View this email in your browser
2019-04-29 | J++ Nyhetsbrev #74
In this week's newsletter we will talk a bit about the drama unfolding over at The Markup and learning from mistakes. Also, don't miss the new series of Headlong's podcast under the Inspiration section. It really is a must-listen.

The rise and fall (?) of The Markup

The Markup was arguably one of the more interesting new publishers set to launch this year. With 20 million dollars from the founder of Craigslist, prize winning journalists from The Wall Street Journal and ProPublica and a clear goal: investigate the tech industry, it seemed like a slam dunk. But it has all gone terribly wrong. Last week five out of seven editors left The Markup, following the firing of its Editor in Chief, Julia Angwin.

If you don't know who Julia Angwin is, it's about time you start googling. There is a lot for reporters and editors to learn from Angwin but one of the most fundamental things is perhaps her approach to investigative journalism and the role of data. These two graphs summarise it quite well:
Now, the story about what has been going on inside The Markup is slowly starting to get out and, disregarding the personal disagreements amongst people involved, it concerns some interesting topics about running a newsroom: personality tests, quantifiable quality measures and the very important albeit fuzzy line between advocacy and journalism. Columbia Journalism Review has summarised what happened here:

Since we are Angwin fans, we also recommend listening to her interview with Kara Swisher at Recode Decode. In it, they touch on the situation at The Markup and a lot more:

Mistakes, what are they good for?

It is always nice when people share their experience of failure and reflections on how to do better. Sarah Leo at The Economist has dived into the publication's archives to find graphs where they messed up either by misleading the reader, confusing her or simply by failing to make a point. She tells us about what went wrong and gives suggestions on how to fix it:


Headlong's new series running from COPS is a must-listen if you are interested in data, crime and policing. First: using "reality" TV  to investigate police, politics and power is excitingly clever. Second: the producers have watched and categorised just under 850 episodes of COPS, gathering around 68,000 data points on suspects and crimes portrayed.

Using grids to visualise correlation seems to be becoming a trend, and we do like it. Previously we have mentioned SVT and their awesome grids in How The Swedes Voted. Now, Yle has done pretty much the same thing, but for the finnish election. Who will make the prettiest grids for the EU election?

Did you know that there is some actual gerrymandering going on in the EU? We didn't. The Pudding has the story.

J++ news

What we've learned about automation and visualisation

If you are near Helsinki, take the chance to come and listen to Jens Finnäs at the Visualizing Knowledge Festival, May 10th at Aalto University. He will be sharing some of the key things that we've learned about newsroom automation and data visualisation in the past years:

We are a new DJA partner

We are a proud new partner of the Data Journalism Awards. This year resulted in 608 nominated projects from 62 countries. Kuek Ser Kuang Keng and Marianne Bouchart analysed all of them and found new actors doing data-driven reporting.

Analysing voting behaviour in the EU parliament

Hot tip: look under the hood of and you'll find an easily scrapable API for EU parliament voting. We used it to help freelance journalist Emanuel Karlsten track which MEPs correct their votes most frequently. There is plenty more data journalism waiting to be done on this data. For example, how loyal have your country's MEPs been to their parties?
J++ Stockholm |

Want to change how you receive these emails?
You can update your preferences or unsubscribe from this list.
Copyright © *|CURRENT_YEAR|* *|LIST:COMPANY|*, All rights reserved.

Our mailing address is:

Want to change how you receive these emails?
You can update your preferences or unsubscribe from this list.