Content analysis approach in critical discourse analysis of newspapers’ news

This paper aims to explain how the content of the analysis can be used as a framework for investigating the ideologies contained in the text, especially in the reporting of the election. The problem in the current ideological discourse analysis is that there is no explicit and systematic method of explaining ideology and discourse criticism to be unscientific. The research question in this paper how the content analysis framework is presented in the news? How the content analysis framework consists of a content analysis framework leading to prescriptive objectives in the form of a news content unit? In this paper, research data were taken from the 2018 West Java governor election campaign news text from online newspapers. The results of the research were news content raised topics such as social actors, namely figure and groups, and social practice, namely politics, and social actions. The ideology presented in the news was an imbalance in terms of the frequency of reporting in the news. The imbalance result was related to asymmetric power ideology in the newspaper.


INTRODUCTION
Newspapers in conveying their ideological message must abide by professional norms and news values. The difference between newspaper news and social media news is that newspaper news has filters different from social media news that has a direct communication message with its readers (Engesser, Ernst, Esser, & Büchel, 2017). In this way, newspaper news does not have the freedom to articulate their ideology and spread their message directly. The study of newspaper ideology was dominated by the study of media, communication, and culture in the 1970s and 1980s (Corner, 2016). In the 1990s and 2000s, ideological studies were dominated by studies of discourse criticism (Wodak & Meyer, 2015).
Research on media ideology has been examined with the theme of political language and mass media (Hamad, 2010) and discussed the bias of mass media in campaign coverage (Aboura, 2005). Chiang and Knight (2011) studied the alignments from newspapers, and Weber and Thornton (2012) revealed the religion in the context of political campaigns. Aduradola and Ojukwu (2013) describe political propaganda and religion during campaigns through language. Ross and Leonard (2008) focused on political communication while Wati, Tallapessy, and Diana (2015) emphasized the use of appraisal analysis in looking at the mass media alignments. Yuan (2016) exposed the bias from the mass media, Bischof and Senninger (2018) explain the campaign message, and (Handayani, 2019) exposed the verbal realization in the presidential election candidate debate.
The concept of ideology can be said to be the entrance of the relationship between power and dominance in terms of critical analysis (Fairclough, 2013). Definitively, ideology is a system of ideas and values and is oriented towards beliefs to explain certain political arrangements, legitimize existing hierarchies and power relations, and preserve group identity (Chiapello & Fairclough, 2002). Based on the critical concept of Fairclough, Ideology explains the horizontal structure of society (relations between communities) and the vertical structure of newspapers (the separation of rulers and those governed). This ideological concept is different from the concept of ideology in political terms (Corner, 2016). The concept of political ideology. This is very different from the descriptive use where the idea of ideology does not lie in the separation of rulers and those governed (domination) The background of the research is that the previous discourse analysis has limitations because it selects and uses only a small amount of text. This leads to concerns in the representation of the selected text, and thus vulnerability to the researcher's bias in the selection of texts for the intended analysis. In this paper, the research question is: can the ideology of newspaper news be explained in detail through a content analysis framework? The problem of research in more detail is how the framework mechanism and criticism of content analysis can reveal the ideology of the newspaper.

