It's increasingly difficult to deny that the [mis]use of statistics has contributed to the divide in many modern, developed societies
Summary statistics rarely describe somebody's lived experience and never ring true for the whole population
When a summary statistic does not apply to you...
At best, it feels irrelevant
At worst, it actively reminds you that you're on the "losing" side of society
In their worst, most top-down form, statistics are un-democratic. They are a pronouncement from on high that the people should act and feel in accordance with an intangible number, even if that number and its implications are a world away from what they experience every day
So how can data and statistics not only find their feet again, but play a key role in bridging the gap they have helped create?
Anscombe's quartet is not just a nerd's party trick, it highlights a very real problem
Income levels in the UK
In April 2016, median gross weekly earnings for full-time employees were £539
In April 2016, median gross weekly earnings for full-time employees were £539
We can do better than that, and it's important that we do better than that
Taking account of housing costs
Do you earn the same amount as someone hundreds of miles away?
Break it down by region
How about someone the other side of town?
We can still do better
Show the full distribution wherever possible, as this demonstrates the full range of circumstances experienced by different people, which in turn:
Try to use metrics that are truly meaningful for your audience
Be empathetic in your use — and interpretation — of statistics
Think about the local context
Don't stop at the simple answer: dig deeper than summary statistics. Sometimes it's only at this stage that the answer reveals itself
You can reach me on Twitter @jburnmurdoch or by email at john.burn-murdoch@ft.com
These slides are at bit.ly/data-divide