How to make the world add up
March 2, 2021

How to make the world add up through more curiosity

By Paul Laughlin

In these turbulent times those seeking to understand what’s really happening often seek to make the world add up. That is they turn to numbers & statistics to give them a more objective or rational view.

However, as Tim Harford deconstructs in his latest book, such a solution can be beset with many traps & self-deceptions. We need to keep our wits about us. To understand better, rather than be lied to via statistics, requires focus. Regular listeners to the BBC’s More or Less podcast will recognise Tim’s engaging style & voice along the way.

Tim’s highly recommended book is entitled “How to make the world add up: Ten rules for thinking differently about numbers“. The majority of this engaging read explains 10 commandments or rules. These are ‘watch outs‘ to help us better critically consider new numbers/data/statistics.

How this book helps leaders of all flavours

The first encouraging aspect of this book is the motivation or heart that shine through Tim’s storytelling. He writes like a man who cares both about the world & about how statistics can be used for good within it. You feel like you are on a quest fighting to make the world a better place. To achieve more truth-telling with numbers & less deception.

This is obvious from the start when Tim introduces readers to the infamous How to Lie with Statistics” by Darrell Huff. Despite highlighting what can learn from this runaway success, Tim also shines a light on the darker side. On how the dark arts of statistics have been used by tobacco companies, those profiting from environmental damage & pandemic deniers to cause real harm.

So begins what feels like a moral quest. Not to be amused by how gullible people can be misled via numbers or graphs, but to improve all our thinking. It is also why this book is suitable for such a wide audience. In a business context, it would help all leaders to improve their critical thinking. It is also a useful reminder to data & analytics leaders of the pitfalls they need to avoid daily. More for that community later.

10 commandments to help you make the world add up

So what of those 10 commandments I mentioned? They are both pleasingly simple and also helpfully eclectic. Covering both the unconscious biases made famous by Daniel Kahneman et al, data considerations (like those highlighted by David Spiegelhalter) and social/cultural norms we need to resist. Here is a brief summary:

  1. Notice your emotional reaction (avoid bias towards what you like)
  2. Consider both the statistical big picture & your personal experience (sense checking by seeing both the wood & the trees)
  3. Ask yourself if you really know what the data represents (clarity on definitions & metadata, exposing misconceptions )
  4. Sense check with comparisons & context (put any new claim in perspective, including having a sense of scale as Nick proposed)
  5. Find out how data was captured (and what might be missing)
  6. Consider who is missing (any potential systemic biases/exclusion)
  7. Demand transparency about big data & algorithms (a great chapter for data scientists to read, a plea for more scientific method)
  8. Value your statistical bedrock by using your national statistics agency (with some moving stories of statisticians who took a stand)
  9. Notice & avoid being lied to by charts/graphs (with references to the great “How charts lie” by Alberto Cairo to help you)
  10. Keep an open mind (How might you be mistaken? How might the facts have changed since you last checked?)

These are a great collection that covers a broad range of attitudinal, behavioural and mental actions that are worth making into habits. As Tim rightly highlights, we can’t be expected to go through the full checklist each time. But, periodic review is worthwhile to help build greater awareness & usage over time.

One simple solution & a plea to data leaders

To help avoid overload, Tim closes his book by simplifying these 10 commandments into one attitude. Be curious. He shares the results of scientific studies that have proven in multiple contexts that if people are more curious they are less trapped in their existing prejudices. They are more able to consider new data and better understand others.

It is a great way to close the book. It prompts me to warn data & analytics leaders of another theme throughout this book. Evidence that more knowledgeable or technically skilled people are more susceptible to their existing biases & seeing only what supports their own views. From my own experience, I’d agree this is true of data & analytics leaders too.

As a community that would recommend others improve their statistical thinking (read Spiegelhalter) or graphicacy (read Cairo), data leaders need to beware of hypocrisy. Being so close to data & analytics day in day out we can become to close & even emotionally committed to our own internal models & hypotheses. We can think we understand our business or customers & just need to educateothers‘ who are less data-led.

As a community, let’s have the humility to recognise these 10 rules are for us too. How could you start to practice considering the above pitfalls when you need see new data or statistics you are tempted to reject?