65%), discourse was not explicitly defined, with the implied meaning being that discourse was the object of analysis (e.g., news content) in any given study. In these cases, CDA was listed as the method or research approach, but the meaning of discourse itself was not situated within a specific theoretical paradigm. For instance, in Robinson’s 2017 article on “Teaching Journalism for a Better Community,” the author’s unit of analysis is interview transcripts with journalists, but she does not situate discourse as a social process or a historical phenomenon. Rather, the reader is left to assume the discourse analyzed is conversational dialogue—not a more specific manifestation of power. Another example comes from Lipari’s (1996) study in Journalism and Mass Communication Quarterly, in which the author explains that news is one of many “forms of discourse” (p. 824), but does not explain how her unit of analysis—stance adverbs in news coverage—fits within broader theoretical approaches to discourse.

In the six journal manuscripts (35%) that situated “discourse” within specific worldviews, definitions varied between discourse as a social process: “a way for those who use language to build social realities and to choose among aspects of reality to be included” (Vultee, 2006, p.327); discourse as an apparatus of power: “a language of ideology with its own rules and conventions” (Kalyango, Myssayeva, & Mohammed, 2015, p. 153); and discourse as communication itself: “language, that is, discourse,” (Erjavic & Volcic, 2006, p. 303). It seems that many journalism scholars published in AEJMC journals understand discourse as a synonym for text, talk, or journalism, rather than as a complex cultural process.

Editorial focus

The majority of CDA studies published in AEJMC journals focused on mainstream print media published in the United States. Of the 17 studies coded, 13 (76%) analyzed U.S. media, such as leading national newspapers and magazines, or news published through English-language wires such as the Associated Press, United Press International, and Reuters. Four articles (24%)—“Mapping the Notion of ‘Terrorism’ in Serbian and Croatian Newspapers,” “Social Control in an American Pacific Island,” “‘You Don’t Understand, This is a New War!’ Analysis of Hate Speech in News Web Sites' Comments,” and “Jews in the News: Representations of Judaism and the Jewish Minority in the Norwegian Contemporary Press”—analyzed news content published in international news outlets. Like the journalism industry itself, journalism CDA research may be nationalist and statist, focusing on official news documents rather than citizen journalism or community-level discourse.

Only four studies (24%) integrated alternative media or counterhegemonic media into their analyses, but 13 (76%) focused on mainstream media at the international, national, or local level. Mainstream media included high-circulation newspapers such as The New York Times, the Washington Post, and the Wall Street Journal, as well as high-circulation consumer magazines such as Forbes and Ladies’ Home Journal. Other analyses of dominant-hegemonic media examined local professionalized outlets such as the Pacific News Daily—Guam’s only newspaper—or European national dailies. The four articles that examined “oppositional” media analyzed news content in feminist magazines, ethnic minority magazines, and online comments on newspaper websites.

Further demonstrating the AEJMC journals’ predilection toward dominant-hegemonic media, print was the medium of choice, as 11 studies (65%) focused on newspaper and magazine content, and one study (6%) examined news wires’ photovisual content. Two articles (12%) analyzed television content, and two more (12%) analyzed digital news. The remaining article—an outlier—coded interviews with news producers. Journalism researchers should be reminded that CDA is a critical-theoretical perspective that can be applied to any journalistic context, not only mainstream textual representations.

Research methods cited

Despite debates regarding whether CDA is a theory, method, or worldview, journalism studies scholars published in AEJMC journals typically treated CDA as a method writ large. Of the 17 sampled studies, 3 manuscripts (18%) described CDA as a theoretical perspective or worldview, while the remaining 14 manuscripts (82%) described CDA as a method or discussed CDA in their methods sections. Typically, manuscripts used a single analytical strategy—CDA if it was cited as the method or textual analysis-plus-coding for studies that described CDA as a theory or worldview. However, six manuscripts (35%) partnered CDA with another method. These mixed-methods approaches used CDA alongside quantitative content analysis, interviews, Google Trends data, questionnaires, semiology, and curriculum development, demonstrating the diversity of CDA approaches in mass communication.

