Data journalism in combating misinformation during the 2018 Bangladesh national election

Размещено на http://www.allbest.ru/

Data journalism in combating misinformation during the 2018 Bangladesh national election

Sarkar Barbaq Quarmal

Md. Aminul Islam

Abstract

This paper explores the following: (1) the use of data by Bangladeshi dailies in terms of statistics and/or infographics in covering the National Election of 2018; (2) the role data journalism played in combating spread of misinformation during the said event. The study is exploratory and descriptive in design. To conduct the study, a mixed method approach has been adopted. Data have been gathered through quantitative and qualitative content analysis of news content published during the election (15 days before and after the election day) in two most widely reached dailies (Prothom Alo, a Bangla daily and The Daily Star, an English daily). Alongside, five in-depth interviews of journalists having experience of covering elections and one journalism faculty member with appropriate academic background have been conducted. The quantitative data were analyzed using SPSS and the qualitative data were studied through thematic analysis and discourse analysis.

Key words: data journalism, misinformation, fake news, elections, Bangladesh.

Аннотация

journalism disinformation newspaper

Мы живем в эпоху датификации. С развитием информационно-коммуникационных технологий (ИКТ) все аспекты человеческой жизни, взаимодействия и производства претерпевают быстрые изменения и часто пересматриваются; сфера журналистики не является исключением из этого правила. Некоторые недавние явления, такие как распространение лжи и/или дезинформации во время президентских выборов в США в 2016 г. и национальных выборов в Бангладеш в 2018 г., вызвали тревогу новостных организаций по всему миру и обострили вопрос о ценности рассказывания историй, основанных на данных. Это, с одной стороны, дает читателям возможность принимать обоснованные решения, а с другой - позволяет журналистам представлять факты более точно, достоверно, привлекательно и увлекательно. Обширная практика журналистики данных не является старым явлением во всем мире; то же самое относится и к академическим дискуссиям по этому поводу. Но в развивающихся странах, таких как Бангладеш, это концепция новая. Если быть более точным, то практика журналистики данных все еще находится в зачаточном состоянии в стране, и мало что было изучено относительно ее динамики с академической точки зрения.

Таким образом, в данной статье рассматриваются следующие вопросы:

(1) использование данных бангладешских ежедневных газет с точки зрения статистики и/или инфографики при освещении национальных выборов 2018 года;

(2) роль информационной журналистики в борьбе с распространением дезинформации во время указанного события. Исследование носит исследовательский и описательный характер. Для проведения исследования был принят смешанный методический подход. Данные были собраны с помощью количественного и качественного контент-анализа новостного контента, опубликованного во время выборов (за 15 дней до и после дня выборов) в двух наиболее распространенных ежедневных изданиях (Prothom Alo, Bangla daily и Daily Star). Кроме того, было проведено пять углубленных интервью с журналистами, имеющими опыт освещения выборов, и одним сотрудником факультета журналистики с соответствующим академическим образованием. Количественные данные были проанализированы с помощью SPSS, а качественные - с помощью тематического анализа и дискурсивного анализа.

Ключевые слова: журналистика данных, дезинформация, фейковые новости, выборы в Бангладеш.

Introduction

We are living in an age of “datafication” [1, 2]. With the advancement of Information and Communication Technologies (ICTs), all aspects of human life, interactions and productions are going through massive and rapid changes, and are often being redefined; data are being used to describe almost everything in the world we are living in today [3], and the field of journalism is not an exception to this [4, 5]. ICTs intervention has transformed the profession by digitalizing the work process [6], which lead to the emergence of new genres of journalism, for example, multimedia journalism [7], mobile journalism [8], data journalism [3, 9], and computational journalism [10-12]. All these have added new dimensions to journalism and require the newsmen specialized ICT skills to practice these. In the several past years, as more information has been added to digital databases, use of quantitative data in journalism kept increasing which Petre [13] noted as “a quantitative turn” of journalism.

In recent years, data journalism attracted a lot of attention both in the digital news production sector and in academic discussions [14-18]. Data journalism basically developed in the past decade or so driven by abundance of digital data triggered by the rapid development of ICTs since early 1990s. This is a specialized area that reflects the rapidly increasing stake of statistical data in today's digital world in production and distribution of information [19]. Data journalism has been widely considered as the “future of journalism” [20]. And the future, noted by William Gibson as the “post-industrial future of journalism,” is already here (cited in [21]).

