This is a practical presentation from NODA 2017, the Nordic Data conference, this year held in Odense. The presentation discusses tools (specifically Datawrapper) and general editorial approaches to data-driven journalism. The presentation advocates a pragmatic approach - based on searching for data, questioning, visualisation and written texts. This approach could provide opportunities specifically for regional/local media.
Be aware that as this is a presentation from the people behind Datawrapper the tool is discussed and presented on a number of slides in this presentation.
Odense, Denmark - January 27-28, 2017
Mirko Lorenz, Cofounder and CEO
Three key points:
Note: This is a slightly edited version of a talk given at NODA
2017 in Odense (Jan. 2017). Changes were made to
make the presentation easier to understand when the
spoken narrative is missing.
development so far
Datawrapper is a web-based
visualisation tool, providing a
workﬂow to create responsive charts
The tool aims to focus on quality and
is optimised for use in newsrooms.
D3.JS and other visualization libraries are
great, but hard to use for journalists...When we started in 2011,
the goal was to close a gap…
with many new
options to create
are not trained to
work with code.
A workﬂow for daily
data search, ﬁltering and
publishing in different ways.
This is how Simon Rogers
created the widely noted
Guardian Data Blog.
Not every data story is a complex interactive.
There are multiple options to publish data,
including simply making it more public.
Even raw data can be very helpful for a community,
if put into context in the right way.
Another good example:
Atlas by vox.com is demonstrates the
value of constant, daily
digging for data to enrich the story.
Why is hardly any regional/local
medium working in this way?
Great current example
how powerful data can be.
If one economist from Oxford University
can start something like this, what keeps media companies from doing the same?
„Our World in Data“ by Max Roser:
Putting things in comparison,
for better understanding.
The Wirecutter shows
how we can and should create value for users.
Building trust is through being trustable.
Announce your intention, then stick to it.
The Wirecutter makes money through afﬁliate links.
A bit more background here.
Why not combine Quartz Atlas,
„Our world in Data“ and The Wirecutter -
for the beneﬁt of local newsrooms?
We need this, because often statistics
remain difﬁcult to understand.
By ignoring statistics and failing to
turn hidden facts into comprehensible stories
we do damage to a source of truth…
Track regional numbers.
Track numbers better than ofﬁcials.
Transform the numbers.
Reduce need to search.
Show what is not known.
Be there when decisions are made.
Do the math, because no one else does it
for many people.
How to do better
Smart way to engage
with the reader.
Source: New York Times
Everyday I write the book
Start doing daily data journalism.
It is not a question of budget,
it’s a quest for depth and quality.
One search technique every journalist can use:
Combine Google search with „data“, then look at results under „Images“.
Be on the lookout when the statistical ofﬁce
(any) publishes new data.
ChangeDetection is just one service which
can be used for that.
No map, no interactive - but well
done, basic charts instead. When
working with charts simplicity is
often the better choice.
Archie Tse: Why we are doing fewer interactives. Presentation from Maloﬁej 2016 (PDF)
The New York Times
thinks twice before
starting a big interactive
data project - although they
know how to it.
Why is simplicity often better?
Nobody will look at a
line chart? Think again.
This is the rise
of beer prices
A key principle,
three simple charts