CDA was typically explained as a tool for ideological critique, with thematic analysis guiding most of the studies in the sample. Two studies (12%; Lipschultz & Hilt, 2011; Luther & Rightler-McDaniels, 2013) included time-based analyses of themes in coverage. One study (6%) examined how journalists use stance adverbs to legitimize certain subjects in news sources (Lipari, 1996). It was the only article in the sample to focus specifically on grammar. Coding procedures were explicated in seven (41%) of the studies, providing readers with a structure for understanding how language or imagery was categorized into themes. Seven of the manuscripts (41%) pointed to vague coding methods such as thematic analysis and pattern identification, and the final three manuscripts (18%) failed to provide a method for coding at all.

Directions for CDA Research in Journalism Studies

CDA studies about news discourse published in AEJMC journals tended to focus on race and ethnicity in domestic mainstream media, and they tended to analyze content at the thematic level, rather than analyzing longitudinal and grammatical trends in discourse. In order to increase the perceived rigor of CDA research in mass communication, our field should intervene by: (1) producing more studies that examine discourse across the ideological media spectrum, not only within dominant-hegemonic mainstream media; (2) examining mediated oppression and stigmas beyond racism and Islamophobia, which might include representations of gender, sexuality, health, social and economic class, ability status, or religion, and so on; (3) sampling media from countries outside of the United States; (4) identifying the historical contexts that allow certain discourses to proliferate or disappear, and (5) comparing journalistic discourse with other types of mediated discourse to determine the ideological consistency of certain mediated messages. Crucially, we must also begin explicating our coding methods—not just by referencing Fairclough, Wodak, or van Dijk, but by outlining the specific procedures we use to identify ideological themes in content. In the following text, I provide guidelines for improving CDA research on journalism topics. In Figures 1 and 2, I offer complementary coding models that will help researchers conduct their analyses.

Sampling

Wodak and Meyer (2009) have asserted that “most of the approaches to CDA do not explicitly explain or recommend data-gathering procedures” (p. 28). I recommend that journalism scholars begin by proposing a research question and then identifying mainstream news sources, alternative media, and self-representational spaces in which that discourse is located (see Figure 2). After identifying research questions and relevant media, the researcher should begin with what Carvalho (2008) calls an “open-ended reading” to “help identify significant debates, controversies, and silences, and possibly suggest specifications and amendments to initial research goals and questions” (p. 166). After reading broadly on their research topic, the author should “characterize relevant genres” and “choose typical texts” (Wodak, 2010, p. 25) for their analysis, discarding media that do not especially exemplify their research problem. Because CDA should incorporate longitudinal analysis across ideologically diverse media, scholars will need to whittle down their corpus of data. Critical discourse analysts do not typically collect representative samples. Rather we purposively select stories in order to describe “critical discourse moments” (Carvalho, 2008; Richardson, 2006)—time periods during which knowledge about a specific topic appears to be growing or changing.

Ideological coding

Many methods for qualitative analysis rely on a loose web of iterative, ideologically invested tools. In order for CDA to be conducted in a systematic, empirical manner, and in order for critical discourse scholars in journalism to build a more comprehensive methodology, it is crucial to chart the coding methods we employ. It is neither enough to articulate that we code our data using frameworks such as Fairclough’s (2015) or Carvalho’s (2008) historical-diachronic approach, nor is it enough to describe what aspects of texts we code. We need to describe precisely how we do it. Previous CDA research in mass communication has fallen somewhat short of this charge. While interpretive approaches allow for flexibility, their lack of precision creates confusion among readers hoping to replicate coding methods in their own research.

Although CDA is often cited by journalism studies scholars as a method, coding is actually the process used to thematically analyze textual data in both qualitative and quantitative traditions. In The Coding Manual for Qualitative Research, Johnny Saldaña (2015) wrote: “a code in qualitative inquiry is most often a word or short phrase that symbolically assigns a summative, salient, essence-capturing, and/or evocative attribute for a portion of language-based or visual data” (p. 3). Coding is the method of applying those “essence-capturing” words and phrases to discourse in order to capture the meaningful qualities of the message. Coding is used primarily to develop theory (and is sometimes confused with the methods of grounded theory), though it is virtually impossible not to code data when using qualitative methods (Auerbach & Silverstein, 2003). In CDA, coding may include textual analysis of media representations or topics, identities and social relations, and message cohesion or coherence, and codes may also be quantified to illustrate trends in discourse (Richardson, 2006). In a 2013 essay, van Dijk implored researchers to “explain by what specific explicit methods” we use in our analyses. In the following section, I will describe methods for sampling and coding journalistic discourse using a comparative approach to CDA.