As an academic discipline, data journalism is still developing globally; it is relatively less explored as far as academic research is concerned. In Bangladesh even the practice of data journalism is quite nascent. The top- ranked media outlets are doing data-driven stories for the past few years; however, it is still not a common practice to see among journalists or media outlets as Islam [22] noted, “Data journalism is still uncommon in Bangladesh, so is computational journalistic field in the country. Data journalism is at an early stage. Bangladeshi media occasionally, now more frequently, does stories based on data.” This work focusing on analyzing the practice of data journalism during the Bangladesh National Election of 2018 would be one of the very first of its kind. The study focuses on achieving the following objectives:

To investigate the use of data by Bangladeshi dailies in terms of statistics and/or infographics in covering the National Election of 2018; and

To understand the role data journalism played in combating spread of misinformation during the said event.

It is worth mentioning here that Bangladesh (officially, People's Republic of Bangladesh), by its Constitution, is a democratic country in South Asia. The country, home for about 160 million people, earned its independence from Pakistan in 1971. It has a long history of colonial oppression. Once known for its poverty and natural calamities, Bangladesh's rapid economic growth in the past decade has made it known as the New Asian Tiger. In March 2018, the United Nations said that with the current economic growth the country is likely to come out of the category of Least Developed Countries (LDC) by 2024. Though the Constitution declared the country a democracy, it went under the rule of military dictators soon after the independence. In 1990, democracy was restored through a mass uprising. Since then the national election takes place once in every five years.

The 2018 National Election in Bangladesh was held on December 30, 2018. The Grand Alliance led by Awami League won the election; they won 258 out of 300 parliamentary seats. The election was criticised by many; the Bangladesh Nationalist Party (BNP), the second largest political party in the country, and its alliance boycotted the election.

Emergence of Data Journalism

Though only recently the term “data journalism” attracted attention globally and it is more of a 21st century construction, the concept is not new; Parasie and Dagiral [23] note that data journalism has been in practice since the beginning of the digitalization and US newspapers have been utilizing digital data in news production since the late 1960s. However, there is an even earlier example of data journalism. Bounegru, Chambers, and Gray [3] state mentioning Simon Rogers that the earliest example of data journalism is a Guardian report in 1821 which concerned the number of students who attended school and the costs per school in Manchester. According to Knight ([20] cited in [24]), the term “data journalism” is attributed to Twitter's first data editor Simon Rogers who mentioned this first in a Guardian Insider Blog post. However, according to Howard [25], Rogers [26] mentioned Adrian Holovaty's name from whom he heard the term first which Hirst [27] confirmed referring to Ingram [28].

Lorenz [29] views data journalism as a process that begins with analyzing and continues by filtering and visualizing data in a form that links to a narrative. Knight [20] defines it as “a story whose primary source or `peg' is numeric (rather than anecdotal), or a story which contains a substantial element of data or visualisation.” Rogers [30] notes that it combines spreadsheets, graphics data analysis and the biggest news stories. Deuze [31], looking at the interactive nature of online journalism as an added value, sees opening of new possibilities of user-content interaction in data journalism when done online. Veglis and Bratsas [32], addressing the power of visualisation and interactivity, present data journalism as “. . . the process of extracting useful information from data, writing articles based on the information and embedding visualisations (interacting in some cases) in the articles that help readers to understand the significant of the story or allow them to pinpoint data that relate to them.”

Bounegru, Chambers, and Gray [3] have rightly put that data can be used both as the source and tool for story-telling. Digitization of data and rapid advancement in ICTs, especially interactive web and web-based data analysis and visualisation tools, contributed to the emergence and growth of data journalism. Parasie and Dagiral [23] note that using large amounts of data for writing an article was a difficult task for journalists even at the end of the past century for it requires specialized skills which were beyond the capabilities of average journalists (which is still the reality for countries like Bangladesh as mentioned by Islam [22] in his article “Making Sense of Data Journalism from Bangladesh Perspective”. Some of the major media/news organizations in the USA and UK hired programmers for this job. Back then, journalists used to rely on information given by governments, officials, research studies, etc. Exceptionally, some investigative reports were produced wherein journalists could manage on their own to find resources for gathering and analyzing their own data. However, with the growing amount of digital data and introduction of smart tools, which anyone could use for analyzing, visualising, and publishing huge amount of data, the scenario rapidly changed [33, 34].

