Approaching Statistical Claims

Statistics by Nick Youngson - Alpha Stock Images
Statistics by Nick Youngson - Alpha Stock Images

The Financial Times discussion on Statistical Claims between Sarah O’Connor and Tim Harford lists various ways in which to approach statistical claims. The first, observe feelings, suggests that emotions and internal mindsets alter both the way we interpret statistics and whether we share them. Taking feelings into consideration allows us to eliminate any bias we may have and view the claim as impartially as possible. The second step, understanding the claim, suggests that the more insight we have into the context and background information of the statistics – the easier the claim will be for us to interpret and build our own opinions around. Harford then suggests “distinguishing between correlation and causation” as another step to approaching these claims. Understanding which of the two the claim is based on gives the reader an insight as to the idea the data is trying to convey and the reason for those statistics to have been used. “look out for what is left out”, Harford encourages listeners to be both proactive and aware when it comes to statistical claims. The audience are also advised to “get a grip on the back story”. Look into where the statistics came from and how they managed to catch your attention. Understanding that social circles tend to impact the statistics/information we are exposed to. Circles tend to share the same opinion thus sharing similar information that they would side/agree with. The last method given to the audience is to “be curious”. When it comes to statistical claims, people are encouraged to research quotes, claims and sources and find out more for themselves to ensure that the statistics do indeed support the claims they are placed alongside.  

Personally, I agree with most of the points that Harford makes. I feel as though the way in which we see the world and the views we share/believe do impact whether we are likely to share or believe a statistical claim. I also believe that as an audience, we should be very aware of what information is missing and why particular information has been used in datasets and other pieces of information haven’t. 

When it comes to myself and my approach to these claims, I have noticed that I tend to be very passive and just believe the claims presented. I feel as though this is why I have been misinformed and wrong on statistics on so many occasions in the past. 

After listening to the podcast, the way in which I took in statistical claims also changed. On reading a BBC News article on the increase of pension contribution in the UK, I saw a claim that stated, “Since 2012, 10 million eligible workers have been automatically signed up to workplace pensions”. I then began to do some research on the claim, looking into how many of these workers had been signed up since the last increase in pension contributions and how many people had opted out of the scheme since the last increase. Something I never would have done before listening to the podcast.