Lies, damned lies, and statistics

Do you find it odd that the whole world suddenly seems to be talking about data? That TV, radio and Twitter are full of debates about daily trends in Covid-19 cases and moving averages in Presidential opinion polls? When did outliers and sampling error become the topics of conversation around kitchen tables?

Does it delight you that discussions about data have become more prominent?

Or do you worry that in an era of fake news, the use of data to solve problems is getting a bad reputation and that many people's instinctive reaction to the latest headlines now echoes Mark Twain's famous line: "There are three kinds of lies: lies, damned lies, and statistics"?

Getting a grip on statistics

This month the Sunday Times in the UK published an article by Tim Harford called Get a grip on statistics and you can illuminate your world. It's an excellent piece, arguing that far from dismissing statistics, the general public should embrace them, but carefully, following a set of principles to ensure that the data we rely on adds light to a subject rather than heat to a debate. Harford's 6 principles are:

  1. Get the back story: evaluate the source of the data and the reason it has been published before using it to draw conclusions
  2. Look at the words, not just the numbers: we should always examine definitions because the data may make no claim to solve the problem to which they are being applied
  3. It's not the sample size, it's what you do with it: it's easy to be impressed by the size of a sample and ignore the extent to which it may not be representative
  4. Correlation not causation: watch the evening news and write down how many times pundits, presenters and claimed experts fail to distinguish between the two
  5. Notice your own feelings: we all have a strong bias to believing data that confirms our own beliefs and trouble accepting the validity of data that does not fit our narrative
  6. Don't give up on statistics: despite dodgy numbers and dubious interpretations, good data can reveal fundamental truths and help us to solve difficult problems

Developing an Insight perspective

The fact that articles like these are appearing in national newspapers - and I would recommend that you read this one - is a real sign of the times. But whilst the public interest in statistics, evidence and the interpretation of data may feel new, it is a daily part of our work as corporate Insight professionals. 

Hans Rosling, the famous Swedish statistician and author of Factfulness, made it his life's work to "fight devastating ignorance with a fact-based worldview" and there are times when I think that would be a pretty good mission statement for many Insight teams. Our organisations are full of good people trying to do the right thing, but impeded by widespread ignorance and competing data.

Amongst other things, it is our role to build a fact-based worldview of how consumers in our market can become value-creating customers of our organisation; to dispel myths, and promote fundamental truths. But along the way we need to make sure that we ourselves follow the same principles that Tim Harford recommends to his readers: evaluating evidence, questioning assumptions and treating data with the respect that it deserves.

This is what the Insight Management Academy refers to as "developing an Insight perspective". It's an essential attribute, and one that our organisations and wider society should learn to prize more highly.

3 ways to learn more

If you are interested in this topic, you could:

  • Read Factfulness by Hans Rosling
  • Read the IMA's online guide: IMP604: How to develop an Insight perspective - please click here for more details
  • Book places on the IMA's Transforming Insight programme of online workshops, in particular:

 - Insight generation part 1: the consultant - please click here

 - Insight generation part 2:  the detective - please click here


James Wycherley

Chief Executive, IMA