Seven Imporant Issues Data Journalists Should Discuss Now

Fusion Tables tutorial, how antique!

This was my first impression when I looked at the program of Nordic Data Journalism Conference held at Stockholm, Sweden.

I´m attending the conference as invited speaker. I have a slot of 30 mins to tell about best practises in Helsingin Sanomat data desk, which I run.

After that, I´ll have plenty of time to listen to others. In the program, there are topics like “what is a data desk” and “using excel”.

These are good things to learn for beginners. But for once, I´d like to discuss real everyday problems about my work with seasoned colleagues in the same field. (And like in every conference, that happens also here in evening sessions.)

If I would organize a professional data journalism conference, there would be presentations on these topics:

One of my main concerns is: do our news apps work?

We know, that the headline determines the click rate and people spend more time in articles with news apps. What we don´t know is: Do our readers click and use our apps?

We just made our first trial to record user behavior few weeks ago. We basically attached a logging event to every possible click in one of our news apps.

People did click it, 96% of all entries to article page clicked on the app. So result was good. But there were problems with the logging, so this high result cannot be trusted.

From a presentation, I´d like to hear: How to measure in-app clicks? What is a good conversion percentage?

“Avalance it!”

This is the new hip slogan in digital newsrooms. After NYT published the avalanche piece Snow Fall, our bosses have been demanding us to do the same.

They look great. South China Sea piece, Tomato Can Blues and also the things we´ve made are visually impressive.

But the real issue is: have you actually read the whole Snow Fall?

For readers, are they better than text? Do Snow Falls work? Do we make these things just to get the WOW! -effect, not to tell the stories best way possible. Perhaps text and images would still be a better way, just like Medium does it.

My hypothesis is, that they are just thought to be nice, but in the long run our readers don´t like them more than regular well-written articles.

In the wake of Snowden revelations of NSA spying on everyone, the tools of data journalists need closer scrutiny.

By combining data bases we have a strong tool to get info even on individual people. As an emerging profession, we have not thought what lies in the end of this road.

All that is public and open data cannot be published, if we are to remain ethical in our work. This applies to names of real people, but also information gathering and crowdsourcing.

For example: is it ok for a (digital) newspaper to publish a database of 160000 richest people in Finland and their income since 1999? I think it is, but there could be good arguments against this.

As mobile penetration nears half of all users, we need to focus on making news apps mobile friendly. Actually, this has been so for the last two years.

It´s not easy. We have to make things work with IE8 (thanks a lot Jquery 2.0 and D3) and at the same time have things work in mobile.

Our solution is to make everything responsive between 298 and 560 pixels (our text area min and max width). We do the app in 560 and then scale it.

So, how to cope with small screens?

Open data folks are obsessed with getting the data. It has been an important issue for many years and the opening of data in Finland is not yet complete.

But at the same time, we should look forward. Analyzing the data is quite primitive still almong journalists.

Where could we get better ideas from?

To Iframe or not to Iframe? That it the question everybody must answer and both replies come with a price.

We have opted for using Iframes. It gives us speed and flexibility, but the price is that we have no control of the whole (parent location href parameters, anyone?).

How should we publish our things? Are there tools to make Iframes better?

We use Trello to organize our projects. We have a development and production servers to test and deploy our stuff. We are starting to use Git as version control tool.

I´d like to know what kind of advanced tools and workflows are used in day to day business in datadesks worldwide.

Leave a Reply