Data Journalism Framework

Multiple taxonomies of data journalism have been suggested by different academicians and professionals. Veglis and Bratsas [24] presented a summary shown in Table 1 that includes three such taxonomies created using different parameters:

Types of data journalism projects/stories proposed by different professionals/academics

Veglis and Bratsas [24] noted in this regard that visualisations supplement the narrative in most of the data journalism stories. However, “in some cases the visualisation is constructed in such a manner so as to in- elude the narrative of the news story. Thus, the visualisation becomes the centre of the data journalism story and the text (which is usually quite limited) supplements or explains the visualisation.” Alberto Cairo refers this as structuring the info-graphics as a story [35]. Based on their analysis of the previous taxonomies, Veglis and Bratsas [24] propose a new taxonomy which addresses both static and interactive data journalism projects. Figure 1 presents the graphical view of the taxonomy:

Figure 1. Taxonomy proposed by Veglis and Bratsas [24]

Methodology

The study is descriptive with a mixed method (quantitative and qualitative) design that includes content analysis of data-driven stories on the 2018

Bangladesh National Election and in-depth interviews. Two most widely reached dailies in Bangladesh, Prothom Alo (Bangla) and The Daily Star (English) have been selected. The Daily Star tops the circulation list of English dailies with a circulation of 44,814 [36]. Prothom Alo (501,800) comes second after the Bangladesh Pratidin (553,300) in the print circulation list [36]. However, it reaches a wider Bangladeshi community through its two online platforms (e-paper and news portal). Issues of the two above- mentioned dailies published between December 15, 2018 and January 14, 2019 (from 15 days before to 15 days after the election; the election was held on December 30, 2018) were analyzed. For ease of access, e-paper (online version with the exact same view of print version) of the two selected dailies (https://epaper.prothomalo.com/ and http://epaper.thedailystar.net) were used for content analysis. A coding sheet was developed and used for this purpose that followed the framework below which has been developed based on the taxonomy shown in Figure 2:

Figure 2. Conceptual framework adopted from Stalph [37] and Veglis & Bratsas [24]

Both the authors participated in the content analysis; hence, before start coding, they ensured inter-coder reliability of 90 percent using the following formula suggested by Holsti (1969): C.R = N1+N2 where “M is the number of coding decisions on which the two coders are in agreement, and N1 and N2 refer to the number of coding decisions made by coder 1 and 2, respectively” (p. 140). SPSS was used for analyzing the quantitative data gathered from content analysis.

Four journalists (two from print, one from television, and one from online media) were purposively selected for in-depth interviews following two criteria: (1) has experience of at least 10 years of active journalism; (2) covered the national election of 2018, and (3) has experience of doing data- driven stories. Alongside one journalism faculty member, who also works as a Mobile Journalism and Data Journalism trainer, was interviewed. The interviews were conducted face to face using a semi-structured interview schedule, they were recorded using digital audio recorder and smart phone (as a back-up). Afterwards, those were transcribed in a summarized form. A third person went through the transcriptions to ensure consistency.

Findings and Discussion

Attention and prominence given to election stories. In general, political news get good coverage in Bangladeshi media be it print, broadcast, or online. Election news get even better treatment. During the recently held National Election of 2018, election news in two major dailies, Prothom Alo and The Daily Star received good treatment as usual. However, the coverage of the election in terms of the number of news items published, Prothom Alo was far ahead. During the study period, 958 out of 3,606 news items (26.6%) in Prothom Alo were related to the election, which is over one-fourth (26.6%) of the total news. On the other hand, 9.6% of 3,417 news items published in The Daily Star related to the election. However, though the use of data was not that significant in terms of the number of data-driven stories published, The Daily Star was a bit ahead of Prothom Alo in this regard. Only about 3% (29 out of 958) of election stories in Prothom Alo were data- driven while this rate was 6% (19 out of 298) in The Daily Star.

Formal characteristics of data-driven election stories. In this section, the formal characteristics of the election stories that used data are discussed: frequency, prominence, number of authors, and topic of election stories.

Frequency and prominence of data-driven election stories. Figure 3 below presents the frequency and prominence of data-driven election stories before and after the election:

Figure 3. Frequency of data-driven election stories and Prominence given to them

It can be seen from Figure 3 that the use of data in election stories, in terms of the number of stories produced, was not that significant. Only 48 election stories were found in the two dailies analyzed. It is only 3.82% of the total number of election stories published during the study period. Also, it can be gleaned that Prothom Alo used data in more stories than The Daily Star. However, the low number of data-driven stories during the election is probably a reflection of the status of organizational readiness in Bangladeshi media outlets. The journalists interviewed in the study informed that there is neither a dedicated desk for data journalism in their outlets nor a comprehensive arrangement for training to equip reporters with skills. Those who are using data need to do everything by themselves including learning how to do it, practically the only help they get is from graphic designers who make visualisation part publicizable (Interview: Quarmal and Islam, 2019).

