A Comprehensive Analysis of the Importance of Programming in
Modern Data Journalism
Un análisis exhaustivo de la importancia de la programación en el periodismo de
datos moderno
doxa.comunicación |
nº 43, pp. 17-34
| 17
July-December of 2026
ISSN: 1696-019X / e-ISSN: 2386-3978
How to cite this article:
Veglis, A. (2026). A Comprehensive Analysis of the Importance of Programming in Modern Data Journalism.
Doxa Comunicación
, 43, pp. 17-34.
https://doi.org/10.31921/doxacom.2878
Andreas Veglis
is a professor of media technology, and head of the Media Informatics Lab at the School of Journalism
and Mass Communication at Aristotle University of essaloniki. He is serving and has served as an editor, member of
scientic boards, and reviewer in various academic journals. Prof. Veglis has more than 200 peer-reviewed papers on
media technology and journalism. Specically, he is the author or co-author of 12 books and 44 book chapters, he has
published more than 115 papers in scientic journals, and he has presented 154 papers at international and national
Conferences. Prof Veglis has been involved in 50 national and international research projects. His research interests
include information technology in journalism, new media, algorithmic journalism, in journalism, digital security, data
journalism, big data, and content verication.
Aristotle University of Thessaloniki, Greece
[email protected]
ORCID:
0000-0002-0286-2304
is content is published under Creative Commons Attribution Non-Commercial License. International License CC BY-NC 4.0
Recibido: 31/12/2024 - Aceptado: 17/02/2025 - En edición: 26/02/2026 - Publicado: 01/07/2026
Resumen:
En los últimos 20 años, se ha experimentado un desarrollo signicativo
en el campo del periodismo de datos, que es una especialización basa
-
da en la búsqueda de noticias en los datos. Hoy en día, existe una ten
-
dencia en el periodismo de datos a emplear técnicas de programación
(normalmente Python o R). Esta tendencia proviene principalmente de
importantes organizaciones periodísticas que participan en proyectos
de periodismo de los macrodatos o big data basados en el análisis de da
-
tos. El objetivo principal de este estudio es examinar la importancia de
la programación en el periodismo de datos actual. Se emplea una me
-
todología mixta, que incluye una revisión bibliográca y un análisis de
estudios de caso. Los hallazgos indican que, si bien la codicación no es
necesaria en el periodismo de datos, puede considerarse una necesidad
Received: 31/12/2024 - Accepted: 17/02/2025 - Early access: 26/02/2026 - Published: 01/07/2026
Abstract:
In the last 20 years, the eld of data journalism has seen signicant
development. Data journalism is a specialization based on nding
news stories in data. Today, there is a trend in data journalism to use
programming techniques (usually Python or R). is trend comes
mainly from prominent journalism organizations engaged in big data
journalism projects based on data analysis. e main thrust of this
study is to examine the importance of programming in data journalism
today. A mixed methodology is employed, including a literature review
and case study analysis. e ndings indicate that although coding is
not necessary in data journalism, it may be viewed as a necessity when
used for complex projects that include large datasets and specic
types of visualization. e study also discusses hybrid approaches and
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
18 |
nº 43, pp. 17-34
|
July-December of 2026
A Comprehensive Analysis of the Importance of Programming in Modern Data Journalism
ISSN: 1696-019X / e-ISSN: 2386-3978
doxa.comunicación
cuando se utiliza para proyectos complejos que incluyen grandes con
-
juntos de datos y tipos especícos de visualización. El estudio también
analiza enfoques híbridos y herramientas especícas que tienen como
objetivo cerrar la brecha entre los enfoques basados en herramientas y
los basados en código en la práctica del periodismo de datos. Los resul
-
tados indican que la programación debería considerarse una habilidad
complementaria que los periodistas de datos deberían considerar.
Palabras clave:
Periodismo de datos, programación, codicación, Python, R,
metodología de estudio de caso.
specic tools that aim to bridge the gap between tool-based and code-
based approaches in practicing data journalism. e results indicate
that programming should be viewed as a complementary skill that
should be considered by data journalists.
Keywords:
Data journalism, programming, coding, Python, R, case study
methodology.
1. Introduction
Since the conception of journalism, its relationship with technology has been symbiotic, with each shaping and inuencing the
other (Siapera and Veglis, 2012). Recently, this relationship has taken a transformative turn with the move towards datacation.
e concept of datacation was introduced by Mayer-Schönberger and Cukier (2013), and it was dened as the process of
transforming large quantities of information into a scalable resource in order to be used as knowledge production and thus
add economic value. e introduction of datacation in journalism can be seen as a signicant shift because it fundamentally
alters the process of news production, dissemination, and consumption. It is worth mentioning that journalism has already
undergone a transformation caused by the introduction of big data technologies (Coddington, 2015). is has fueled the
development of data journalism, a journalism specialty that employs computational methods to analyze large datasets and
extract meaningful information (Gray, Bounegru, & Chambers, 2012; Veglis and Bratsas, 2021).
Data journalism has faced skepticism from journalists who are resistant to technological changes. However, with the rise of
digitalization, data started to inuence newsrooms through metrics, resource allocation, topic selection, and news format, and
thus gained more attention in scientic publications (Angelou
et al.
, 2020). Various terms, such as “data-driven journalism,”
“interactive journalism,” and “big data journalism” has been employed to depict this new form of journalism (Ausserhofer
et
al.
, 2020). us, datacation has gradually become an important parameter in the production of journalistic content, with data
journalism playing a signicant role in this process (Porlezza, 2023).
e journalist’s datacation of their toolkit has provided them with the ability to report complex issues and provide empirically
grounded narratives (Lewis & Westlund, 2015). It has also democratized the eld, allowing for greater public engagement
through interactive visualizations and open data platforms (Young
et al.
