7 countries, 9 teachers: a dossier of data journalism teaching strategies

Like many journalism educators, I was initially employed by a university because I had worked as a journalist for many years and can bring real world experiences into the classroom. But data journalism presents a bundle of new challenges. I don’t have lived experience of being a data journalist to draw on, and neither do my colleagues.

My way-forward was research. I wrote an academic paper that involved interviewing 35 other Australian journalism academics, at 25 universities, about what they were doing to address the problem. In brief, I found that I was not alone. The others agreed (almost unanimously) that upskilling was an issue. Very few could access funding to bring in experts to teach students, or support staff. Most were trying to cram their own upskilling into their already overflowing workloads and puzzling over what specifically they should dedicate their limited upskilling time to learning. Should it be Tableau, or more Excel, or coding or testing data scrapers, or data visualisation tools?

The next ugliest problem was squeezing it into crammed curricula, already bursting at the seams because of pressure to produce graduates who can work on every platform.

Then there are the students themselves. They present with a mix of skill levels, which is tricky enough, but in this case the mix is due largely to a fascinating phenomenon called ‘math aversion’.

In my own data journalism course, I tackle this issue in my first class. We talk about it, confess to having it and make a pact to deal with it as if we are in recovery mode, feeling the fear and doing it anyway. I took heed of what others in my study said, in comments like: “I map everything step-by-step, so they don’t get frightened” and “we have to find workarounds like online percentage converters”. It’s an approach that, according to the student feedback at the end of semester, is working -- although it means that our exploration of the intricacies of cutting edge data journalism is minimal for now. But we are laying the groundwork, and by tackling the fears we are setting people up for lifetimes of learning.

Other solutions discussed in that paper included using blended learning approaches, such as requiring students to complete Excel Courses, and recruiting specialist trainers to teach students, or to upskill staff.

Since that paper, other researchers have documented additional strategies to build the data capacity of students. For example, the University of New South Wales introduced a data interrogation activity into a first year unit, because thinking about data “from the beginning reinforces for students that it is a core skill, just like interviewing or checking facts”.

The University of Canberra also introduced a digital campaigning unit that is relevant to students in a range communications fields, and includes quantitative literacy and data visualisation skills useful for all students mastering data journalism. In describing the program, Glen Fuller explained that the context of ‘changing the world’ via campaigning injected passion into what students could otherwise frame as dull work.

At Swinburne University, they’ve taken a silo breaking approach and employed a data analytics expert to teach data journalism to students, having considered the difficulties of upskilling existing journalism teaching staff.

All of these approaches are experiments and the results of them are being shared at conferences, in academic journals (such as this special edition of Asia Pacific Media Educator), and in books like News, Numbers and Public Opinion in a Data Driven World. Progress in academia can be slow (because everyone is busy) but this ball is rolling now and a revolution in how we think about the role of data in journalism education is underway.

In 2017, through an online survey we conducted with Tunisian journalists, working across newspapers, radio, and television, we demonstrated the value of technical skills for new graduates. Unsurprisingly, our survey found that 33% of journalists identified writing skills as important -- yet, almost the same number, at 32%, highlighted the importance of technical skills.

In Tunisia, discussions around the teaching of journalism have often focused on the linguistic practice of journalism. The technical dimension, which has disrupted the job market, has not been fully considered, although some universities have added standalone modules into their journalism education programs. But we still yet to see any in-depth change.

Even so, these last seven years, we’ve seen more conversations about open data, open source, big data, data mining, data warehouses, and the potential for their use in journalism. In academia, we have really started talking about databases in Tunisia, particularly in the fields of computer science or statistics. This means that most of the movement has been in engineering schools -- two universities offer a Masters in Big Data -- and, while there is one public school in journalism and three private ones, they neither offer a degree in data journalism or a separate subject on it.

As a result, the teaching of data journalism is always by individual initiative, on the part of a professor who usually teaches web writing or editing, online journalism, or similar. For my own part, this was how I introduced my third year web journalism students to the field. Here, my students learnt how to change data into graphs:

There has also been more movement, following the Arab Spring, when academic staff at the Higher Institute of Arts and Multimedia of Manouba, launched a Masters in Media Engineering. This degree has led to three specialties: web development, 3D visualisation, and community management. Students learn how to develop data stories, to manage data, and even to write, by taking courses in data visualisation, data information, and more generalist courses, such as online editing, which include data skills.

