When to trust data visualisation (and when not to)

If you ask whether people prefer to see images rather than text to process an information, I’m pretty sure the answer would be a resounding yes. Why?

Because humans are visual creatures.

Research from 3M corporation has found that we process images 60,000 times faster than text. This might explain why we find visual data is more appealing and attractive:

simply because we can understand it quicker.


Image 1

This might also explain the increasing number of data journalism we see everywhere we go, whether it’s on TV, social media, and even newspapers. The emergence of data journalism certainly, has not been ignored by journalists or even amateur bloggers.

A staggering number of people and businesses are racing and competing against each other to make the best and most creative infographics that are appealing to the audience.

However, often, at the expense of credibility and accuracy.

As discussed in the previous blog post, there are some problems associated with infographics and data journalism. Fisher’s ‘map of the world’s most and least racially tolerant countries‘ can perhaps serve as a perfect example of how data journalism are often flawed and misleading, yet, it is blindly accepted and believed by millions of people in a heartbeat.

The fact that colours and designs have more impact on people’s perception of messages, rather than the actual credibility of the data source, says a lot about the issue of interpreting infographics.


Image 2

So how can this issue be solved?

First of all, it is important for anyone that create infographics or data visualisation to disclose where the sources associated with their data and graphic are coming from, and more importantly, how their data/work should or should not be treated as scientific fact.

And secondly, by raising public awareness about the issue of accuracy in data visualisation, to prevent the spread of fake news or misinformation.

But how do people identify inaccurate/faulty data?

John Burns Murdoch came up with this list you have to check before believing in any data visualisation. It is not anything revolutionary, it is just the kind of thing that people can do mentally and automatically in their mind when seeing a data. If the data failed to check all the lists provided, then it is probably best to not trust the data.


The emptiness of Data Journalism

Data Journalism is a new trend of storytelling. Here, traditional journalistic methods are combined with data analysis, programming, and visualization techniques to create readable and interesting news stories. Some rock stars of this new trend include The Upshot of The new York Times, DataPoint of The Sydney Morning Heralds, The Guardian and The Vox.

However, not every journalist knows how to utilize the potentials of Data Journalism.


This is my favorite illustration! Thanks, Barry Blitt!

Every day when browsing through news on the internet, have you ever felt that, somehow, they are all the same? If journalists simply think of data journalism as a process of shuffling mountain of data and putting together a readable summary, then it’s easy to understand why we have to read hundreds of identical news every day. It’s all aggregation!

Many fans of data journalism believe that the data will show us all. And their mission is trying to answer every question using data, not anecdote. However, sometimes what the data tells us is biased rather than objective. In his article, Michael Kinsley argues that the nature of life is complicated and motives can be subtle.

If all data point to the conclusion that the earth is square, then will you say so in your article? The story of Nate Silver may make you rethink of this.

Nate Silver is one of the most famous data journalists in the US. He is also the founder of FiveThirtyEight website, which he describes as a regime of data journalism. Silver has been criticizing the value of traditional newspaper and television. And he honors only investigative journalism and data journalism. Given his data analysis, Nate Silver was down on Trump’s prospects throughout the US election. He even gave Hillary Clinton a 71.4 percent chance of beating Trump. That’s why many old-school journalists can’t wait to laugh at him after Trump’s victory.

Data journalism hillary

Talking about Silver’s failure, Timess media columnist, Jim Rutenberg concludes that “a good place to start would be to get a good night’s sleep, and then talk to some voters.”

And I totally agree with him.

Up with the facts! Down with the cult of facts!

Quantitative data should be an aid to journalistic practices, but it shouldn’t be a replacement for traditional journalistic methods. Such measurements are appropriate to some particular subjects. However, the assumption that it is appropriate to all subjects is not at all adequate. There is no numerical answer to the question of who deserve to receive food relief from the government, or whether gay marriage is acceptable. Interviews, opinion columns and all kinds of journalistic commentary need to be given equal honor to vary the voices and pave way for a free society.