, 2018). is transformation has resulted in the
redenition of the journalist’s role, which now includes tasks such as data analyst, visual designer, and interpreter of complex
socio-political phenomena (Ausserhofer
et al.
, 2020). e renovated journalism space is a new kind of journalism in the shape
of altered journalistic practices through the growing incorporation of data and programming. Ruppert
et al.
(2017) argue that
data has turned into a tool of power, thus, making such inuence felt in the public and private sectors, even in the domain of
journalism. ey claim that the arrival of data journalism is part of the general reconguration of power structures, with data,
doxa.comunicación |
nº 43, pp. 17-34
July-December of 2026
Andreas Veglis
ISSN: 1696-019X / e-ISSN: 2386-3978
| 19
code, and algorithms as the most important actors. On the other hand, data journalism was practiced before the introduction
of coding and was mainly focused on statistical analysis through computer-assisted reporting (CAR) (Parasie & Dagiral, 2013).
e necessity of data journalism coding is often debated within the media industry. Proponents on one side say that coding
skills are important for journalists (and especially data journalists) in the current media landscape. ey argue that coding
enables journalists to collect, analyze, and present data in new and creative ways, thus enabling them to detect hidden
patterns and provide a thorough explanation of complicated issues. In addition, they state that the usefulness of coding for
storytelling and investigative journalism (Simon, 2021) has increased. Porlezza (2023) argues that coding is a prerequisite for
the practice of data journalism. Data journalism involves the collection, analysis, and visualization of data to tell stories and
disclose insights. To deal with information eectively, journalists usually utilize programming languages (such as Python and
R) to clean data, conduct statistical analyses, and create data visualizations. Coding skills enable journalists to use big data,
data processing automation, and interactive data visualizations. However, it must be said that the integration of coding with
journalism practice also raises some concerns. In particular, critics insist that journalists must keep their focus on storytelling
and information-gathering skills and leave the technical part to data experts and developers. e fear is that mastering coding
may cause journalists’ attention to be diverted from their main duties, and thus, journalistic integrity may be compromised
(Cruz, n.d.; Hannaford, 2015). Heravi and Lorenz (2020) argue that the use of coding in data journalism is a matter of debate.
Even if coding talents may be useful within the framework of data-driven stories, it should be noted that not all data journalists
are obliged to be coding specialists. Porlezza (2023) does not exactly oppose that conclusion, but he states that the knowledge
of coding may strengthen a journalist’s capacity to deal with data properly. Coding is not considered a core component of data
journalism but rather a tool that can meaningfully assist in the process (Heravi & Lorenz, 2020).
Finding the equilibrium point between the pros and cons of coding in journalism is crucial. Journalists endowed with
programming skills can mine vivid data and attract audiences using various new techniques; however, it is important that
journalistic standards and ethics are not compromised. Colman
et al.
(2018) explain the ethical issues that arise when coding
is applied by journalists in their work.
us, the use of coding in journalism presents both opportunities and challenges. It is essential for journalists to critically
evaluate the benets and risks associated with coding, considering the specic context and goals of their reports. By leveraging
coding skills responsibly and ethically, journalists can harness the power of technology to enhance their storytelling and
deliver impactful journalism.
e main thrust of this study is to examine the signicance of programming in data journalism practice today. Based on the
previous discussion, two research questions were formulated. Specically, the disagreement over whether programming is
essential or supplementary in modern data journalism is the basis for the rst research question:
RQ1: Is coding considered a necessity for practicing data journalism?
Additionally, the discussion in cases where programming is utilized points to the idea that the characteristics of a specic
project likely determine the need for code-based approaches, thus helping us articulate the second research question:
RQ2: What are the features of a data journalism project that inuence the decision to utilize coding for it?
20 |
nº 43, pp. 17-34
|
July-December of 2026
A Comprehensive Analysis of the Importance of Programming in Modern Data Journalism
ISSN: 1696-019X / e-ISSN: 2386-3978
doxa.comunicación
1.1. Practicing data journalism
Simon Rogers is credited with rst using the term data journalism in a post on the Guardian Insider Blog, as noted by Knight
(2015). is concept encompasses a workow that begins with data analysis, followed by ltering and visualizing the data to
complement the narrative, as outlined by Lorenz (2010). It integrates aspects such as spreadsheets, graphics, data analysis,
and major news stories (Rogers 2008) and is essentially about creating news graphics, incorporating design and interactivity
elements (Bradshaw 2018; Lorenz 2010; Rogers 2008). Megan Knight (2015) denes data journalism as a narrative primarily
based on numerical data or one that signicantly involves data or visualization.
Veglis and Bratsas (2017a) oered a denition that more eectively encapsulates the importance of visualization and
interactivity in data journalism. ey describe it as a process of extracting valuable insights from data, crafting articles based
on this information, and integrating visualizations (sometimes interactive) into these articles to aid reader comprehension or
allow them to engage with data relevant to them. ey also proposed the categorization of the data journalism process into
six distinct stages: Data Compilation, Data Cleaning, Data Understanding, Data Validation, Data Visualization, and Article
Writing (2017b).
Data Compilation
: is initial phase of a data journalism project begins with either a question requiring data or a dataset
that needs to be explored. Data compilation can occur in various ways, such as receiving data directly from an organization
(often as open data), using advanced search techniques (mainly with the help of Google advanced search), searching with
search engines that focus on datasets (for example, Data Search tool), web scraping (with dedicated scraping tools or through
widely used applications that support web scraping), converting document formats to enable analysis (for example, extracting
tables from PDF les and exporting them to CSV format), or gathering data through observation, surveys, online forms, or
crowdsourcing (Veglis & Bratsas, 2017a).