It can also be gleaned from the table above that Bangladeshi dailies treated data-driven election stories with high prominence (83%, or 40 out of 48). The Daily Star made it even more significant by putting all their stories in the high prominence area. These stories were given such a good treatment because usually readers have interest in election stories and (probably) visuals make them more suitable for such a treatment.

Number of authors in data-driven election stories. As can be seen in Figure 4 below, more than half (56%) of the data-driven election stories published in Prothom Alo and The Daily Star had no bylines; those were published with “Staff Reporter” or “Staff Correspondent” as the author. In stories with bylines, only about one-third (38%) were collaborative projects, a mere 16% of the total data-driven election stories. This is probably a reflection of a lack of strategic vision about seeing data as a primary source or “peg” [20]. Responses from the interviews regarding availability/access to data supports this assumption as well. One of the interviewees said: Access to data is sometimes difficult (even practically impossible) to get, however, generally Right to Information Act helps in getting data from government entities though sometimes the process is too lengthy, and we try to get the data from alternative sources in such cases. Also, a lot of publicly available data are out-dated and some are not well-structured and `realistic'. Still, there are a lot of publicly available data which can be source of good investigative reports. But, most of the journalists are not even aware of it let alone knowing how to use the data (Interview: Quar- mal and Islam, 2019).

Figure 4. Number of authors in data-driven election stories

Topics in data-driven election stories. Topics covered in data-driven election stories are presented in Figure 5 below.

As can be seen from Figure 5, about two-thirds (64.5 %) of the data- driven election stories were devoted to providing general election updates like information about constituencies, candidates, campaigns of political parties, clashes and conflicts among rival political groups, etc.

Figure 5. Topics in data-driven election stories

This trend can be seen both before and after the election; however, this is striking that all such stories after the election were devoted to providing general election updates though there are often mishaps following the election. Also, there is always a huge room for analyzing the election by comparing/contrasting it in many ways with the previous elections. It is also noteworthy here that all kinds of election stories were stopped about a week after the election. When asked during the interviews about this, none of the interviewees could shed light on this issue except saying “we (reporters) do not have control over this” (Interview: Quarmal and Islam, 2019).

Visualisations in data-driven election stories. Figures 6 and 7 below present the analysis of visuals, namely, the number of visualisations and types of visualisations used in data-driven election stories.

It can be seen from Figure 6 that a limited number of visuals were used in the data-driven election stories; three-fourths of the stories used just one visual each. The Daily Star used more visuals on average compared to Pro- thom Alo; seven out of their 19 stories used two visuals each and two stories used three visuals each.

Figure 6. Number of visualisations in data-driven election stories

Types of visualisations used in Prothom Alo and The Daily Star were not that diverse as well. As can be gleaned from Figure 7, mostly infographics (56%) were used as visualisation. Also, 12.5% of the stories used more than one visualisation each. However, though Roam's [38] visual frameworks say charts, for example bar charts, pie charts, area charts, line charts, radar charts, are more effective forms of visualisations for they can easily show comparative quantitative values and Kirk [39] mentions bar charts as the most reliable and useful visualisation as they can provide “categorical comparisons”, uses of charts were significantly low in numbers; only 12.5% in Prothom Alo. This is again, probably, the reflection of a lack of readiness among journalists, which is quite understandable as they have few training facilities.

Providing data-sets to readers is very important and cannot be ignored in any way as Stalph [37. P. 11] rightly reflected from Weinacht and Ralf [40] that “providing used data-sets to readers as defined by open data is an essential characteristic of data journalism.” Hence, providing just a textual reference is not enough; instead, providing a link to the source should be a general practice among data journalists so that readers can have direct access to data sets and can view/download them. However, Bangladeshi print media is not yet in that practice as can be seen from Figure 8 below, which presents the facts regarding provision of data in the data-driven stories during the Bangladesh National Election of 2018 in Prothom Alo and The Daily Star, nation's two most widely reached (and most credible) dailies. From their observation of print media in Bangladesh, the authors cannot assume any different scenario in other dailies or news industry as whole. However, this could be an exploration point for future studies.

Figure 7. Types of visualisations in data-driven election stories (Never used: timeline, pictogram, area chart, bubble chart, scatter plot, radar chart, flow chart)

Figure 8. Provision of data in election stories

Except for few infographics, source was mentioned in the visualisations. However, all visualisations were produced using data from single sources. Also, a few visualisations (7 of 48) did not mention any source in it. As for the number of sources in each story, most stories indicated only one source. Figure 9 below presents the facts.