Despite these developments, for many journalism schools, integrating specialised coursework in data and computation presents something of a chicken-and-egg conundrum. Without a data education, there’s few educators to teach data. To help, I’ve developed the following steps on bringing data into a curriculum:

1. While schools may wish to prepare their graduates for this emerging field, the field itself may not yet have enough teachers in its ranks. So, try scheduling guest lectures as a transitional solution.
2. New tools are developing quickly, and it is critical for the faculty to continue to grow, learn, and change as the field itself develops. Everyone has to follow the new wave.
3. Many universities provide computer labs and studios for classes. The primary advantage is the certainty that each student will have a workstation with the necessary technical specifications and software installed. The primary disadvantage is that students may graduate without these tools at home, which they need to practice the skills that they have learnt.
4. Using MOOCs in a complementary fashion with university data journalism courses could help professors integrate new skills into their offerings.
5. Journalism schools should build collaborative partnerships with other disciplines.

Journalism is not a narrow set of traditional newsroom skills, but instead encompasses whatever tools and methods have, in one way or another, been made journalistic.

Several journalism schools around the world have begun building bridges with computer science departments by opening research centers, co-teaching and cross-listing classes, and even developing joint degree programs. This should be the way forward in Tunisia, where expertise is siloed in our journalism schools and multimedia institute. Instead, let’s gather experiences and build a comprehensive approach to teaching the country’s future data journalists.

In Hong Kong, which boasts the most diversified, competitive, and democratic media market in the Asia-Pacific region, skills in data driven reporting are highly valued by middle-level journalists. Irene Jay Liu, one of Hong Kong’s public faces for data journalism, noted in a 2013 interview with the Pacific Media Center (she was then news editor for data at the Thomson Reuters Hong Kong division) that data journalism might be a solution to the crisis in journalism, including layoffs in the newsroom, closure of local and regional newspapers, and decline in circulations brought about by media convergence. While the digital revolution has made it possible for the ordinary public to find, produce, and distribute information they previously relied on journalists for, data journalism, with its presumed sophistication, can be regarded as one crucial saving grace for professional journalism and the mainstream media to justify their continued existence.

However, professional data journalists are still rare in Hong Kong media industry. This may be due to the fact that journalism programs here have a long history of predominantly focusing on conventional skills sets (writing, interviewing, news judgement) and overlooking the necessity of equipping journalism students with data and computational skills.

While newsrooms are making tremendous efforts in using data to produce news, data training in the higher education sector has lagged behind. Whether journalism programs should emphasise data skills in training future journalists has long been a subject for debate. Adding to this, teaching journalism students (who are generally afraid of numbers) data skills and convincing them to build a data frame of mind has proven to be a thorny task. At times, it is a big challenge, if not a battle.

To find out more, we conducted a study of 121 journalism students at Hong Kong Baptist University, which has a top journalism program in Asia. Our survey, combined with in-depth interviews, generated three major findings for journalism educators and future journalists:

1. While journalism students are eager to understand what data journalism is and how it’s practiced, they do not have comprehensive knowledge of data collection, data analysis, and interpretation.
2. Computational tools are absent from current journalism curricula, which leads to students’ misperception about data usage in news reporting.
3. While students have a high willingness for learning data journalism, about half of those surveyed expressed a dislike of data work.

Interestingly, we found that gender and major also played a role in our student’s perception of data journalism. Male students mastered more data-related knowledge than their female counterparts. Those majoring in Chinese journalism showed the least interest in data, compared to our financial journalism majors, who seemed much more comfortable with it -- among the Chinese journalism majors, only 20% identified themselves as having an interest in data, compared to 75% of their financial journalism counterparts. International journalism majors lay in between.

The overall irony emerged from our study is: although our students recognised that knowing about data journalism is a must for their career development and will be advantageous in their professional practice, many showed minimal interests or possess minimal data skills. The reason for this gap lies in the lack of support and insufficient training in data by journalism educators. Consequently, it is hard for students to embrace a comprehensive data skill set, not to mention its integration into journalistic core values. To this end, the foremost priority for journalism programs should be to cultivate a data-friendly culture, that is, for the administration to see data literacy and data skills as an area for investment rather than a cost. Only from there can a data-savvy curriculum be situated for our future journalists.