Data Cleaning
: Also known as data scrubbing, involves detecting and correcting erroneous or corrupted records within a
dataset (Wu, 2013). It involves removing human errors and standardizing data formats for consistency with other data used
by journalists (Veglis & Bratsas, 2017b). For this task, general-purpose applications such as Microsoft Excel and Google Sheets
can be utilized, which include many features, some of which can be used for data cleaning directly or indirectly (Bauzon
et al.
,
2021; Guerrero
et al.
, 2019; Setiyanto & Setiawan, 2022).
Data Understanding
: In this stage, journalists must decipher various codes in datasets representing categories, classications,
or locations, along with specialized terminology. Additional data are often required to make the existing data meaningful.
Journalists need to be data literate and capable of understanding, articulating, and critically analyzing data (Veglis & Bratsas
2017a). For this purpose, it may be necessary to nd additional data; therefore, the tools of the previous stage may be employed.
In addition, tools for combining datasets may be useful. Such a process can be accomplished more easily with applications
that support easy data combination, such as Tableau (Balaji,
et al.
, 2021; Batt,
et al.
, 2020; Loth, 2019). It is worth noting that
dataset combination can be accomplished with the help of general-purpose spreadsheet applications such as Microsoft
Excel or Google Sheets; however, it is not an easy process because these applications do not directly support such processes
(Guerrero, Guerrero, & Rauscher, 2019).
doxa.comunicación |
nº 43, pp. 17-34
July-December of 2026
Andreas Veglis
ISSN: 1696-019X / e-ISSN: 2386-3978
| 21
Data Validation
: is involves cross-checking original data and acquiring additional information from sources to enrich
the data (Silverman, 2014; Veglis, 2013). It is important to recognize that datasets, like any source, have inherent biases and
objectives. Journalists must investigate the origin, purpose, and collection methodology of a dataset (Bradshaw, 2018). is
can be done by exploring the dataset’s creation history, nding dataset references, or using other information sources related
to the investigative subject (Silverman, 2014; Veglis & Bratsas, 2017a). us, general-purpose searching techniques, such as
those employed in the data compilation stage, may be utilized.
Data Visualization
: It is the visual representation of abstract information for analysis and communication (Cairo, 2012).
Because statistical information is abstract, transforming it into a physical representation requires an understanding of visual
perception and cognition. Eective data visualization follows design principles based on human perceptions (Card
et al.
,
1999; Few, 2013). Several applications (desktop or online services) can be used to create visualizations (Protopsaltis
et al.
,
2020). Most of them oer multiple templates for creating dierent types of visualizations. Typical examples include Infogram,
Piktochart, DataWrapper, and Tableau.
1.2. Available solutions for adopting coding.
In the media industry, there are two alternative solutions for utilizing coding: the programming languages Python and R (and in
some cases Ruby) (Parasie & Dagiral, 2013). Both oer robust capabilities for data processing and visualization. Python, which
is characterized by its simplicity and readability, can be used for tasks such as web scraping, data analysis, and interactive
visualization. It oers extensive libraries, such as Pandas and Matplotlib, that enable journalists to eciently handle large
datasets and present complex information in an accessible format. Python supports integration with web frameworks, a
feature that supports the creation of interactive and online presentations (Lutz, 2001).
e R programming language is highly regarded for its statistical analysis and data visualization capabilities, making it an ideal
choice for in-depth analysis of complex datasets. It oers many packages that can support the discovery of hidden patterns and
trends in the data. R Markdown oers a platform for combining code, visualizations, and narratives to enhance storytelling in
data journalism (Crawley, 2012).
Both Python and R are open-source, which fosters a collaborative environment, with a vast community of developers
contributing to their development and oering support. is aspect is crucial for journalists, who often rely on shared
resources and community support to learn and troubleshoot. Additionally, there is an abundance of educational resources
and online tutorials for both programming languages, which facilitates their adoption by journalists (Appelgren & Nygren,
2014; Bradshaw, 2018; Bounegru
et al.
, 2018; Hamilton, 2016; Knight, 2015).
Setting up Python or R language on a computer is not an easy process. Nevertheless, there are alternative solutions that utilize
cloud computing, such as Google Colab, which supports writing and executing Python code through a browser (Tock, 2019).
Specically, Google Colab is a free, cloud-based Jupyter Notebook environment that enables writing and executing Python in a
web browser with free access to GPUs and TPUs, making it suitable for machine learning and data science (Burke, 2023). us,
journalists do not need to set up a personal computer to use Python.
22 |
nº 43, pp. 17-34
|
July-December of 2026
A Comprehensive Analysis of the Importance of Programming in Modern Data Journalism
ISSN: 1696-019X / e-ISSN: 2386-3978
doxa.comunicación
Another recent development is the integration of Excel with Python. Microsoft combines Python’s advanced computational
abilities with Excel’s spreadsheet features (Kinnestand, 2023). Users can directly employ Python in Excel cells by leveraging
Python libraries for data manipulation and visualization
1
. is integration in Excel allows journalists to gradually migrate to
Python coding while using the familiar Excel environment.
Another approach that is available for working with Python involves utilizing AI interfaces like ChatGPT, with the help of
specic plugins
2
. Of course, it is worth noting that such solutions are still at an early stage of development, with plugins
appearing and disappearing rapidly. With the help of specic plugins, ChatGPT can serve as an intuitive interface with its
natural language processing capabilities for journalists who may not possess much programming knowledge and are not eager
or have the time to learn programming. is could involve querying datasets, generating statistical summaries, or creating data
visualizations using text commands. us, the utilization of Python for data journalism becomes easier, allowing journalists
to focus more on the narrative and less on the technical aspects of data handling (Knight, 2015). e use of ChatGPT as an
interface with Python can also be viewed as a democratization of data journalism, allowing a broader range of journalists to
engage in data-driven storytelling. is approach aligns with the trend towards more accessible and collaborative forms of
journalism (Hamilton, 2016). By lowering the barrier to entry for data journalism, AI plays a signicant role in the evolution of
the eld, making it more inclusive and diverse than before.
