A matter of stats

Statistics aren't synonymous with facts but communicating them correctly can bring us closer to the truth

I like alternative facts.

To me, if 8 out of 10 people are x then the alternative fact is that 2 out of 10 are y. The alternative fact is just the other side of the same coin. 

Knowing these alternative facts increases the breadth of my vision and my understanding of the world. As a communicator, it makes me better at my job and more socially-aware. If I was fifteen years younger, I'd probably be describing myself as ‘woke’. 

What I don't like are alternatives to facts.

Sometimes they originate in vacuums. Other times they fill a void or represent an underlying uncertainty or societal anxiety. As humans we have a propensity for storytelling and the ones that resonate with us are the ones that we share. Adding facts to these stories justifies, validates and verifies our position and assures us that we are right.

As a recent convert to the efficacy of focus groups, I have witnessed this first-hand. The most overwhelming take-away was the participants’ deference to readily accept information coming from those whose values they share – for example, a friend or a loved one.

The provenance of the information did not matter, only who shared it. 

This isn’t helped by the fact that many people consider statistics to be the ultimate fact. There is the quest and an innate desire to find the absolute truth and it is assumed that data provides this. It doesn’t.

At its best it is a snapshot in time and/or the best estimate that is available. There will always be flaws and limitations; nothing is ever perfect or unequivocally absolute. The question is whether it’s fit for its intended purpose.

And that’s the problem with being woke. Facts can be tactically deployed at an appropriate – and entirely subjective – time. Whilst it’s impossible to say this is categorically wrong, without appropriate context, it can be incredibly dangerous.

This is further compounded because data-literacy rates are generally poor.

What is too often missing is a user manual or guide for the correct application of each particular dataset, a health warning, if you like, that allows those who access it to know how to use it appropriately. It would also mean those that attempt to use it maliciously are held more publically accountable.

As communicators there are things we can and should be doing to help fight the good fight:

We should give our audience what they need to make the best decisions possible. The days of spin are over as far as I’m concerned. It’s a new age and citizens are unafraid to call out those in authority.

We should highlight long-term trends that are visible in datasets; not focus on erratic up or down movements over short periods of time. The volatile may present more of a story today but it’s rarely static and it may not still be true tomorrow.

International comparisons are also important in our increasingly global world. Utilised correctly, they frame the data and gravitas comes from presenting the bigger picture. If we as communicators can’t do this, who can?

It may sound tired but it should also be remembered that correlation does not imply causation. We’re wordsmiths, not number people, but that doesn’t mean we shouldn’t probe deeper. We should allow our inner-child to surface and ask ‘why?’ more often. There are often gems hidden in the depths.

Beyond our professional lives, as data-consumers we also have an obligation to question why the data matters. Does it have an agenda and, if so, what is it, who set it and why? There is rarely harm in curiosity.

And, ultimately, so what? What has changed as a consequence of knowing the information? Am I thinking differently, do I feel different or am I actively doing something different as a result?

As communicators, we take pride in our nous and there is always a reason something is being shared. By stopping and questioning that reason we can all be a little better at fact checking.

Finally, John Pullinger, the UK’s national statistician, recently wrote a seminal piece for the Guardian on echo-chambers and how in a post-truth world, statistics could provide an essential public service. I’d encourage everyone to read it.

The views expressed in this article are personal and do not necessarily reflect those of the Office for National Statstics or British Civil Service.


Connect with Martin on Twitter.

Image: Thinkstock