Although data journalism has been progressively implemented in the study programs of North American and European universities, programs in Latin America have developed slowly, with little presence even at this time.

That said, data journalism training does have some presence in the region’s public and private universities -- for example, Adrián Pino’s program at the University of Concepción del Uruguay (Argentina), where he has incorporated a specific journalism unit dedicated to data. In this unit, Adrián focuses on critical skills, including how to use basic scraping tools, harnessing spreadsheets for processing and analysing data, and foundational concepts and tools for data visualisation. Each student also prepares their own data journalism project, independently carrying out all stages of the story production process until it’s published.

Yet, as with many students around the world, embracing calculations and numbers is one of the central challenges for journalism students in Argentina. Often, they find it difficult to focus on statistical rigor, to incorporate basic mathematics, and to understand that data journalism requires these skills. As educators, we’ve found that the best way to counter this hesitancy is to focus on teaching strategies that foster enthusiasm. For example, when going through the entire process of developing a journalistic project based on data, students have shown greater enthusiasm by taking ownership of the task. Similarly, showcasing the power of data visualisation has lead them to investigate visualisation tools outside of class.

In addition to standalone university programs, a broader initiative by Datos Concepcion has sought to spread data journalism in higher education institutions across the country. Launched in 2017, the Data Journalism Training Program for Universities began with seven universities and added another four in 2018. The program works as an intensive two day workshop (10 hours total) where the basics of data journalism are presented to students through practical exercises that cover data downloading, cleaning, processing, and analysis. It also culminates with the generation of a data journalism project, which again has helped participants connect with the team, become enthusiastic about data, and even realise their first data driven news project.

But it is difficult to measure the success of these programs because there have been no studies looking at the status of data journalism education in Latin American universities, or even in the media teams of our region. The absence of impact measurements, which show the capacity for change and the real impact of data journalism teachings, presents a limitation for educators. Moving forward, we believe that developing new impact indicators for data journalism and surveying the state of its instruction would be a great way to develop effective teaching strategies tailored to Latin American students.

For example, Latin America, known for its booming open data scene, should be the perfect staging ground for data journalism education. Many of those who teach data journalism in the region’s universities participate in the open data movement and many of the organisations that push open data and transparency processes have built trainings that could be better used by university programs. Yet, there are questions around how to better leverage these alliances and take advantage of capabilities on both sides.

Although we have academics and researchers who could advance this area, a lack of financing is generally the factor that delays these opportunities. Without these advances, it is clear that challenges will await us in the years to come, requiring articulated and shared efforts among the educational organisations across Latin America.

I’ve taught students, interns, and professionals across South America, New Zealand, South Africa, and the United States for anywhere between a single hour-long session to a semester-long course. It’s hard work, and I’ve not gained full mastery. While the daily work of teaching is an enormously joyful experience and I’m grateful for my students’ efforts and progress, I often arrive at the end of the semester feeling drained, and slightly dissatisfied. There is always a gap between what I know is possible and how far the students have gone, and I’m all too aware of the many ways I could have taught better. At the university level, where I’ve worked for the past five years, I’ve wrestled with the challenge of how to teach my students skills, integrate them into the global community of people doing data journalism, and carry out a meaningful project in 15 weeks, with all of the competing challenges of work, other classes, internships, family, and social life.

At the same time, thinking back to my fourth grade teacher and hearing from students I taught more than two decades ago, helps me remember that any teaching is an entry point into a subject’s universal principles and their long-term connection with them. The spark that I have lit with these students helps me keep going through the inevitable struggles I face with the current group. I also draw solace from the knowledge that I’ll have another chance in the following semester, as well as from the humbling experience that we never know what, despite our greatest efforts, our students will take from their time with us.

Even with this uncertainty and continual desire to improve, I’ve developed certain approaches that guide me through each semester of teaching data journalism.

Here are six learnings that ground me and my students:

1. Investigative Reporters and Editors (IRE) Contest entries are your friend

Reading these entries teaches students how to reverse engineer an investigation, albeit with some chest thumping and the larger objective of winning a coveted IRE award. By examining them, and then reaching out to their respective reporters, I’ve found that students start to both learn how an investigation is done and to join the global community of people doing that work.