1.3. Tool based vs code-based data journalism
Based on the previous discussion, a direct comparison between tool-based and code-based approaches to practicing data
journalism is conducted. e comparison is based on the possibilities oered by each approach in facilitating the various
stages of developing data journalism, as depicted by Veglis and Bratsas (2017b) and presented in Section 1.1. To facilitate
such a comparison, Table 1 was constructed. Table 1 includes indicative tools (services and apps) that can be used in the ve
data journalism stages in the tool-based approach, along with the available capabilities that the two mentioned programming
languages (Python and R) oer directly or with the help of specic libraries or packages.
1
https://www.microsoft.com/en-us/microsoft-365/python-in-excel
2
https://openai.com/index/chatgpt-plugins/
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
doxa.comunicación |
nº 43, pp. 17-34
July-December of 2026
Andreas Veglis
ISSN: 1696-019X / e-ISSN: 2386-3978
| 23
Table 1. Direct comparison of the possibilities oered by the tool-based and code-based approaches.
Data Journalism StageTools utilizedCoding possibilities (Python / R)
Data Compilation
Search engines
: Google, Bing, etc.
Data Search tool
: Google Data Search
Web scraping
: Microsoft Excel, Google Sheet.
PDF scraping
: Tabula
Conduct web searches in Python through the
Google Custom Search API and in R with the
help of GoogleSearchR package.
Web scraping is supported by both Python (i.e
BeautifulSup4), and R (i.e rvest).
Data Cleaning
General purpose application
: Microsoft Excel,
Google Sheet.
Dedicated cleaning tool
: OpenRene
Data cleaning can be implemented by both
Python and R since both programming
languages the support data cleaning operations
Data Understanding
Applications/tools that support combining data
sets:
Tableau, Microsoft Excel, Google Sheet.
Python supports the combination of datasets,
primarily using pandas library. e same is
valid for R.
Data ValidationNo additional application/tools are employedNot applicable
Data Visualization
Visualization tools/applications:
Infogram,
Datawrapper, Tableau, Piktochart etc.
Both programming languages possess
extensive capabilities in creating visualizations.
From Table 1, one can easily conclude that both approaches can support all stages of data journalism. Nevertheless, it is worth
noting that in the code-based approach, all data journalism stages are supported through the same platform (i.e., Google
Colab), whereas in the tool-based approach, no interoperability exists, and thus journalists have to manually export/save and
import/load the data from one tool to another. is adds complexity to the process and creates opportunities for errors that
may slow the production of data journalism articles.
2. Methodology
For this study, a mixed approach was selected, which included a literature review and case studies. e case Study
methodology is a widely used research approach that can support the study of complex phenomena or processes. It is a
qualitative research technique widely adopted in various disciplines, including social sciences, education and business. e
case study methodology involves an in-depth exploration of one or more cases, utilizing various data collection techniques
such as interviews, documents, and observations. It oers a holistic perspective and generates rich and detailed insights into
the research topic (Kratochwill
et al.
, 2013; Tellis, 1997).
Despite its advantages in providing comprehensive and in-depth insights and contributing to theory building and practical
applications, as Yin (2014) suggests, the case study methodology has limitations. Baxter and Jack (2008) caution about the
limited generalizability of case study ndings due to the focus on a single or a few cases.
24 |
nº 43, pp. 17-34
|
July-December of 2026
A Comprehensive Analysis of the Importance of Programming in Modern Data Journalism
ISSN: 1696-019X / e-ISSN: 2386-3978
doxa.comunicación
For the purpose of the study, two data journalism articles were selected. One utilizes data that can be characterized as big data,
and the second is based on data of moderate size. Both were published by iMedLAB
3
, a non-prot organization that aims to
enhance transparency, credibility and independence in journalism. iMedLAB was founded in 2018 and funded by the Stavros
Niarchos Foundation (SNF)
4
. e specic organization was chosen because it publishes articles in English (and in some cases,
also in Greek) and datasets that have been employed in its data journalism articles. It also provides information on the process
and tools employed for the collection of datasets and the creation of the visualizations, which is crucial for the purpose of this
study. e two selected articles are as follows:
Article 1:
MyCoast app: A data analysis of leasing agreements for beaches in Greece (published August 2
nd
, 2024) written by
anasis Troboukis, available at
https://lab.imedd.org/en/analysame-ta-dedomena-gia-tis-ellinikes-paralies-tou-my-coast/
Article 2:
In Greece, more people abstained from voting than participated in the European elections (published June 18
th
,
2024), written by Kelly Kiki and Chrysoula Marinou, available at
https://lab.imedd.org/en/more-people-abstained-from-voting-than-participated-in-the-european-elections/
3. Analysis of case studies
Next, the content of each article is briey presented. Article 1 reports on a quite popular issue in Greece during the summer,
using data from the MyCoast app, an application that was deployed through which Greeks could check whether the concession
terms were met on the beaches they visited and, if needed, submit a report. e article includes several static and two more
complex visualizations (see gure 1) that oer some degree of interactivity (the user can scroll and alter the data presented in
the visualization). is article also includes a detailed map of the data used. e dataset can also be downloaded in CSV format.
e inspection of the dataset revealed that it comprised 9170 rows with a total size of 4,26 Mbytes. e size of the dataset is
quite large; therefore, it can be characterized as big data. e dataset was publicly available, and thus, we can assume that
the dataset was obtained directly from the Greek government and no special process (i.e., web scraping, search) was utilized.
Nevertheless, the iMed team studied 9,170 coastal concessions and analyzed 3,957 contracts to derive nancial considerations
for each lease. is means that the team managed and linked related data types. Notably, 43.0% of the records in the MyCoast
app did not include relevant lease contracts. Consequently, the team implemented ltering, excluding municipalities (i.e.,
Paros, Mykonos, ira, and Patmos) to ensure that the results were representative of reality. It is evident that the entire process
(managing, cleaning, and ltering) was quite complex, which strongly implies the use of programming.
