Believe half of what you see and none of what you hear

Fake News and Clickbait are subjects which have generated countless articles, blogs and discussions and are rarely out of the news. How to differentiate between what is real and what is fake (or at least exaggerated) is a useful skill in all areas of life.

This post is about reflecting on how you should apply critical thinking (and a bit of scepticism) within your data work in the same way you would these viral stories.

Recently there was a video circulating which supposedly showed a statue of the Virgin Mary floating down a street during a flood in 'Venice'.

The real version, while a bit less divine, is still worth watching:

You may say that the floating Virgin Mary was obviously a fake (the video had an el supositorio watermark which should have provided another clue even if your Spanish isn't great). The data equivalent of this may be some of the 'creative' data visualistions from the Liberal Democrats which often have little to do with good chart design.

Sometimes the misdirection is less obvious, and can even be well intentioned with this video going viral. The video below shows what appears to be a dramatic rescue of someone about to throw themselves in front of a train, but in reality it is a well edited piece splicing the train and people together from separate shots.

Compared to the Virgin Mary video, it's a lot less obvious it isn't real, although there a few pointers such as no obvious effects of turbulence on the people/papers etc., but also this happens to be my local train station (Tring) so I know that platform is rarely used for departures and also that the stairs down to the platform are further up the platform so the woman who 'saves' the man has no reason to be at this part of the platform where trains would never stop.

A data equivalent of well meaning but misleading was a highlighted in a tweet on how apparently inequality is far worse in UK than other countries that went viral in part due to Jeremy Vine but was correctly called out by an FT Journalist.

The more detailed Full Fact explanation is here and worth reading but essentially The Economist chart is flawed as it compares more granular (and therefore more unequal) levels of region for the UK compared to other countries.

Inequality is obviously real and it's admirable that people want to highlight it, but doing it through misleading data isn't the right way to go about things.

A similar thing can happen within your own analysis, if you're tasked with evaluating the efficacy of a project for example: You go off and code it all up and if, when you've produced some results, things look positive, you're far less likely to double check and do a thorough QA than if the results were negative.

In an ideal world, we'd be as critical about good things as we are about things we consider bad. In one of our recent events, I talked about using the Paxman approach of 'Why is this bastard lying to me' when looking at any analysis (including your own).

On the same theme, the tweet below came up in my feed which has a similar message, ultimately critical thinking and scepticism are two things you should never be short of in any facet of your life.