2. Build incrementally

It helps to start with smaller stories and projects that become the basis for larger, more ambitious ones. I taught a class one semester where I had the students go through the process twice. The first story was with a dataset I provided, while the second was with either an additional one related to the first, or one that they had selected. Building muscle memory through repetition and leading into increased complexity is a positive way to go.

3. Emphasise the importance of failure

It’s hard to overstate how critical it is for students to feel comfortable embracing failure as a necessary part of growth, particularly in data journalism. Often, we not only learn more from our mistakes, but this is where gains come as we strive and reach for different elements.

This has happened to me numerous times, but one that comes to mind is a lesson on the importance of the margin of error. The day before a major public presentation, I learnt that, during the fact-checking process, an analysis that I thought showed major disparities between neighborhoods turned out to be completely within the margin of error. My editor and I scrapped the entire analysis, regrouped over some Chinese food, and launched into a frenzied effort to come up with a new angle. We eventually found one, but only after staying at the office until the early hours of the morning. You can be assured that I check the margin of error every time now!

4. Provide a lot of opportunities for students to get comfortable as a first step towards developing expertise

Some students are very comfortable with digging into and splashing around in the sea of data. Others like to do visualisations. Others feel maps are their thing. Still others excel at moving from working with data to taking their findings into the real world. Offering all kinds of learning options provides more opportunities for students to find their groove and to understand how different storytelling pieces connect to each other.

5. Encourage students to be patient and enjoy the hunt

Going ‘down the rabbit hole’ is a phrase that often has negative connotations of losing the point and wasting time. While that perspective has some merit, there is value, too, in thinking about and exploring all kinds of questions before deciding on a specific course.

6. Remember that the data is never the story

I once said to my wife, “Honey, it’s Saturday night. I’m here with you. I’ve got a tasty glass of wine and a fresh new dataset. Can life get any better?” (“I certainly hope so,” she answered without a second’s hesitation.) As much as I love data, I’ve come to understand more and more that many of the most powerful stories come from weaving together findings with stories of individuals and communities whose experiences reflect the impact of a policy or phenomena.

Although arduous and at times frustrating, working with data and teaching data journalism are tremendously rewarding parts of my life. The ability to make a real impact through the data driven illumination of abuse and oppression is a deeply meaningful experience; to work with young people as they begin their journeys to do the same is even more so.

It’s into this environment that we first introduced Kyrgyz students to data reporting, with the help of UNDP at the Data Journalism Summer Institute in 2017. Following the success of this programme, I decided to stay on to build a formal classroom framework at the American University of Central Asia in Bishkek.

Although ambitious, it was decided with the department leadership to divide the necessary theory and practice into five courses and offer them continuously at the Bachelor level over the course of 2.5 years -- from an introduction to data journalism to data analysis, storytelling, and visualisation. We wanted to promote a holistic understanding of the various processes and skills involved in the production of solid data journalism work.

This change turned out to be easier done in practice, than implemented on paper. Even though I was invited to lead the university’s data journalism concentration, some of our classes sat under pre-existing courses such as ‘Newswriting’ and ‘Advanced Reporting’. Others, like ‘Data Storytelling and Data Visualisation’, were offered as electives and most of our students were eager to finish the concentration. From my limited observation of two student groups so far, data journalism is not perceived as rocket science but rather as something cool. Even pivot tables -- sometimes, especially the pivot tables.

But, in developing the data journalism concentration, we tried to change the set of required courses in the journalism schedule, which presented a unique challenge: these conflicted with the State’s standards for journalism education. And so, because it is becoming clear that a knowledge of basic Excel is as much of a must for journalists as grasping basic grammar and document editing software, we have to offer many of our data journalism courses as electives to complement the State-approved curriculum. By allowing this combination of required and elective courses, we hope that every future journalist in Kyrgyzstan will be at least data literate, with the opportunity to learn more advanced skills.

This is how data journalism courses can spread from the American University of Central Asia, which is often a pioneer and innovator in the country’s higher education system, to other journalism schools, bounded by limited budgets and the skills of their staff. Yet, to facilitate this growth, we need proper research into how journalism education itself should be updated to equip future graduates with the skills for the market in Kyrgyzstan.