It is not evident whether the visualizations were created using specic tools, as they do not seem to employ template-
based layouts or include any app-specic elements (watermarks, branded fonts, recognizable UI components). In contrast,
visualizations have consistent and precise styling and employ simple color schemes (Figure 1). Additionally, the types of
3
https://lab.imedd.org/
4
https://www.snf.org/en/
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
doxa.comunicación |
nº 43, pp. 17-34
July-December of 2026
Andreas Veglis
ISSN: 1696-019X / e-ISSN: 2386-3978
| 25
visualizations, thermal maps, the rst displaying concession concentration, and the second cost per square meter (the spot
size and color relate to higher cost), are very specialized, and it is very dicult to create them with standard data analysis tools
(that is, DataWrapper and Flourish). Overall, there is a high probability that the visualizations in article 1 were created via
programming. It is also probable that other stages of the data journalism project (data cleaning, data understanding, and data
validation) were also completed using similar methods.
Figure 1. Two static visualizations exhibiting (a) the highest concentration of concessions for large areas of the seashore or beaches and (b) the
cost of leasing per square meter (
https://lab.imedd.org/en/analysame-ta-dedomena-gia-tis-ellinikes-paralies-tou-my-coast/).
<p>[Enlace de URL / hc (has AS)]</p>
26 |
nº 43, pp. 17-34
|
July-December of 2026
A Comprehensive Analysis of the Importance of Programming in Modern Data Journalism
ISSN: 1696-019X / e-ISSN: 2386-3978
doxa.comunicación
Article 2 discusses the issue of participation in the European elections in Greece. Specically, it presents data that indicate
that the abstention from the European elections in 2024 has increased in Greece by 17,3 percentage points in comparison with
the previous European election in 2019, although abstention remained stable in the European Union as a whole. e article
includes one static and two interactive visualizations (Figure 2) and embeds another visualization of the issue of abstention
that was created by a cross-border research by EDJNet
5
. Visualizations were created using DataWrapper and Flourish. e
article includes an option for readers to download the datasets used. ere are four txt les and two small CSV les with the
data utilized for the creation of the DataWrapper visualizations (an option that is embedded in the visualizations). As the
authors of the article state, data were collected from various ocial online sources and scraped from PDF les (that were
available online). While scraping can imply the utilization of programming, the moderate size of the data and the use of
DataWrapper for the visualizations force us to assume that no programming was used for the needs of this article. Various
tools that support the extraction of tables from PDFs are available and could have been employed. e same thing has probably
happened with data cleaning, understanding, and validation. A spreadsheet application was probably employed, which can
easily handle small data sets.
It is worth noting that although it is evident that the visualizations were created with the help of specic applications, datasets
in CSV and TXT les can also be accepted in Python and R (in the case of R the process is slightly more complicated)
5
https://abstencao.divergente.pt/en/home
<p>[Enlace de URL / hc (has AS)]</p>
doxa.comunicación |
nº 43, pp. 17-34
July-December of 2026
Andreas Veglis
ISSN: 1696-019X / e-ISSN: 2386-3978
| 27
Figure 2: Visualizations (static and interactive) of various data on the issue of abstention in the European Elections (
https://lab.imedd.org/en/
more-people-abstained-from-voting-than-participated-in-the-european-elections/).
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
28 |
nº 43, pp. 17-34
|
July-December of 2026
A Comprehensive Analysis of the Importance of Programming in Modern Data Journalism
ISSN: 1696-019X / e-ISSN: 2386-3978
doxa.comunicación
4. Discussion
Based on the results of the two case studies examined in the previous section, this section discusses the research questions of
the study.
Concerning
RQ1,
whether coding is considered a necessity for practicing data journalism, we can answer that coding is not
an absolute must, and in many cases, the tool-based approach can provide sucient results. As is evident from Article 2
(which was presented in Section 3), the tool-based approach can be employed with good results. Heravi and Lorenz (2020)
reported that although 69% of data journalists use some form of coding in their work, many successful data journalists mainly
employ spreadsheet software and visual analytics tools. Borges-Rey states that the situation varies signicantly between media
organizations (2016). Usually, large media organizations have teams that include members with technical skills in coding
and working with data, whereas in medium to small media organizations, they are more in favor of the tool-based approach.
Other media organizations create teams that include both journalists with coding skills and traditional investigative reporters
working with data, promoting complementary rather than mandatory coding (Parasie and Dagiral, 2013). On the other hand,
Young and Hermida found an increase in no-code and low-code tools in newsrooms (2015), tools that can make complex
data analysis possible for journalists without coding skills. Parasie and Dagiral view the use of coding in data journalism as
an evolution in the eld and not a prerequisite (2013). However, the majority of data journalists exhibit a strong interest in
acquiring programming skills, as they will be able to be more productive in the eld (Heravi, 2018).
In other words, while coding cannot be considered an absolute requirement for practicing data journalism, it is becoming an
important skill in the eld. It is worth noting that in some cases where data journalism projects involve big data, this is the only
viable approach.
Moving to
RQ2
, which concerns the features of a data journalism project that inuence the decision to utilize coding, we
can draw some conclusions related to the creation of visualizations from the two case studies. Specically, when specic
visualizations are needed that cannot be created with the help of the available data analysis tools, the only option is to use
coding. is means that if the project requires very unique visualizations, such as article 1, where specic areas need to
be marked on a map, it is quite dicult or even impossible to nd data analysis tools that can create such a visualization.
However, in article 2, the visualizations are not very specialized (line chart, bar chart, and choropleth map); thus, well-known
data analysis tools have been employed. Nevertheless, such visualizations can also be created by coding, as both programming
languages (Python and R) oer extensive libraries and packages for creating such visualization types.