Coding procedures

Many qualitative scholars use diverse, multistep, iterative coding methods without articulating our research within a particular coding language. Saldaña offers a solution to the problem of describing qualitative coding methods. The Coding Manual for Qualitative Researchers is an encyclopedic review of hundreds of coding strategies Saldaña compiled from external sources. It is a useful companion text for all qualitative researchers and not only those working on CDA in journalism studies. Saldaña’s (2015) approach asks scholars to begin with open coding, develop categories from those codes, further refine categories into themes, and eventually build theory after considering the relationships between codes, categories, and themes. This procedure is similar to open coding and axial coding in grounded theory.

This essay builds a coding procedure for CDA based on Saldaña’s recommendations for qualitative coding. I have added ideological and longitudinal analysis to Saldaña’s (2015) “generic” coding methods (p. 48), and I have introduced transideological analysis as a primary intervention for CDA in journalism studies. Figure 1 provides an outline of the coding procedures a researcher should follow for coding a single data set of journalistic discourse, for example, stories published in mainstream media. Figure 2 demonstrates how first-and-second-cycle coding can be used to compare ideological formations in mainstream media, alternative media, and self-representational media, building theory about discourse across the ideological media spectrum. In comparative research, coding for each data set should follow an identical first- and second-cycle coding procedure, examining the same types of codes in each sample, then comparing their presence and prevalence across multiple data sets.

First-cycle coding methods. The CDA research project should begin with first-cycle coding, the initial interpretative process involved in codebook development. As outlined in Figure 1, journalism scholars should conduct first-cycle coding in this order: (1) attribute coding, (2) holistic coding, (3) descriptive coding, (4) values coding, and (5) versus coding. Precoding should begin in a codesheet such as Excel or Google Sheets, and more nuanced coding methods should continue in a qualitative analysis software such as NVivo. CDA research demands multiple passes over qualitative data. The researcher should conduct an initial coding run using Steps 1 through 5, returning to the data to confirm, alter, or reject codes as needed.

Attribute coding. Attribute coding is also known as precoding. Descriptive attributes of the data are systematically noted in the codesheet. Notation includes “basic descriptive information such as: the fieldwork setting, participant characteristics or demographics, data format, time frame, and other variables of interest” (Saldaña, 2015, p. 70) After an “open-ended reading” of a news sample (Carvalho, 2008), the researcher should code the descriptive attributes of each story, such as publication title, publication date, article headline, and author name, while selecting relevant articles for analysis. It is especially helpful to organize the codesheet chronologically (with sampled articles organized based on publication date), as this will help with second-cycle longitudinal coding during the next phase of research.

Holistic coding. Holistic coding identifies a story (or discourse unit’s) theme. It is similar to framing analysis in mass communication research. Holistic coding “applies a single code to each large unit of data in the corpus to capture a sense of the overall contents and the possible categories that may develop” (Saldaña, 2015, p. 118). After precoding the sample for basic attributes, code each story for the presence of one or two dominant themes—the holistic codes. Using more than one holistic code is called “simultaneous” coding (Saldaña, 2015), which is common practice in CDA. Attribute codes and holistic codes should be logged alongside each story’s title in the researcher’s codesheet.

Descriptive coding. Descriptive coding describes “what is talked or written about” (Wolcott, 1994, p. 119). It is often the first step in getting intimate with the problems of a study, and it is also essential for longitudinal analysis as the researcher analyzes topical changes over time (Saldaña, 2015). Although attribute coding and holistic coding are logged in the scholar’s codesheet, they will want to switch to a qualitative analysis software such as NVivo to track descriptive codes. Descriptive coding can be used to identify subjects of discourse, such as topics discussed, individuals referenced, sources cited, specific themes in news coverage, and events described. Most qualitative researchers naturally conduct descriptive coding as they get to know their data.

Values coding. Values coding accounts for “perspectives or worldview” (Saldaña, 2015) present in discourse. All CDAs therefore engage in values coding. The researcher may want to ask themselves: “Whose perspectives are being validated by this story?” Values coding will illustrate biases in media, such as misogyny, feminism, normativity, racism, and other forms of power embedded in representations. CDA scholars can conduct values coding to provide more nuanced insights into the construction of ideology surrounding news sources and news subjects described in media content. Values coding should also be conducted in a qualitative analysis software after descriptive coding.