In addition to building a comprehensive course, I felt the need to look beyond our cosy classroom and consider the broader data journalism environment. Where will my students publish their first data journalism projects? Who will have vacancies for them once they graduate? Are there enough designers, developers, and other professionals who understand the principles of data storytelling and would be able to work with them on their projects?

With this in mind, I started working in the other direction, too. Together with the alumna of that 2017 Data Journalism Summer Institute, we co-founded the School of Data Kyrgyzstan, a chapter of the global network, aimed at promoting data literacy across Central Asia. To truly prepare students, we run hackathons, where students work alongside journalists, designers, and developers on a data story; offer internships and jobs to help students build experiences; and, we provide them with the opportunity to step up as assistant trainers at our workshops.

We’ve also taken some steps to make data journalism education more available and accessible to other students. For example, we created an online course on data communication, available for free here in both the Russian and Kyrgyz languages. This is already being used by the teaching staff at Osh State University as a blended learning component. And, to really help spread data expertise, we’ve also planned for a two-week retraining programme for journalism professors from all over the country. Thanks to various grants, this is entirely possible to implement and its impact on Kyrgyz data journalism is only a question of time.

To sum up, two years into teaching data journalism at the university-level in Kyrgyzstan, and there are still lots of things to be improved. We are still yet to update the State’s standards and the required courses for university-level data journalism -- which will be a long process. However, our experience has shown that this challenge can be overcome by revisiting the design of the concentration and seeking better cooperation with other departments such as sociology, statistics, or IT; involving actual journalists in the learning process and in the format of production labs; or, even experimenting with the spectrum of formal and informal education to find an ideal balance. So, while I’ve been working to teach data to my Kyrgyz journalism students, my work the university has also taught me to accept things in transition.

As my personal attempt to try to remedy this, I started Ireland’s first postgraduate data journalism module for the Journalism MA programme at the National University of Ireland in Galway in September 2015. Later that year, I was asked to teach a Dublin City University undergraduate module on Data Journalism too.

In 2016, I joined the School of Information and Communication Studies at University College Dublin (UCD), which gave me the opportunity to start Ireland’s first dedicated data journalism programme. Gearing up to design this new programme, I -- of course -- approached the problem as most academics would: by studying the field.

I first studied the state of data journalism practice and data journalism educational needs globally through a survey I ran in collaboration with Mirko Lorenz of Datawrapper. I then studied and analysed around 220 data journalism related courses I could find across the globe. Following these studies and a feasibility assessment locally, we decided that a part time postgraduate certification programme aimed at professional journalists and journalism graduates was our best way forward at UCD, and in Ireland. Consequently, we designed and launched the UCD Data Journalism ProfCert programme for a September 2017 intake.

As part of UCD programme, our students are introduced to a variety of data journalism techniques and the tools needed to complete a data journalism project lifecycle, as well as being trained in quantitative data analysis, statistics, and R. In the second semester, we run a data journalism studio, which takes students through the production of data driven stories, maintained on a fully-fledged data journalism publication website.

One of the key challenges, for any journalism educator, is to prepare students for real-life storytelling. So, through the data journalism studio, we partner with Irish news organisations for co-publications of a selection of our students’ data stories. Examples of such co-publications are RTÉ’s investigations on ‘Personal Injury Claims in Ireland’ [UCD version, RTÉ version], and ‘Most dangerous cities for Gardai’ [UCD version, RTÉ version], as well as the Irish Independent stories on ‘Hottest travel destinations for Irish sun-seekers’ [UCD version, Irish Independent version] and ‘Patients on trolleys in Irish hospitals’ [UCD version, Irish Independent version]. These co-publication initiatives have paved a way for exciting collaborative opportunities between the programme and news media organisations, while adding a real, tangible industry-focused aspect to the programme.

Since 2014, I have trained around 100 data journalists in Ireland. Despite this figure, and the good quality of work produced by my students, positions hiring specifically for ‘data journalists’ are still very rare to come by. My hope is to train as many data journalists as possible -- and, as the number of skilled graduates grows, that they themselves will change and re-shape the data journalism landscape in Ireland, and beyond.

Are you a teacher? Share your thoughts in our comments below.