As for the other stages of development of the data journalism projects in our case studies, there is no indication of what
approach has been employed. Nevertheless, we can make some speculations. Specically, in Article 1, the data reside in a
single PDF le, which, although quite extensive in size, is still manageable by regular spreadsheet applications (Microsoft
Excel and Google Spreadsheet). However, because the visualizations were probably created with programming, there is a high
probability that other stages of development (for example, data cleaning, data understanding, and data validation) were also
completed through coding. In the case of article 2, the data employed were available from four dierent les (with relatively
small sizes). Because the datasets were obtained from dierent sources and, as mentioned, scraped from PDF les, there is
some probability that a programming approach was used. However, if we consider the facts that a) there are several tools that
doxa.comunicación |
nº 43, pp. 17-34
July-December of 2026
Andreas Veglis
ISSN: 1696-019X / e-ISSN: 2386-3978
| 29
can support this process, and b) the visualizations were created by data analysis apps, it is more probable that no programming
was employed for the needs of article 2. e size of the dataset is a very important factor in selecting a suitable approach (tool-
based or code-based). Additionally, visualization requirements are an important factor. It is also worth noting that it is not very
common to use a mixed approach (tool-based and code-based) during a project.
In addition to the data journalism features that were examined in our case study analysis, there are other parameters that aect
the selection of the approach in developing a data journalism project, that derive from the scientic literature. Personal interest
in technology, prior exposure to programming concepts, and perceived career benets are important parameters inuencing
coding adoption by journalists (Appelgren & Nygren, 2014). Organizational context (newsroom size, available resources, and
institutional support) is signicantly responsible for selecting the code-based approach (Fink & Anderson, 2015). Similarly,
Joung and Hermida highlighted the role of existing technical infrastructure and workow integration as positive parameters for
embracing the code-based approach. Coddington emphasized the factor of working with large datasets and the requirement
of frequent data updates as a reason for choosing to drop the programming approach (2015).
5. Conclusions
is study examined the utilization of programming in modern data journalism. e study’s conclusions reveal the relationship
between coding and practicing data journalism. First, while coding is not a requirement for data journalism, it has become
a valuable skill in the eld. e analysis shows that both tool- and code-based approaches can work in many cases for data
journalism projects. e choice between the two depends on many factors, including project complexity, dataset size,
repetition needs, and visualization requirements.
Specically, the case study analysis shows that the decision to code is mainly driven by the size and complexity of the data being
analyzed, with larger datasets beneting more from a code-based approach. is decision is also driven by the specicity of
visualization requirements, where unique or complex visualizations require coding solutions. Other factors include workow
integration and automation, available resources and technical infrastructure, and journalists’ technical background and
skills. Most importantly, there is no one-size-ts-all approach to data journalism. Coding can provide more exibility and
capabilities for complex data analysis and visualization, but traditional tool-based approaches are still viable and work for
many data journalism projects, especially those with moderately sized datasets and standard visualization needs. New tools
and platforms, such as Google Colab and Excel’s Python integration, are gradually breaking down the technical barriers to
coding in journalism. is means that the distinction between coded and non-coded approaches will become less clear as
hybrid approaches become more accessible to journalists with less technical backgrounds. ese ndings have signicant
implications for journalism education and newsroom practices. is means that while coding skills are useful, they should
be seen as a complementary, not mandatory, tool in the data journalist’s toolkit. e focus should be on strong journalistic
principles and storytelling, with technical skills serving to enhance, not replace, the core of the profession.
However, it is worth noting that this study exhibits certain limitations, since it includes a small sample of case studies from one
media organization. Additionally, it is not easy to dene with complete certainty the approach used for each data journalism
article without acquiring the necessary information from the actual journalism team that developed the projects. Nevertheless,
30 |
nº 43, pp. 17-34
|
July-December of 2026
A Comprehensive Analysis of the Importance of Programming in Modern Data Journalism
ISSN: 1696-019X / e-ISSN: 2386-3978
doxa.comunicación
the current paper should be treated as a concept study, with the aim of opening a debate about the role of programming in
modern data journalism.
Future research could focus on the long-term impact of programmatic literacy in newsrooms, the eectiveness of hybrid
approaches combining traditional and coded methods, and the evolution of data journalism as new tools and technologies
emerge by conducting surveys of data journalists. is could provide the scientic community with insights into how data
journalism continues to adapt and evolve with technological advancements and changing newsroom needs.
6. Acknowledgements
is article has been translated into Brian O’halloran, to whom we are grateful for his work.
7. Specic Contributions of Each Author
Name and Surname(s)
Conception and Study DesignAndreas Veglis
Methodology
Andreas Veglis
Data Collection and AnalysisAndreas Veglis
Discussion and ConclusionsAndreas Veglis
Writing, Formatting, and RevisionAndreas Veglis
8. Conicts of Interest
e author declares that they have no conicts of interest.
doxa.comunicación |
nº 43, pp. 17-34
July-December of 2026
Andreas Veglis
ISSN: 1696-019X / e-ISSN: 2386-3978
| 31
9. Bibliographical references
Angelou, I., Kansaras, V., Kourkouridis, D., & Veglis, A. (2020). Social media followership as a predictor of news website trac.
Journalism Practice
,
14
(6), 730-748.
https://doi.org/10.1080/17512786.2019.1635040
Appelgren, E., & Nygren, G. (2014). Data journalism in Sweden: Introducing new methods and genres of journalism into ‘old’
organizations.
Digital Journalism
, 2(3), 394–405.
https://doi.org/10.1080/21670811.2014.884344
Ausserhofer, J., Gutounig, R., Oppermann, M., Matiasek, S., & Goldgruber, E. (2020). e datacation of data journalism
scholarship: Focal points, methods, and research propositions for the investigation of data-intensive newswork.