Figure 9. Number of sources mentioned in each visualization in data-driven election stories

Figure 10. Providers of data used in election stories

It was found that government agencies, e.g. Election Commission (EC), Bangladesh Bureau of Statistics (BBS), etc., are major data providers of the election stories. This was learned from the interviews with the journalists as well. The journalists opine that none other can provide data as much as EC can. EC provides updates regarding election procedure on a regular basis. Also, EC makes open documents submitted by the candidates containing all kind of information about them including wealth, and these are a great source of data-driven stories. Alongside, local NGOs also do some research, and data-driven stories are done using their data. Figure 10 presents the scenario found in the content analysis.

Form and content of data-driven election stories. When looking at form and content, three aspects of data-driven election stories, namely format, reference of foreign news, and juxtaposition, were analyzed. As can be gleaned from Figure 11 below, almost all the stories were presented in combination of text and visuals. In 71% of the stories, visualisation was part of the story; on the other hand, visualisation was structured as story in 25% of the stories (12 of 48). However, no visualisation was used in two stories--both published in The Daily Star. Though visualisation is often thought to be an integral part of data-driven stories, such absence of visualisation does not impact the story as Weinacht and Spiller [40] argue that data-driven stories are not bound to graphically represent the data. Also, the interviewed journalist from Prothom Alo opined the same. He stated: “We try to use visualisations as that can provide the reader with ease of summarizing the main facts and information regarding the subject matter, and it helps to attract the reader as well. However, content is the key, not the visualisations” (Interview: Quarmal and Islam, 2019).

Figure 11. Structure of the data-driven election stories (Two stories did not use any data, hence were not categorically presented in the chart)

It is also seen in the analysis that none but one story, published in Prothom Alo, did refer to foreign news. This is quite understandable that the national election is a domestic event, hence, not much related to issues external to the country.

The researchers also tried to see whether there is any “perceived recognizable conflict” [37] by examining if any contrasting juxtaposition of “protagonists” and “antagonists” was put into action. No such juxtaposition was found.

Role of data journalism in combating misinformation during election. As mentioned earlier, four journalists having wide experience of covering election and doing data-driven stories were interviewed; among them two are from print media, one from television, and one from online media. Among the two working in print, one was from Prothom Alo and the other from The Daily Star. The journalist from Prothom Alo not only covered a number of elections in Bangladesh (including National Election, Upozilla Election, City Corporation Election, etc.) but also covered the recent National Election in India. Both the interviewees agreed on the importance of data in telling stories, in establishing truth. The one from Prothom Alo said: “Readers want quality news, quality story. Numbers do mater in quality storytelling, to tell the truth, to interpret the truth, to describe the social, political or economic process . . . Number is a determinant of a good news story, whether we put it in the headline or in story, as part of texts only or in the form of charts, tables or infographics.”

The journalist from The Daily Star said: “The practice of data journalism in Bangladesh is at very early level. Both print and electronic media are trying to tell stories by using data. We at the Daily Star are trying to practice data journalism for past few years. We try to use data in story and present visually. To make them more communicative, we emphasize on info-graphics so that stories are interactive and readers feel more connected to the story or issue.”

The other two interviewed journalists did not say anything much different. The TV journalist said: “We often use data in the form of infographics in our reports. However, if you ask me - it's not really `data journalism' in proper sense; not much insights are drawn from the infographics shown on the screen, sometimes they seem to me just decorative pieces. And, why not? Most of us don't have much idea of `data journalism'; the concept is new.”

The one from online media noted: “I have been working as a journalist for about 20 years. Data is not new to me. I am familiar with using data in news stories. Because, I deal with the issues related to economy, business, and finance. And, I completely agree that there is no better way to tell stories to convince your audience, to show them the truth; the data itself tell the story. But as a concept data-driven journalism or data journalism is very new in our country. Most of colleagues are unaware of the workflow that this specialized domain requires, let alone the skills.”

All the interviewees mentioned data provided by EC, specially the affidavits submitted by candidates to EC, as the main source of data of data- driven election stories. The Prothom Alo journalist stated: “We tried to use the information by putting in context, by digging out their embedded meaning. Data can help in challenging the official claims which are not true. We made stories by using the data on wealth of candidates from their `HOLOFNAMA' (affidavits) and showed the comparison between official claim and reality” (Interview: Quarmal and Islam, 2019).

The Daily Star journalist made similar point stating: “We used all `HOLOFNAMA' (affidavits). We analyzed them, tried to crunch data and presented using graphs, tables, charts their wealth before the previous election and before the recent one.”