Versus coding. Versus coding is dichotomous coding used to identify “individuals, groups, social systems, organizations, phenomena, processes, concepts, and so on, in direct conflict with one another” (Saldaña, 2015, p. 115). As the first cycle of coding nears its end phases, the researcher should reread each story to determine whether certain groups are pitted against one another in an us-versus-them storyline. This is especially useful for stories in which police, legislators, politicians, clergy, and other moral authorities surveil or condemn social “outsiders.” Versus coding can thus be used to identify the people and policies that produce certain ideologies in media discourse. This approach is particularly suited to “critical discourse analysis, and qualitative data sets that suggest strong conflicts or competing goals within, among, and between participants” (Saldaña, 2015, p. 115). As the reader may recall from the first section of this essay, van Dijk’s ideological square uses a type of versus coding.

Importantly, first-cycle coding may require “careful reading and rereading of your data as your subconscious, not just your coding system, develops connections that lead to flashes of insight (DeWalt & DeWalt, 2002)” (Saldaña, 2015, p. 47). The CDA researcher will likely proceed from first-cycle coding to second-cycle coding, then return to first-cycle coding to finesse coding categories (see Figure 1). After all, coding is an iterative method.

Second-cycle coding methods. After the scholar is initially satisfied with the first-cycle codes they have developed, they should begin looking for patterns and trends across stories in the sample. For this reason, second-cycle coding is also known as code synthesis, or collapsing codes. Second-cycle methods “develop a sense of categorical, thematic, conceptual, and/or theoretical organization” from those earlier developed codes (Saldaña, 2015, p. 149). Although Saldaña’s “generic” coding method asks researchers to synthesize patterns from first-cycle codes, CDA scholars must take their analyses one step farther, contextualizing patterns and themes within the broader lieu de mémoire. After all, “one important characteristic arises from the assumption of CDA that all discourses are historical and can therefore only be understood with reference to their context” (Meyer, 2001, p. 15). The researcher may find it helpful to conduct pattern coding in a qualitative analysis software, then return to her codesheet’s chronological attribute codes and holistic codes to determine how patterns have changed over time.

Pattern coding. Pattern coding corresponds with axial coding in grounded theory. Scholars look for meanings within and across their samples that cohere into metalevel patterns. Pattern codes “pull together a lot of material into a more meaningful and parsimonious unit of analysis” (Saldaña, 2015). This requires significant critical thinking. During this phase of second-cycle coding, the researcher synthesizes first-cycle codes under larger umbrella codes, which represent the dominant discursive formations of the stories examined. The scholar should ask herself how the first-cycle codes cohere into more parsimonious units of meaning.

Longitudinal coding. Longitudinal coding is a primary tool that separates CDA from other ideologically invested qualitative research approaches. While pattern coding will help surface trends in a body of discourse as a whole, longitudinal coding can evidence changes in discourse over time. Longitudinal coding corresponds with Carvalho’s (2008) historical-diachronic strategy for CDA. Journalism scholars should look for changes in the distribution of first-cycle codes and pattern codes in their news timeline. This constructivist approach acknowledges the instability of discourse. Longitudinal coding can also help identify paradigm shifts and essential discourse moments, which make compelling junctures around which to structure research findings.

After reading and rereading the data set to ensure codes are saturated, collapsed into parsimonious patterns, and organized along a meaningful timeline, the scholar should draft a rudimentary findings section. Qualitative findings should be explicated in their discursive contexts, paying careful attention to the roles of ideology, power abuse, and social dominance within and between news articles, journalists, news sources, and news subjects. CDA research should always be attentive to the “social processes of power, hierarchy building, exclusion and subordination,” according to Meyer (2001, p. 30), because “in the tradition of critical theory, CDA aims to make transparent the discursive aspects of societal disparities and inequalities.” When writing initial findings, journalism scholars should be attentive to the social construction of power through news discourse, and to its relationship with larger social processes.

Transideological Critique: A Comparative Approach

I have shown that journalism CDA research published in AEJMC journals has tended to reproduce nationalist perspectives, relying on mainstream media analysis to investigate topics related primarily to race and ethnicity during limited time periods. This confirms findings published decades ago, which found that discourse analyses of journalistic messages focused largely on mainstream media representations (Chen, 1989). It is crucial that journalism scholars understand that “through access to the mass media, dominant groups also may have access to, and hence partial control over the public at large” (van Dijk, 2013b, p. 86). If our field hopes to advance research about media ideology, CDA in journalism studies must embrace a methodological intervention. Our responsibilities are threefold: We must investigate ideological factors beyond race and ethnicity, incorporate longitudinal coding (described in the previous section), and move toward transideological analysis. Because mainstream media ideology is theorized as being equal to and exemplary of oppression (Ono & Pham, 2009), we must identify messages produced in order to challenge discourse found in mainstream journalism.