Journalism
,
21(7), 950-973.
https://doi.org/10.1177/1464884917700667
Balaji, N., Pai, B. K., Bhat, B., & Praveen, B. (2021, February). Data visualization in splunk and Tableau: a case
study demonstration. In
Journal of Physics: Conference Series
(Vol. 1767, No. 1, p. 012008). IOP Publishing.
https://doi.org/10.1088/1742-6596/1767/1/012008
Batt, S., Grealis, T., Harmon, O., & Tomolonis, P. (2020). Learning Tableau: A data visualization tool.
e Journal of Economic
Education
,
51
(3-4), 317-328.
https://doi.org/10.2139/ssrn.3438993
Bauzon, J., Murphy, C., & Wahi-Gururaj, S. (2021). Using macros in Microsoft Excel to facilitate cleaning of research data.
Journal
of Community Hospital Internal Medicine Perspectives
, 11(5), 653-657.
https://doi.org/10.1080/20009666.2021.1954282
Baxter, P., & Jack, S. (2008). Qualitative case study methodology: Study design and implementation for novice researchers.
e
Qualitative Report
, 13(4), 544-559.
https://doi.org/10.46743/2160-3715/2008.1573
Borges-Rey, E. (2016). Unravelling data journalism: A study of data journalism practice in British newsrooms.
Digital
Journalism
, 4(3), 336-352.
https://doi.org/10.1080/17512786.2016.1159921
Bounegru, L., Gray, J., Venturini, T., & Mauri, M. (2018). A eld guide to’Fake News’ and other information disorders.
A Field
Guide to” Fake News” and Other Information Disorders: A Collection of Recipes for ose Who Love to Cook with Digital Methods
,
Public Data Lab, Amsterdam (2018).
https://doi.org/10.2139/ssrn.3097666
Bradshaw, P. (2017). Computational thinking and the next wave of data journal- ism. Online Journalism Blog.
onlinejournalismblog.com/2017/08/03/computational-thinking-data-journalism
Burke, J. (2023). Why and how to use Google Colab. T
echTarget.
https://www.techtarget.com/searchenterpriseai/tutorial/
Why-and-how-to-use-Google-Colab
Cairo, A. (2012). e Functional Art: An introduction to information graphics and visualization. New Riders.
https://doi.org/10.12775/tsb.2014.025
Card, S. K., Mackinlay, J., & Shneiderman, B. (Eds.). (1999).
Readings in information visualization: using vision to think
. Morgan
Kaufmann. ISBN: 1558605339
Coddington, M. (2015). Clarifying Journalism’s Quantitative Turn: A typology for evaluating data journalism, computational
journalism, and computer-assisted reporting.
Digital Journalism
, 3(3), 331–348.
https://doi.org/10.1080/21670811.2014.976400
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
32 |
nº 43, pp. 17-34
|
July-December of 2026
A Comprehensive Analysis of the Importance of Programming in Modern Data Journalism
ISSN: 1696-019X / e-ISSN: 2386-3978
doxa.comunicación
Colman, F., Bühlmann, V., O’Donnell, A., & van der Tuin, I. (2018).
Ethics of coding: A report on the algorithmic condition
.
https://ualresearchonline.arts.ac.uk/
id/eprint/14236/1/Ethics_of_Coding_A_Report_on_the_Algorit.pdf
Crawley, M. J. (2012).
e R book
. John Wiley & Sons.
https://onlinelibrary.wiley.com/doi/epdf/10.1002/9781118448908
Cruz, J. (n.d.).
Why journalists need data and coding skills more than ever
. International Journalists’ Network. Available at
https://ijnet.org/en/story/why-journalists-need-data-and-coding-skills-more-ever
Few, S. (2013). 35. Data visualization for human perception.
e Encyclopedia of Human-Computer Interaction,
International
Design Foundation.
https://data.europa.eu/sites/default/les/course/4.1_DataVisualizationForHumanPerception.pdf
Fink, K., & Anderson, C. W. (2015). Data Journalism in the United States: Beyond the “usual suspects”.
Journalism Studies
,
16(4), 467-481.
https://doi.org/10.1080/1461670x.2014.939852
Gray, J., Bounegru, L., & Chambers, L. (2012).
e Data Journalism Handbook
. O’Reilly Media.
https://datajournalismhandbook.
org/uploads/rst_book/DataJournalismHandbook-2012.pdf
Guerrero, H. (2019).
Excel data analysis
. Springer International Publishing.
https://educons.edu.rs/wp-content/
uploads/2020/05/2019-Excel-Data-Analysis.pdf
Hamilton, J. T. (2016).
Democracy’s Detectives: e Economics of Investigative Journalism
. Harvard University Press.
https://doi.org/10.4159/9780674973572-001
Hannaford, L. (2015). Computational journalism in the UK newsroom.
Journalism education
, 4(1), 7-21.
https://e-space.mmu.
ac.uk/575953/1/Article%20Computational%20Journalism.pdf
Heravi, B. R. (2018). Data journalism in 2017: A summary of results from the global data journalism survey. In T
ransforming
Digital Worlds: 13th International Conference,
iConference 2018, Sheeld, UK, March 25-28, 2018, Proceedings 13 (pp. 107-
113). Springer International Publishing.
https://doi.org/10.1007/978-3-319-78105-1_13
Heravi, B. R., & Lorenz, M. (2020). Data journalism practices globally: Skills, education, opportunities, and values.
Journalism
and media
, 1(1), 26-40.
https://doi.org/10.3390/journalmedia1010003
Kinnestand, S. (2023). Introducing Python in Excel: e Best of Both Worlds for Data Analysis and Visualization, Microsoft 365
Blog.
https://rb.gy/nshr7e.