Newspaper Discourse
News and market share of newspapers not only provide news to the public but also provide information. Besides, newspapers consider politics, culture, and economics in their communication globally integrated (Machin & Van Leeuwen, 2007). News from newspapers is submitted to the editor by looking at political persuasion following the rules of journalism, so that news must be politically neutral, informative, and pure facts. In news discourse, the more powerful social actors regulate how ideas are put into practice and used to regulate the conduct of others. Mullainathan and Shleifer (2005) divided the news market share with two assumptions: 1) readers hold beliefs about what they like to be confirmed, and 2) newspapers can influence readers' views. Nie et al. (2010) explain that the market share of online newspapers, in terms of production costs, is lower in production costs than online newspapers. Therefore, the tendency of readers to look for newspapers that are under their own beliefs (ideology) can quickly, so that online newspapers target readers as their political views. News forms currently take the form of mainstream media and new models of newspapers such as online newspapers, including internet news networks, radio talk, and amateur bloggers (Baum & Groeling, 2008). New forms of newspapers tend to refer to and compare their stories with news from general newspapers.
Discourse functions as a re-contextualization of social practices and socially constructed knowledge about (several aspects) of reality are developed in certain social contexts and in a manner that is following the interests of social actors. The initial foundation of discourse is the knowledge that is ultimately based on social practice, on shared concrete experiences. In this case, the practice of discourse is the practice of observing, recording, and carrying out mathematical and linguistic operations (Bernstein, 2006). Ideology in the media is a conceptualization of journalism as a professional ideology (Deuze, 2004). Ideology is part of the journalist's self-perception and is not merely a bias.
Furthermore, discourse talk about the properties of language use, focus on a larger unit than words and sentences (text, discourse, conversation, speech acts, or communicative events), emphasis outside of grammar sentences leading to the study of actions and interactions (Wodak & Meyer, 2015). Critical discourse analysis can be in the form of an expansion of aspects of non-verbal interaction (semiotic, multimodal, visual) and communication: movements, images, films, internet, and multimedia, focusing on dynamic (social) -cognitive or interactional strategies and interactions. CDA describes the function (social, cultural, conditional, and cognitive) context of language use, is a phenomenon of text grammar and language use: (coherence, anaphora, topic, macrostructure, speech act, interaction, turn-taking, signs, politeness, argumentation, rhetoric, mental models, and many other aspects of the text and discourse) A discourse could be put into binary opposition systems where one side is always given a higher status, marginalizing the other, such as ('white' versus 'black'; 'citizen' versus 'non-citizen'; etc.). Things like this can make diversity and complexity invisible. The ideological analysis was developed through an analytical framework of the theory and method of language analysis concerning power (Fairclough, 2013). This framework, seen from the aspect of the struggle against domination and oppression in the form of linguistics. The concept of ideology is asymmetric conceptualized power between participants in discourse events with unequal capacity. Within the critical paradigm, ideology and inequality, and power figures are socio-terrifically insights in language studies (Blommaert & Bulcaen, 2000). In this crucial paradigm, the ideological analysis framework is an investigation and elaboration of theoretical linguistics. The theoretical framework of ideology is a multidisciplinary method that is articulated by the triangulation of discourse, cognition, and society (Van Dijk, 2006). This framework is very critical and rests on the nature and structure of socio-cognitive. Ideological analysis with big data is to use the corpus linguistics methodology, where qualitative analysis is done to identify the context when words are used (Orpin, 2005).
In discourse, an explanation of identity is how the community (people) who play a role, in this case, there are three very different types of social groupings namely: a society with identity as 'Symbol of success,' as 'Community ties,' and as 'Urban intelligence.' This identity arises when the media looks for new ways to create market demand, not as 'demographics' but as 'psychographic,' 'behavioral' groups, 'attitudes,' and consumption patterns. Some of the identities are 'Actualizers,' who are skeptical of advertising, have large financial resources, and read the newspaper every day; 'Experiencer,' who follows fashion, buys on impulse, and listens to a lot of music; or 'Striver,' which is status-oriented and spends a lot of money on leisure activities.

The Content Analysis Framework
Content analysis is a research method that uses a series of procedures to make valid conclusions from texts (Weber, 1990). The outcome is how the inferential process of the message sender, the message itself, and the audience. Content analysis is a qualitative research technique that shows three different approaches: conventional, directed, or summative (Hsieh & Shannon, 2005). These three approaches are used to interpret the meaning of the text and adhere to the naturalistic paradigm. Content analysis is a research method used to analyze content in various formats (e.g., books, newspapers, television, internet) that aims to understand message patterns (Manganello & Fishbein, 2010). Understanding content analysis is a systematic analysis of the text, images, and symbolic material (Krippendorff, 2018).  (Krippendorff, 2018) Content analysis is a very flexible research method that has been widely used in library studies and information science with various research goals and objectives (White & Marsh, 2006). In the process, content analysis is a strategy and has developed techniques and approaches for analyzing texts known as extensive textual analysis. A critical aspect of content analysis, as revealed by Lombard, Snyder-Duch, and Bracken (2002) and Elo et al. (2014) is that content analysis is a method, theory, logic in its use. The further content analysis measures computational reliability to categorize units (programs, scenes, articles, stories, words, etc.), then use this categorization to calculate indexes. Text is qualitative, even Data; Text Input