The comparative method proposed in Figure 2 helps researchers code for counterhegemonic as well as dominant-hegemonic discourses. If a primary goal of CDA is to visualize ideology and hegemony in discourse, we must establish the stability (or instability) of discursive formations among ideologically diverse media. CDA scholars in journalism should examine hegemony and counterhegemony in two situations: First, we should analyze discourse formations during paradigm shifts, during which outgroups bring a new discourse into the forefront through community solidarity over time (Hamilton, 1997); and second, we must examine discourses in vernacular media, which are media produced by and for local communities or marginalized groups (Ono & Pham, 2009). Paradigm shifts will become evident during longitudinal coding of a given data set. Vernacular media, however, must be analyzed alongside mainstream media in order to paint a fuller picture of ideological media formations. Vernacular media may include community newspapers and magazines such as the black press or lesbian, gay, bisexual, transgender, and questioning–alt weeklies. They can also include self-representational media such as online forums, where the regulatory function of mass dissemination does not limit the representations presented. On one hand, analyzing vernacular media helps scholars visualize the marginalized imaginary. On the other hand, vernacular media are subject to many of the same limitations of production, distribution, and regulation that burden the mainstream press (for an overview, see Tuchman, 1978 and Schudson, 2003). These narratives can be, and often are more complex, but they do not always carry a counterhegemonic function (Ono & Pham, 2009).

Ideological analysis

The coding model provided in Figure 2 directs CDA researchers to sample media across the ideological media spectrum, starting with a systematic analysis of discourse in mainstream newspapers or mass media (see Figure 1). The researcher should then follow up their mainstream media analysis by coding messages published about the same topics in alternative and online media, making direct comparisons between codes present in mainstream journalism, alternative media, and self-representational discourse. Based on the models provided in Figures 1 and 2, the scholar will eventually produce a number of pattern codes and longitudinal codes that describe discourse across their data sets. The final stage of analysis will describe similarities and differences in discourse observed across the ideological media spectrum. Codes should evidence the prevalence and distribution of power in ideologically diverse media. Similarities between codes across the samples may indicate the presence of hegemonic logics in “counter-hegemonic” media (see Ono & Pham, 2009). Differences between codes across the samples may demonstrate the outcomes of meta- and microlevel sociological influences on content (see Tuchman, 1978). Findings should make links between power structures, journalistic production, and representations in media content, both thematically and over time. Theoretical conclusions should examine the roles mass media, alternative media, and self-representational media play in maintaining and challenging ideological formations.

Conclusions

This essay’s three purposes were to explain CDA to journalism scholars, demonstrate trends in CDA research in journalism studies, and provide a conceptual roadmap for coding mass communication discourse in the CDA tradition. In the first half of the essay, I offered a brief background on the foundations of CDA research, then I outlined themes in each of the 17 CDA articles published in AEJMC division and interest group journals since 1996. Those articles tended to focus on representations of race and ethnicity in mainstream domestic media, tended not to explicitly describe their objects of analysis or coding methods, and tended to focus on thematic analyses rather than linguistic analyses of discourse. The exploratory analysis of AEJMC articles points to the need for comparative, transideological research in journalism studies, and for transparency and rigor in coding.

The second half of this essay followed by making an argument for a systematic coding method for CDA in journalism studies and for the implementation of a transideological, multimedia approach to our research problems. In Figure 1, I bridge gaps in journalistic research by offering an empirical coding method CDA scholars can cite and apply to their own data sets. Journalism research should follow Saldaña’s guidelines for qualitative analysis, focusing specifically on the first-cycle method of coding for attribute, holistic, descriptive, values, and versus codes, and progressing into the second-cycle method of coding for longitudinal and pattern codes. CDA researchers may cite this method and coding model in their own work, increasing transparency about how we code our data.