Knight, M. (2015). Data Journalism in the UK: a preliminary analysis of form and content.
Journal of Media Practice
, 16:1, pp.
55-72.
https://doi.org/10.1080/14682753.2015.1015801
Kratochwill, T. R., Hitchcock, J. H., Horner, R. H., Levin, J. R., Odom, S. L., Rindskopf, D. M., & Shadish, W. R. (2013).
Single-case intervention research design standards.
Remedial and Special Education
, 34(1), 26-38.
https://doi.
org/10.1177/0741932512452794
Lewis, S. C., & Westlund, O. (2015). Big Data and Journalism: Epistemology, expertise, economics, and ethics.
Digital
Journalism
, 3(3), 447–466.
https://doi.org/10.1080/21670811.2014.976418
Lorenz, M. (2010). Data driven journalism: What is there to learn? Presented at
IJ-7 Innovation Journalism Conference
, 7–9
June, Stanford, CA.
https://www.slideshare.net/slideshow/datadriven-journalism-what-is-there-to-learn/4644302
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
doxa.comunicación |
nº 43, pp. 17-34
July-December of 2026
Andreas Veglis
ISSN: 1696-019X / e-ISSN: 2386-3978
| 33
Loth, A. (2019).
Visual analytics with Tableau
. John Wiley & Sons. ISBN: 9781119560203
Lutz, M. (2001).
Programming python
. O’Reilly Media, Inc. ISBN: 9781449398712
Mayer-Schönberger, V., & Cukier, K. (2013).
Big data: A revolution that will transform how we live, work, and think
. Houghton
Miin Harcourt. ISBN: 0544227751
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), 853-871.
https://doi.org/10.1177/1461444812463345
Porlezza, C. (2024). e datacation of digital journalism: A history of everlasting challenges between ethical issues and
regulation.
Journalism
, 25(5), 1167-1185.
https://doi.org/10.1177/14648849231190232
Protopsaltis, A., Sarigiannidis, P., Margounakis, D., & Lytos, A. (2020, August). Data visualization in internet of things: tools,
methodologies, and challenges. In
Proceedings of the 15th international conference on availability, reliability and security
(pp.
1-11).
https://doi.org/10.1145/3407023.3409228
Rogers, S. (2008). Turning ocial gures into understandable graphics, at the press of a button. Inside the Guardian Blog,
https://www.theguardian.com/help/insideguardian/2008/dec/18/unemploymentdata
Ruppert, E., Isin, E., & Bigo, D. (2017). Data politics.
Big data & society
, 4(2). DOI: 10.1177/2053951717717749
Setiyanto, S., & Setiawan, I. (2022). Data science with excel.
International Journal of Computer and Information System
(IJCIS),
3(3), 104-110.
https://doi.org/10.29040/ijcis.v3i3.79
Siapera, E., & Veglis, A. (Eds.). (2012).
e handbook of global online journalism
. John Wiley & Sons.
https://doi.org/
10.1002/9781118313978
Silverman, C. (2014). Verication Handbook. European Journalism Centre
http://vericationhandbook.com/downloads/
verication.handbook.pdf
Simon, B. (2021). Coding With Data in the Newsroom, in Bounegru, L., & Gray, J. (eds).
e Data Journalism Handbook:
Towards a Critical Data Practice
(p. 124-127). Amsterdam University Press.
https://doi.org/10.1017/9789048542079.019
Tellis, W. (1997). Application of a case study methodology.
e qualitative report
, 3(3), 1-19.
https://doi.org/10.46743/2160-3715/1997.2015
Tock, K. (2019). Google CoLaboratory as a platform for Python coding with students.
RTSRE Proceedings
,
2
(1).
https://doi.org/10.32374/rtsre.2019.013
Veglis, A. (2013). Education of Journalists on ICTs: Issues and Opportunities.
Journal of Applied Journalism & Media Studies
,
2(2), 265-279.
https://doi.org/10.1386/ajms.2.2.265_1
Veglis, A., & Bratsas, C. P. (2021). Data Journalism: Denition, Skills, Diculties, and Perspectives. In
Encyclopedia of
Information Science and Technology
, Fifth Edition (pp. 1140-1151). IGI Global.
https://doi.org/10.4018/978-1-7998-3479-3.
ch079
Veglis, A., & Bratsas, C. (2017a). Journalists in the age of Data Journalism: e case of Greece.
Journal of Applied Journalism &
Media Studies
,
6
(2), 225-244.
https://doi.org/10.1386/ajms.6.2.225_1
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
34 |
nº 43, pp. 17-34
|
July-December of 2026
A Comprehensive Analysis of the Importance of Programming in Modern Data Journalism
ISSN: 1696-019X / e-ISSN: 2386-3978
doxa.comunicación
Veglis, A., & Bratsas, C. (2017b). Towards a taxonomy of data journalism.
Journal of Media Critiques
,
3
(11), 109-121.
https://doi.org/10.17349/jmc117309
Yin, R. K. (2014).
Case study research: Design and methods
(5th ed.). Sage publications. ISBN 9781452242569
Young, M. L., Hermida, A., & Fulda, J. (2018). What makes for great data journalism? A content analysis of data journalism
awards nalists 2012–2015.
Journalism practice
, 12(1), 115-135.
https://doi.org/10.1080/17512786.2016.1270171
Young, M. L., & Hermida, A. (2015). From Mr. and Mrs. Outlier to Central Tendencies: Computational journalism and crime
reporting at the Los Angeles Times.
Digital Journalism
, 3(3), 381-397.
https://doi.org/10.1080/21670811.2014.976409
Wu, S. (2013). A review on coarse warranty data and analysis.
Reliability Engineering and
System
, 114, 1-11.
https://doi.
org/10.1016/j.ress.2012.12.021
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>
<p>[Enlace de URL / hc (has AS)]</p>