Regarding the role of data journalism in combating misinformation during the elections, all the interviewees made positive notes that data journalism can do this not only in the case of election but other issues too. However, they stressed the credibility of the media outlet doing data-driven stories as well as the credibility of the data provider mentioning that people in Bangladesh often do not tend to rely on data provided by the government entities; they perceive them to be “cooked”. However, credibility of the media outlet adds a layer of trust: “That's what our newspaper has earned in these many years,” said the journalist from Prothom Also; “people think if Prothom Alo is saying something then there must be some point,” he added. The journalist from The Daily Star stated: “. . . we used data against false claims during Shahbag movement, Hafejat movement, Ramu incident. The blending of data and word can be really a great tool to tell the truth.” However, none of the interviewees reflected much on the electoral issues except for those already mentioned about using data from HOLOFNAMA.

As mentioned earlier, the interviewees also reflected on the current state of data journalism in Bangladesh. They agreed that data journalism is in its early days in Bangladesh. Few media outlets are trying to practice it but there are still a lot of challenges to overcome among which the first is the readiness of journalists and media outlets. Journalists need training and the organizations need to provide facilities required for data journalism, for example, a dedicated desk that would work in gathering/mining, analyzing, interpreting, and visualizing data and regular training programs for journalists to make them skilled in these processes. Another important issue they mentioned was data archiving. According to them, data archiving is poor in Bangladesh, especially in case of government entities; a lot of publicly available data are outdated, and, worse, sometimes the respective office does not have the data. The journalist from Prothom Alo mentioned one such incident off the record. The online journalist mentioned the usability of the available data. According to him, most of the publicly available data are not readily usable, they are not structured. Both the authors experienced this themselves by exploring a number of websites. The interviewees also mentioned lack of training facilities on data journalism--both internal and external. They could not mention any institute that provides training on data journalism.

To get more insights on the current status of data journalism in the country, a journalism faculty member having a PhD in Journalism from the University of Liberal Arts Bangladesh was interviewed. He also works as a Mobile Journalism and Data Journalism trainer. He has a wide range of experience of working as a multimedia journalist and consultant at home and abroad. He echoes the journalist interviewees regarding the current status of data journalism in the country that it is in its early days in the country. He said:

I have been extensively working as a trainer and consultant before joining here. From my observations, I can say that there is lack of readiness both at the organizational level and individual level. I did not come across any media outlet that has a dedicated section for dealing with data journalism workflow. Dealing with data is a specialized skill and most of the journalists does not have the training for it. In my knowledge, the very first trainings on data journalism started taking place only in 2019. DW Akademie organized these trainings. The journalism schools are not providing any training on this as well. Most of the journalism curricula in the country focus on the traditional skills like news writing, news sourcing and gathering etc. which also came out in a study commissioned by the DW Akademie. The Media Studies and Journalism Department that I work in is one of the finest media, communication and journalism schools in the country; only of its kind to receive international recognitions. We have recently introduced a data journalism major - the first of its kind in the country. I think this information alone tells the story of data journalism in this country.

Conclusion

The sampled widely reached and largest circulated dailies produced a very small number of data-driven stories during the 2018 election. However, though the number of data-driven election stories was not that significant in both the dailies analyzed, those received good treatment in terms of prominence. Data visualisations were found to be less diverse--mostly infographics which reflects the lack of readiness again.

The interviewees agreed that data-driven stories can play an important role in combating misinformation; however, this was not much visible in the stories on the Bangladesh National Election of 2018. These stories did combat some misinformation and disinformation by debunking false claims made by the candidates; however, the efforts were limited. The interviews also revealed that the journalists rely mainly on the government entities for data, including during elections. However, data archiving and access to data can be seen as a significant challenge for the practice of data journalism in Bangladesh at this point of time.

The practice of data journalism is still in an early stage in Bangladesh. The media outlets in the country are not yet equipped with appropriate facilities and manpower, and they do not have mechanisms to produce/develop such manpower. From the interviews, it was also surfaced that there are almost no training facilities on data journalism; the curricula in journalism schools in the country are mainly focused on the traditional journalism skills like news writing, news sourcing and gathering, etc.; the first academic program on data journalism in the country has been introduced only recently.

The main limitation of the study is the sample size. The study included only two top reached news media outlets of the country where thousands of newspapers and magazines are published regularly. Moreover, it used only Dhaka-based newspapers and conducted interviews with Dhaka-based journalists.

The authors suggest for more researches on the topic with a larger sample size to get a broader perspective of the issue, to have deeper insights of the phenomenon, and to draw an authoritative conclusion.

References

1. Arsenault, A. (2017) The datafication of media: Big data and the media industries. International Journal of Media & Cultural Politics. 13. pp. 7-24.