The method in Figure 2 helps scholars identify and interrogate counterhegemonic discourses, which have been primarily articulated as an outcome of alternative media and online media (see Althusser, 1971; Ono & Pham, 2009). Although alternative media and online media may sometimes challenge normative discourse, this is not always the case. The method proposed in this essay will help future researchers question whether alternative media are necessarily oppositional and whether the techno-utopian discourse prevalent in the past decade is warranted. This essay contributes a method to help build knowledge about ideology in communicative technologies, contributing to theories based on the ideological consistency of discourse across media contexts.

The comparative approach outlined in this essay demands sampling discourse across multiple types of media. Therefore, it is essential to advance research questions about topics that create controversy in ideologically diverse contexts, and thus drive coverage in alternative media and online forums as well as mainstream newspapers and magazines. CDA “focuses on social problems, and especially on the role of discourse in the production and reproduction of power abuse or domination” and provides a voice for the oppressed (van Dijk, 2001, p. 96). However, oppression is not merely race-based, and journalism researchers should address gender, sexuality, socioeconomic class, religion, health and mental health, and other intersectional aspects of identity. Comparative CDA should develop theory from comparisons of representations over time and across the ideological media spectrum, paying careful attention to how marginalized groups represent themselves. As Wodak and Meyer (2009, p. 2) remind us:

The objects under investigation do not have to be related to negative or exceptionally “serious” social or political experiences or events; this is a frequent misunderstanding of the aims and goals of CDA and of the term “critical,” which, of course, does not mean “negative” as in common-sense usage.

Because CDA demands that researchers sample vast data over time and media, it is not an “easy” method for qualitative scholars to tackle. It demands intense critical thinking and a rigorous method for winnowing data. Carvalho (2008) recommends that we focus on discourse shifts, or what she calls “critical discourse moments.” Because our data sets may be so large, “the analysis of those ‘moments’ allows for the identification of discursive turns and/or continued lines of argumentation at particularly important times in the social construction of an issue” (Carvalho, 2008). CDA is not a method for scholars on tight deadlines. CDA projects require significant investment, both in terms of data sampling and time committed to analysis. Researchers would do well by investing in an NVivo subscription, given the scope of codes developed in this version of CDA.

Limitations and Future Research

This essay offered recommendations for conducting CDA scholarship in journalism studies. The exploratory analysis outlined in the first half of this article paints a picture of how journalism-focused CDA research has been articulated in AEJMC division and interest group journals. Based on the professional orientation of Journal of Communication Inquiry’s audience, I did not sample journalism CDA research published in International Communication Association or National Communication Association journals. It is possible that the critiques I articulated are limited to studies selected for publication by AEJMC editors and peer reviewers. Further, because I selected studies that matched a keyword search for “critical discourse analysis,” it is possible I missed articles that approached research from a CDA orientation, but that did not label their theoretical or methodological commitments as “critical discourse analysis.” Future research should examine a larger data set of journalism-focused CDA research published across disciplinary boundaries.

As a mass communication scholar trained in qualitative research traditions, and as a professional journalist-turned academic, I am not fluent in CDA approaches used across communication disciplines. The methodological interventions proposed in this essay will prove helpful for academics working on problems of representation at the intersections of journalism studies and cultural studies, but may not be appropriate for scholars building theory in the linguistic tradition, or working with other forms of mass media (e.g., strategic communication, computer-mediated communication, etc.). Politically motivated activist-scholars in diverse communications subfields should consider building methodologies for CDA appropriate to their areas of inquiry. At the time of publication, I was preparing to host a Blue Sky Workshop on CDA methods at the meeting of the International Communication Association in Prague, May 2018, to begin advancing methodological inquiry concerning CDA across the broader communication research profession. I look forward to reporting the outcomes of the workshop in future research.

Appendix: CDA Articles Sampled from AEJMC Journals

Dalisay, F. (2009). Social control in an American Pacific Island: Guam’s local newspaper reports on liberation. Journal of Communication Inquiry, 33(3), 239–257.

DeFoster, R. (2015). Orientalism for a new millennium: Cable news and the specter of the “Ground Zero Mosque.” Journal of Communication Inquiry, 39(1), 63–81.

Demont-Heinrich, C. (2007). Globalization, language, and the tongue-tied American: A textual analysis of American discourses on the global hegemony of English. Journal of Communication Inquiry, 31(2), 98–117.

Døving, C. A. (2016). Jews in the news—Representations of Judaism and the Jewish minority in the Norwegian contemporary press. Journal of Media and Religion, 15(1), 1–14.

