Data journalism survey: a mixed picture




A word cloud showing frequency of types of visualisation mentioned by survey respondents


A survey of data journalism training needs shows a fractured picture of activity across the news industry.

The survey, by the European Journalism Centre (EJC), asked 16 questions covering how journalists use data, their levels of expertise, and the barriers they faced. The survey was distributed through networks commonly used by data journalists, including the Open Knowledge Foundation blog and conference, Journalism.co.uk, and the EJC community.

Over 200 people responded, from 40 countries in Europe, North and South America, Australasia, Asia and Africa, including many from outside traditional newsrooms, including consultants, PR professionals entrepreneurs and academics. Online publishers employed the largest number (37%) followed by newspapers (18%), magazines (13%) and broadcasters (10%).

While 70% of respondents unsurprisingly felt that data journalism was 'very important', an even higher proportion felt that a 'lack of adequate knowledge' was the biggest hurdle to using data.

On an organisational level, there was an almost even split between those who had already published a data project, those who were working or planning to produce one, and those who had no immediate data journalism plans.

And there was little evidence of this being the preserve of a generation of 'digital natives': 34% of respondents were aged between 35 and 50, and another 33% between 25 and 35. Only 11% were under 25. But these were highly educated individuals, with 42% having completed Masters level education.



Of the 144 respondents who answered questions on working practices, almost half had collaborated with programmers; a similar proportion had worked with visual designers; and around a third with statisticians.

Notably, 53% said that they did statistical analysis alone, and 16% felt that they had never needed to do such analysis.

When it came to training, most respondents wanted to learn how to integrate data into stories. 65% wanted to focus on techniques to interrogate data; and 54% wanted to learn how to develop data as a service, through products such as news apps.

More specifically, there was a wide spread of interest in a range of skills. Even the least appealing skill listed - programming - was of interest to a third of respondents, while the most appealing - analysing data - interested almost three-quarters.

The spread of interest in different skills is shown in the chart below.




What data is being used by data journalists?

When respondents were asked about what types of data they had recently
analysed, the variety of responses was notable:


Word cloud of responses to 'What kind of analyses did you need to do recently?'


And when asked what data respondents had needed but were unable to access, many replied "none".


Responses to 'What data did you recently need but were unable to access?'


When asked what areas would most benefit from data journalism, however, education emerged as a clear favourite (the phrasing of the question may be key here):


Responses to "Which of the issues that you report on might benefit from raw data?"


It's not clear from the data why this was the case. Reporters' own interests may be one reason (most gave their profession as 'journalist' or 'reporter' without any sector identified) - although the fact that education did not dominate the types of data being used suggests this is not the case. There were also a small number of journalism educators in the sample.

The depersonalised data is available below, so that's one question to start with.



  • Spreadsheet with the depersonalized survey data (the categories: 'name', 'email', 'age', 'gender' were removed and open-ended responses were edited to remove personal information)
  • Survey summary (no open-ended responses)