Content Analysis
Contexts Inference Result Output when certain characteristics of a text can then be converted into numbers. Therefore, the use of computer software can process large volumes of text in a concise time and not eliminate the qualitative nature. With a content analysis framework, it can analyze topics in the form of politics, foreign propaganda, ethnographic research, and discourse analysis, with the help of computer text analysis (Krippendorff, 2018) Content analysts have a starting point for analysis in the form of text (printed material, recorded speeches, visual communication, artwork, artifacts) (Krippendorff, 2018). Content analysts recognize that all texts are produced and read by others and are expected to be significant for them, not just for analysts. In content analysis, there is a framework on which to base the methodology. This framework is intended to serve three objectives as follows: • The prescriptive objective is to guide the conceptualization and research design.
• The practical analytic content and the analytical purpose is to facilitate a critical examination and comparison of the analysis of published content. • The methodological objective is to show performance and caution criteria that researchers can apply in evaluating ongoing content analysis. In content analysis, the initial step is to determine the text to be analyzed. Text is the data available for content analysis to start analytical efforts. Next, the research questions to be answered will be examined through the context of the text in understanding the contents of the text. Prasad (2008) describes what characteristics of content are represented? How? And for whom? What research question? The next step, to start analytical construction is to operationalize what is known about the context to answer the research question. The next step in content analysis is to validate the evidence, which is the main justification for content analysis.

METHOD
This research used a content analysis study. The use of content analysis methods in this study was a qualitative method consisting of textual language analysis, contextual analysis, and computational analysis approaches.

Design
The content analysis described analytical construction is to operationalize the context to answer the research question with AntConc software help. The content analysis used to validate the evidence to justify the result itself. Data sources were taken from newspapers produced in Indonesia. Data sources were taken from news from online newspapers during the 2018 West Java Governor Election campaign. The corpus of data was from the Pikiran Rakyat online amounted to 331 news articles with data collection periods during the official campaign period of 15 February 2018 to 23 June 2018. Data collection procedures in research using systematic documentation retrieval techniques. This procedure consists of collecting and sorting text data based on news on the campaign topic of the 2018 governor election in West Java with a predetermined time susceptibility. This technique refers to a systematic sampling technique. The systematic sampling method is the selection of data units from a sample list. It allows researchers to obtain language and textual words, present weighted data, and save researchers time in transcribing data (Creswell, 2012). In taking data, a systematic sample is when the text comes from publications that appear regularly, such as newspapers, television series, or other recurring or ongoing events. In obtaining all the news on the internet according to the desired period, the search for newspaper news pages used the Google search engine. After the news newspaper pages were obtained in digital form (html), the last step in data retrieval was the conversion of data from html to txt files with the help of 'html to text' software.

Procedure
In general, the procedure in this study used the following steps: • Sampling: relying on a sampling plan • Unitization relies on unitization and coding schemes: relies on coding instructions • Reducing data into manageable representations to simplify data. • Summing up contextual phenomena: relying on analytic constructions or context models chosen as warrants • Tells the answer to a research question: relying on narrative or discursive traditions of discourse

Data Analysis
Data in the form of a corpus in the form of a .txt file is entered into the corpus linguistic processing machine (Anthony, 2018). After a concordance engine processes the data, the next step in data processing is searching for word lists to determine in the form of contextual which words stand out most in the news based on their frequency. This process is essential because to explain all the research questions, the context of the text is the focus of the analysis. The context of the text in this study can explain the media ideology in the form of alignments in the frequency of reporting. Context is the construction of a person, who is in the conceptual environment of the text, and the situation in which it plays a role. In other words, in the framework of content analysis, the context is ideas, actors, and roles in news texts. It can be concluded that the context in question is the actors who are candidates for the 2018 elections in West Java, for example