Erjavec, K., & Kovačič, M. P. (2012). “You don’t understand, this is a new war!” Analysis of hate speech in news web sites’ comments. Mass Communication and Society, 15(6), 899–920.

Erjavec, K., & Volcic, Z. (2006). Mapping the notion of “terrorism” in Serbian and Croatian newspapers. Journal of Communication Inquiry, 30(4), 298–318.

Guzman, A. L. (2016). Evolution of news frames during the 2011 Egyptian Revolution: Critical discourse analysis of Fox News’s and CNN’s framing of protesters, Mubarak, and the Muslim brotherhood. Journalism & Mass Communication Quarterly, 93(1), 80–98.

Izadi, F., & Saghaye-Biria, H. (2007). A discourse analysis of elite American newspaper editorials: The case of Iran’s nuclear program. Journal of Communication Inquiry, 31(2), 140–165.

Jenkins, J. M., & Johnson, E. K. (2017). Body politics: Coverage of health topics and policy in U.S. feminist magazines. Mass Communication and Society, 20(2), 260–280.

Kalyango Jr., Y., Myssayeva, K. N., & Mohammed, A. (2015). Visual representation of Shiite Muslim mourning rituals. Visual Communication Quarterly, 22(3), 146–159.

Kumar, A. (2014). Looking back at Obama’s campaign in 2008:“True Blue Populist” and social production of empty signifiers in political reporting. Journal of Communication Inquiry, 38(1), 5–24.

Lipari, L. (1996). Journalistic authority: Textual strategies of legitimation. Journalism & Mass Communication Quarterly, 73(4), 821–834.

Lipschultz, J. H., & Hilt, M. L. (2011). Local television coverage of a mall shooting: Separating facts from fiction in breaking news. Electronic News, 5(4), 197–214.

Luther, C. A., & Rightler-McDaniels, J. L. (2013). “More trouble than the Good Lord ever intended”: Representations of interracial marriage in U.S. news-oriented magazines. Journal of Magazine & New Media Research, 14(1).

Marcellus, J. (2006). Woman as machine: Representation of secretaries in interwar magazines. Journalism & Mass Communication Quarterly, 83(1), 101–115.

Robinson, S. (2017). Teaching journalism for better community: A Deweyan approach. Journalism & Mass Communication Quarterly, 94(1), 303–317.

Vultee, F. (2006). “Fatwa on the bunny:” News language and the creation of meaning about the Middle East. Journal of Communication Inquiry, 30(4), 319–336.

Footnotes

Declaration of Conflicting Interests

The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author received no financial support for the research, authorship, and/or publication of this article.

Note

Author Biography

Chelsea Reynolds, PhD, is an assistant professor of communications at California State University, Fullerton, where she researches sexuality and gender in media and teaches journalism. Her work has been recognized with numerous awards, including the AEJMC Mary Ann Yodelis Smith Award for Feminist Scholarship and the AEJMC Magazine Media Research Grant. Dr. Reynolds is a former magazine journalist and health reporter. She currently lives in Los Angeles, California.

References

Althusser

(1971). Ideology and ideological state apparatus (Notes towards an investigation). In Althusser

(Ed.), Lenin and philosophy and other essays (pp. 127–186). New York, NY: Monthly Free Press.

Auerbach

Silverstein

L. B.

(2003). Qualitative data: An introduction to coding and analysis. New York, NY: NYU Press.

Carvalho

(2008). Media(ted) discourse and society: Rethinking the framework of critical discourse analysis. Journalism Studies, 9(2), 161–177.

Chen

K. H.

(1989). Deterritorializing “critical” studies in “mass” communication: Towards a theory of “minor” discourses. Journal of Communication Inquiry, 13(2), 43–61.

DeWalt

K. M.

DeWalt

B. R.

(2002). Participant observation: A guide for fieldworkers. Walnut Creek, CA: AltaMira Press.

Du Gay

Hall

Janes

Madsen

A. K.

Mackay

Negus

(1997). Doing cultural studies: The story of the Sony Walkman. Thousand Oaks, CA: Sage.

Fairclough, N. (2001). Critical discourse analysis as a method in social scientific research. Methods of critical discourse analysis, 5, 121–138.

Fairclough

N. L.

(2014). Language and power (3rd ed.). New York, NY: Routledge.

Fairclough

N. L.

Wodak

(1997). Critical discourse analysis. In van Dijk

T. A.

(ed.), Discourse Studies. A Multidisciplinary Introduction: Vol. 2. Discourse as social interaction (pp. 258–284). London, England: Sage.

