Recent research has shown that we feel that outcomes classified into large categories are more likely to occur than outcomes classified into small categories. Turns out that this is because we incorrectly think that individual outcomes take on properties of their overall category.
Imagine for a moment that you’re in a casino, standing at the roulette wheel. With one ball in the croupier’s hand, you’ve one chance to win BIG…
On the wheel, there are 18 black pockets but only 2 green ones. Now, how much more likely are you to land a specific black pocket over a green one?
Despite the fact that we’ve a completely equal chance of landing on either, many make the mistake of thinking that success will be more likely with the black choice than the green. This is because there are more black pockets (18) than green ones (2). Our judgment on the probability of landing on a ‘Black two’, for instance, is affected by the category it sits within. Black’s a bigger category than green, so we assume it to be the more probable of the two.
Put more broadly, we mistakenly believe items in larger categories to be more likely to occur than ones in smaller categories. This, dear friends, is known as the Category Size Bias, and it affects all sorts of things.
Let me tell you how you can use it to nudge people to be healthier, to spend more, and be more likely to take action.
We think that when outcomes are put into relatively large categories, they’re more likely to occur.
In a recent study by Isaac & Brough (2014), 223 people were asked to play a simple lottery, of which there were two versions, each with 15 balls.
The first contained 5 black, 5 grey and 5 white balls.
The second contained 2 black, 11 grey and 2 white balls. You can see here in the second version that the grey is a relatively large category.
Despite the fact that the chance of each ball being plucked from the lottery bowl being equal, people were then asked to estimate the probability of a grey number 8 ball being chosen.
And the answers were crazy! People judged chance of choosing the grey 8 ball to be higher in the lottery with 11 grey balls, than the one with 5 grey balls. Put another way, because there were more grey balls than black or white, they thought that the grey #8 ball was more likely to come up.
We’ll spend more money on items put in larger categories.
Another experiment from the research involved a glass urn filled with 90 coloured tickets. 81 were blue, and 9 were yellow.
81 students were given one of these blue or yellow tickets, and asked both how likely it was that their ticket would be chosen. They also were asked how much they’d like to bet.
The results showed that those given blue (large category) tickets bet 24% more than those given yellow tickets. They also thought the probability of their ticket being chosen would be higher (24% for blue vs 10% for yellow), despite the fact that each ticket has an equal chance of being chosen (1.67%)! Crazy!
Clearly, this research shows that we feel more confident in choosing something if it’s put within a relatively-large category, and are more likely to take higher risks. Gambling companies take note!
The more clearly you highlight the difference in category size, the more likely we are to choose something from the larger category.
In a second study, 175 people were asked to guess the probability of a 26-sided alphabet die landing on an ‘A’ (part of the small vowel group) or a ‘T’ (part of the larger consonant group).
The researchers wanted to see whether actively communicating the size of the two groups to the people would have any effect on their judgements. And boy did it!
They tested for 3 levels of information: high, medium and low.
They found that when actively mentioning group size: that T was one of 21 consonants in the alphabet, and A was one of 5 vowels, people were much more likely to think that a T was more likely to be rolled than an A. This is despite the fact that both outcomes have exactly the same chance of happening! Crazy!
The graph below shows you just how much merely communicating category size can affect our decisions.
We are more likely to take action from tasks when they’re in a bigger list, over a smaller list.
One more study from the paper that we’d love to share tested a group’s likelihood to take safety measures against an IT threat. They were given a list of 9 tasks that would protect them, such as “Use a pop up blocker” or “Change your password frequently”.
They were then asked to these 9 tasks into either one of two categories: “identity theft” or “loss of data”. One category would need to end up with 7 items (large), and the other with 2 items (small).
The results of this study are extremely interesting. In the 3 months following the exercise, participants were more likely to engage in those preventative behaviours that they’d classified into a large (vs. small) category of tasks.
And the reason for this greater willingness to act on advice given? People simply felt that the IT threat associated with a large number of preventive behaviours would be more likely to happen than one associated with fewer of them.
In conclusion, this fascinating new research shows us that something as simple as creating and sizing categories has a huge potential to motivate long-lasting behavioural change.
Judgment under Uncertainty: Heuristics and Biases PDF (Tversky & Kahneman, 2007)
Judging a Part by the Size of Its Whole: The Category Size Bias in Probability Judgments PDF (Isaac & Brough, 2014)