Have you ever considered what the best choice of numbers is to communicate an idea? Or why always following a hard and fast rule might end up being confusing? I recently came across an infographic that got me thinking about how important it is to choose carefully when to follow a rule, and when to go for something a little more … bespoke!
A very common approach when trying to communicate prevalence or risk is to express the quantity as a proportion. Often, a percentage is used, or the more general idea of ‘so many out of one hundred’. This leads to a very useful visual representation sometimes called a ‘waffle diagram’ – a one hundred grid colour coded to show proportional information. These can be useful to make comparisons of absolute and relative risks; for example, in this task from NRICH that explores the health risk of eating processed meats: How Risky is My Diet?
Adapted with permission of NRICH, University of Cambridge, from video 4 'How risky is my diet’ (2020).
Recently, when discussing my partner’s work as a midwife, we looked at a diagram together that represents the frequency of twin births in the population of the UK and noticed something very odd. The proportion was given not as a number out of 100, but out of 64, which got us thinking: Why the odd choice of numbers?
Used with permission from Gallimore I, Evans T, Kurinczuk J, Kenyon S, Draper E, on behalf of the MBRRACE-UK Collaboration, “Learning from deaths in twin pregnancies”, Leicester: The Infant Mortality and Morbidity Studies, Department of Population Health Sciences, University of Leicester, 2021.
In percentage terms, this represents 3.125% or 3.125 babies in every 100 that are twins. Clearly it makes sense to avoid using this decimal figure in relation to the very discrete quantitative unit of 'a human child' and so x out of 100 is no use. But why 2 in 64 rather than the perhaps more obvious choice of 1 in 32?
As a mathematician, I immediately focused in on the relationship between 2 and 64 as a power of two, and wondered whether this had some significance. It turned out that the number 2 was important but not in the way I thought – my partner observed the far more relevant feature of the piece of information here … twins come in pairs!
In fact, if you consider the information diagrammatically it becomes obvious that presenting the data as 1 in 32 would result in a far more abstract diagram, because only one of the twin babies would be visible. And there may be an uncomfortable implication that a further 31 births would need to take place before the second of the twins emerges noisily into the world, no doubt wondering what was the delay and can I please have some milk now!
Even so, with the more natural choice of 2 in 64, it still requires some stylistic flourishes to communicate the difference between the purple icons, each representing a single child born to a different parent, and the blue icon representing the twin birth with the twins visibly squooshed together, potentially representing the unfathomable urge to dress them up in creepily similar clothing.
This is a great example of where knowing the context of numbers leads to different decisions about how the quantities are expressed and the related stylistic choices that best communicate graphically the underlying message.
What do you think? Does the choice of 2 in 64 births make the prevalence of twins easier to understand? Do you have any alternative suggestion that would communicate the twin birthrate in the UK better? And can you think of other situations where the context dictates that a more effective representation than x out of 100 could be chosen? Let us know in the comments below.