10.

Foucault

(1978). Nietzsche, genealogy, history. Semiotexte, 3(1), 78–94.

11.

Golafshani

(2003). Understanding reliability and validity in qualitative research. The Qualitative Report, 8(4), 597–606.

12.

Gramsci

(1971). Selections from the prison notebooks (Q. Hoare, & G. N. Smith, trans.). New York, NY: International Publishers.

13.

Hall

(1997). Representation: Cultural representations and signifying practices ( Vol.2). New York, NY: Sage.

14.

Hamilton

(1997). Representing the social: France and Frenchness in post-war humanist photography. In Hall

(Ed.), Representation: Cultural representations and signifying practices (Vol. 2, pp. 75–150). New York, NY: Sage.

15.

Jensen, K. B., & Neuman, W. R. (2013). Evolving Paradigms of Communication Research: Editorial Introduction. International journal of communication, 7, 9.

16.

Lovejoy

Watson

B. R.

Lacy

Riffe

(2014). Assessing the reporting of reliability in published content analyses: 1985–2010. Communication Methods and Measures, 8(3), 207–221.

17.

Meyer

(2001). Between theory, method, and politics: Positioning of the approaches to CDA. In Wodak

Meyer

(Eds.), Methods of critical discourse analysis (pp. 14–31). London, England: Sage.

18.

Neijens

(2017). Working on the fundamentals of journalism and mass communication research: Advancing methods in Journalism & Mass Communication Quarterly, 2007–2016. Journalism & Mass Communication Quarterly, 94(1), 12–16.

19.

Ono

K. A.

Pham

(2009). Asian Americans and the media ( Vol.2). Cambridge, England: Polity Press.

20.

Philo

(2007). Can discourse analysis successfully explain the content of media and journalistic practice? Journalism Studies, 8(2), 175–196.

21.

Ray

L. J.

(1979). Critical theory and positivism: Popper and the Frankfurt School. Philosophy of the Social Sciences, 9(2), 149–173.

22.

Richardson

J. E.

(2006). Analysing newspapers: An approach from critical discourse analysis. New York, NY: Palgrave Macmillan.

23.

Saldaña

(2015). The coding manual for qualitative researchers. Thousand Oaks, CA: Sage.

24.

Sandelowski

(1986). The problem of rigor in qualitative research. Advances in nursing science, 8(3), 27–37.

25.

Schudson

(2003). The sociology of news. New York, NY: W&W Norton.

26.

Tuchman

(1978). Making news: A study in the construction of reality. New York, NY: Free Press.

27.

van Dijk

T. A.

(1989). Structures of discourse and structures of power. Annals of the International Communication Association, 12(1), 18–59.

28.

van Dijk

T. A.

(1990). Discourse & Society: A new journal for a new research focus. Discourse & Society, 1(1), 5–16.

29.

van Dijk

T. A.

(1998). Opinions and ideologies in the press. In Bell

Gerret

(Eds.), Approaches to media discourse (pp. 21–63). Chichester, England: Wiley.

30.

van Dijk

T. A.

(2001). Multidisciplinary CDA: A plea for diversity. Methods of Critical Discourse Analysis, 1, 95–120.

31.

van Dijk

(2013a). CDA is NOT a method of critical discourse analysis. Retrieved from http://www.edisoportal.org/debate/115-cda-not-method-critical-discourse-analysis

32.

van Dijk

T. A.

(2013b). Discourse, power and access. In Caldas-Coulthard

C. R.

Coulthard

(Eds.), Texts and practices: Readings in critical discourse analysis (pp. 93–113). London, England: Routledge.

33.

Wetherell

Taylor

Yates

S. J.

(2001). Discourse theory and practice: A reader. London, England: Sage.

34.

Wodak

(2010). Critical discourse analysis: Some important concepts and considerations. Paper presented at the Accept Pluralism Second Project Meeting, University of Milan, Italy. Retrieved from http://accept-pluralism.eu/ACCEPT/Events/ACCEPTPLURALISMMeetings/SecondMeeting

35.

Wodak

Meyer

. (Eds.). (2009). Methods for critical discourse analysis. Thousand Oaks, CA: Sage.

36.

Wolcott, H. F. (1994). Transforming qualitative data: Description, analysis, and interpretation. Thousand Oaks: Sage.