2. Lycett, M. (2013) `Datafication': making sense of (big) data in a complex world. European Journal of Information Systems. 22. pp. 381-386.

3. Bounegru, L., Chambers, L. & Gray, J. (2012) The Data Journalism Handbook. [Online] Available from: http://datajournalismhandbook.org/.

4. Loosen, W. (2018) Four Forms of Datafied Journalism: Journalism's Response to the Datafication of Society. Bremen: “Communicative Figurations” Research Network, Centre for Media, Communication and Information Research (ZeMKI) (Working Paper Series “Communicative Figurations Working Papers” No. 18).

5. Baack, S. (2015). Datafication and empowerment: How the open data movement re-articulates notions of democracy, participation, and journalism. Big Data & Society. 2 (2). p. 205395171559463.

6. Veglis, A. (2009) Cross Media Communication in newspaper organizations. MCIS 2009 Proceedings. 4th Mediterranean Conference on Information Systems. Athens Greece: MCIS.

7. Bull, A. (2010) Multimedia Journalism: A Practical Guide. London and New York: Routledge.

8. Martyn, P.H. (2009) The Mojo in the Third Millennium. Journalism Practice. 3 (2). pp. 196-215. DOI: 10.1080/17512780802681264.

9. Felle, T., Mair, J. & Radcliffe, D. (eds) (2015) Data Journalism: Inside the Global Future. Suffolk: Abramis.

10. Diakopoulos, N. (2011) A Functional Roadmap for Innovation in Computational Journalism. April 22. [Online] Available from: http://www.nick-diakopoulos.com/2011/04/22/a-functional-roadmap-for-innovation-in-computational- journalism/.

11. Gynnild, A. (2014) Journalism Innovation Leads to Innovation Journalism: The Impact of Computational Exploration on Changing Mindsets. Journalism. 15. pp. 713730. DOI: 10.1177/1464884913486393.

12. Hamilton, J.T. & Turner, F. (2009) Accountability through Algorithm: Developing the Field of Computational Journalism. Report given at the Behavioral Sciences Summer Workshop, Stanford, CA, July 27-31. [Online] Available from: http://www.stanford.edu/~fturner/Hamilton%20Turner%20Acc%20by%20Alg%20Fina l.pdf.

13. Petre, C. (2013) A Quantitative Turn in Journalism? Tow Center for Digital Journalism, October 30. [Online] Available from: http://towcenter.org/blog/a- quantitative-turn-in-j ournalism/.

14. Appelgrena, E. & Nygren, G. (2014) Data Journalism in Sweden Introducing new methods and genres of journalism into “old” organizations. Digital Journalism. 2 (3). pp. 394-405. DOI:10.1080/21670811.2014.884344.

15. Bradshaw, P. (2011) The inverted pyramid of data journalism. Online Journalism Blog. [Online] Available from: http://onlinejournalismblog.com/2011/07/07/the-inverted-pyramidof-data-journalism.

16. Bradshaw, P. (2011) Data journalism survey: A mixed picture. Data Driven Journalism. [Online] Available from: http://datadrivenjournalism.net/news_and_analysis/data_journalism_survey_analysis.

17. Abramis, F.K. & Anderson, C. (2015) Data Journalism in the United States. Journalism Studies. 16 (4). pp. 467-481.

18. Mair, J. & Keeble, R.L. (2014) Data Journalism: Mapping the future, Suffolk: Abramis.

19. Thibodeaux, T. (2011) 5 Tips for Getting Started in Data Journalism. Poynter.

20. October 6. [Online] Available from: http://www.poynter.org/how-tos/digital-strategies/147734/5-tips-for-getting-started-in-data-j ournalism/.

21. Knight, M. (2015) Data Journalism in the UK: A Preliminary Analysis of Form and Content. Journal of Media Practice. 16 (1). pp. 55-72. DOI:10.1080/ 14682753.2015.1015801.

22. Andersen, C., Bell, E. & Shirky, C. (2014) Post Industrial Journalism: Adapting to the Present. In: Howard, A.B. (2014) The Art and Science of Data-Driven Journalism. Tow Center for Digital Journalism, Columbia University. [Online] Available from: https://academiccommons.columbia.edu/doi/10.7916/D8Q531V1. DOI: 10.7916/D8Q531V1.

23. Islam, M.A. (2018) Making Sense of Data Journalism from Bangladesh Perspective. Global Media Journal. 16 (31).

24. Parasie, S. & Dagiral, E. (2013) Data-driven journalism and the public good: “Computer-assisted-reporters” and “programmer-journalists” in Chicago. New Media & Society. 15 (6). pp. 1-19. DOI:10.1177/1461444812463345.