RESULTS AND DISCUSSION
The first stage of analysis was analyzing the content of newspaper news. This analysis aimed to see the tendency of the topic and content when it is delivered. This analysis uses the help of linguistic concordance analysis tools, namely AntConc. The content analysis framework in this research had many steps, including determining unit, coding, and context unit. If the conventional news content analysis uses humans as the code, in this study, the analysis steps used the help of corpus linguistic tools as a coder. In the analysis, there were 331 corpus units analyzed. Furthermore, the main corpus units analyzed based on word lists and keywords. For the context analysis, word list and keywords were given context based on semantic category units so that meaningful contexts were obtained (Fajri, 2019). The table in the appendix shows the 50 highest keywords in the main corpus of total occurrences-the main corpus keywords number 729 from the Pikiran Rakyat.com newspaper. Because the keyword method can show the contents of the corpus and the topics discussed, Keynes analysis requires words with the same semantic meaning or the same grammatical function (Baker, 2006). In this analysis, keyword categorization is needed as a reference for discourse analysis. Referring to this, all the main corpus keywords that appear in AntConc need to be categorized according to their semantic context. This can be explained by the keywords that appear are categorized semantically based on USAS or UCREL Semantic Analysis System (Rayson, Archer, Piao, & McEnery, 2004). The next step in data analysis is the conclusion of the analysis results. This inference answers problems from research in the form of newspaper ideology. Inferences made in this study are inductive with conclusions based on the results of the semantic category and from theories about the media and theories from previous research.
In names and grammatical words, the concept was named itself referred to as the actor of the process. Actors, just like in discourse, were represented by rather economy, social, political control, and domination. Actors in the news discourse were described as powerful or less power in social and political status that dominated the entire discourse. In this study result, social actors can be referred to as individuals through a process of social practices of political and social domination-the social actors represented by personalization. The personalization of social actors found in the study was represented as a human being through proper names involved in the news discourse (e.g., Ridwan Kamil, Emil, rindu, dedi mulyadi deddy Mizwar, etc. In the aspect of participation, the main social actor was described as the main essential participant in the news discourse represented by many coverages. The other social actors found in the news discourse described as non-essential social actors. A small number of coverages represented the non-essential participants. In other words, the results from the news discourse showed that the domination (asymmetry ideology) from news coverage only covered one of the campaign candidates (social actor), and reflected the entire election result as the news became strategic tools to promote the election campaign.
In describing news discourse, actors were not realized through agentive roles, but through other processes, and one of the processes was social practices. Social practices in news discourse described social process categories rather than Social actor's categories. From the semantic categories perspective, social practices were the processing activity or status of the participant in the news discourse (e.g., Calon (Candidate), Pasangan (Pair in candidate), Pilgub (Governor Election), Kampanye (Campaign), Wakil (Vice), Debat (Debate), etc.). Social practices found the study described the actors and their social action or status in the hierarchy (e.g., Gubernur (Governor, Kpu (Election Watch), Partai (Political Party), etc.). Social practices found in the study showed that the main social action or status represents the actors' position (as election candidate). The social practices described political activity, social relationships, identities, and other values. Other social practices found in the study are related to other social actions or status reflected by other actors or participants. These social practices described as embedded social practices because it was reflected by the domination of the main social practices. The hegemonic of social practice described the process of the social actor being a candidate (calon) of the election campaign dominated the entire discourse. This hegemony results best described to promote the candidate, and it was one of the objectives of the news discourse social process.

CONCLUSION
This research was a content analysis research that aimed to find how newspapers represent their ideology in the 2018 West Java governor election. With a content analysis framework, there was unitization analyzed from word lists, keywords, and semantic categories. The conclusion of this research was the discovery in the preaching of the People's Mind online; some topics stand out, topics such as actors represented with social figures, groups, and other topics like social context namely politics, and elections. It was found in many online newspaper reports that one of the election participants was the most mentioned or reported on the election. In other words, the news content of the online newspaper is dominated by the news of one of the eligible participants. And the contents of newspapers in the form of news topics can be concluded that the newspapers favor one candidate more than the other eligible participants. Overall, the ideology displayed by the newspaper is realized and explained through asymmetric news content. The critical dimension was described from the news discourse contents. The discourse contents found in the study was the description of the relationship from social actors (participant) with the social practices (action and status).