25. Veglis, A. & Bratsas, C. (2017) Towards a taxonomy of data journalism. Journal of Media Critiques. 3 (11). pp. 109-121.

26. Howard, A.B. (2014) The Art and Science of Data-Driven Journalism. Tow Center for Digital Journalism, Columbia University. [Online] Available from: https://academiccommons.columbia.edu/doi/10.7916/D8Q531V1. DOI:10.7916/D8Q531V1

27. Rogers, S. (2012) Facts are sacred: The power of data. Guardian. 6 Jan. 2012. [Online] Available from: http://www.theguardian.com/news/datablog/2012/jan/06/ facts-sacred-guardian-shorts-ebook.

28. Hirst, T. (2014) Personal Recollections of the "Data Journalism" Phrase. OUsefulInfo. 29 Apr. 2014. [Online] Available from: http://blog.ouseful.info/ 2014/04/29/personal-recollections-of-the-data-journalism-phrase/.

29. Ingram, M. (2009) The Golden Age of Data Journalism? Nieman Journalism Lab. May 2009. [Online] Available from: http://www.niemanlab.org/2009/05/the- golden-age-of-computer-assisted-reporting-is-at-hand/.

30. Lorenz, M. (2010) Data driven journalism: What is there to learn? Presented at IJ-7 Innovation Journalism Conference, 7-9 June, Stanford, CA.

31. Rogers, S. (2008) Turning official figures into understandable graphics, at the press of a button. Inside the Guardian Blog. [Online] Available from: http://www.Guardian.co.uk/ help/insideGuardian/2008/dec/18/unemploymentdata.

32. Deuze, M. (2001) Online journalism: Modelling the first generation of news media on the World Wide Web. First Monday. 6 (10). [Online] Available from: http://firstmonday. org/oj s/index. php/ fm/ article/ view/893.

33. Veglis, A. & Bratsas, C. (2017a) Journalists in the age of Data Journalism: The case of Greece. Journal of Applied Journalism & Media Studies. 6 (2). pp. 225-244.

34. Bradshaw, P. & Rohumaa, L. (2011) The Online Journalism Handbook: Skills to Survive and Thrive in the Digital Age. Harlow: Longman.

35. Sirkkunen, E. (ed.) (2011) Trends in Data Journalism. [Online] Available from: http://virtual .vtt.fi/ virtual/nextmedia/Deliverables-2011/D3.2.1.2.B_Hyperlocal_Trends_In%20Data_Journalism.pdf.

36. Bradshaw, P. et al. (2015) Data Driven Journalism - Doing Journalism with Data: First Steps, Skills and Tools. LEARNO.NET Professional Video Training by the European Journalism Centre. [Online] Available from: http://learno.net/courses/doing- journalism-with-data-first-steps-skills-and-tools.

37. DFP. (2018) Statistics of Registered Newspapers and Magazines. [Online]

38. Available from: https://dfp.portal.gov.bd/sites/default/files/files/dfp.portal.gov.bd/page/b5b99e8e_85fe_465d_992c_1fee17302efc/-%E0%A6%AA%E0%A6%A4%E0%A7%8D%E0%A6%B0%E0%A6%BF%E0%A6%95%E0%A6%BE%E0%A6%B0%20%E0%A6%AA%E0%A7%8D%E0%A6%B0%E0%A6%9A%E0%A6%BE%E0%A6%B0%20%E0%A6%B8%E0%A6%82%E0%A6%96%E0%A7%8D%E0%A6%AF%E0%A6%BE%20%E0%A6%93%20%E0%A6%AC%E0%A6%BF%E0%A6%9C%E0%A7%8D%E0%A6%9E%E0%A6%BE%E0%A6%AA%E0%A6%A8%20%E0%A6%B9%E0%A6%BE%E0%A6%B0%20(7).pdf.

39. Stalph, F. (2017) Classifying Data Journalism. Journalism Practice. 12 (10). pp. 1332-1350. DOI:10.1080/17512786.2017.1386583.

40. Roam, D. (2009) The Back of the Napkin: Solving Problems and Selling Ideas with Pictures. Expanded ed. New York: Penguin Books.

41. Kirk, A. (2016) Data Visualisation: A Handbook for Data Driven Design. London: Sage.

42. Weinacht, S. & Spiller, R. (2014) Daten journalismus in Deutschland. Eine explorative Untersuchungzu Rollenbildern von Daten journalisten. Publizistik. 59 (4). pp. 411-433.

Размещено на Allbest.ru