Using Context and Comparison to Enable Meaning
Let’s begin by considering a study in which researchers asked two groups of people to rate the attractiveness of a simple gamble. They asked Group A to consider a gamble in which they were given a 7/36 chance of winning $9. They asked Group B to consider a similar gamble, but added the potential for a small loss—that is, Group B’s gamble gave them a 7/36 chance of winning $9 or, conversely, a 29/36 chance of losing 5¢. Which group do you think rated their option as most attractive?
Although Group A’s gamble was clearly better than Group B’s, people actually rated the gamble with some potential for a small loss as more attractive. The mean response to the first gamble was 9.4, while in the gamble that added a chance of a small loss, the mean was 14.9. 
Why would the adding the potential for a small loss make such a difference? Researchers concluded that, in the gamble with no potential for loss, people focused primarily on their probability of winning rather than on the possible monetary outcome. This is because probability is easy to assess—there are known upper and lower bounds that provide context, and in the example of this gamble, the probability of winning is relatively unattractive—only 7 out of 36 chances of winning.
The $9 monetary value of the gamble is harder to assess. Is winning $9 good or bad? Attractive or not? Adding the chance of a small loss puts the $9 payoff in perspective and gives it meaning. The value of $9 was much clearer and more positive in the presence of the 5-cent loss. The combination of a possible $9 gain versus a 5-cent loss is an attractive win/lose ratio—that is, it offers a good-value win versus a very small loss.
Remember, as I’ve described in my previous columns, that people determine value through a process of comparing things. It’s not so much that people can’t comprehend the value of $9. The real issue is: what does $9 really mean in the context of this gamble? Is a $9 payoff good or bad? Without the context of a reference point—in this case, a 5-cent loss—it’s hard to judge the worth of the $9.
Numeric Data and Decision Making
Research shows that the ease with which people can derive meaning dictates the extent to which they will use data. For people to use numeric data effectively, they need to be able to do three things:
- Comprehend it.
- Interpret it—and attach the right meaning to it.
- Act on it.
People must be able to successfully comprehend and attach meaning to data before they can act on it. UX designers can play a critical role in enabling this process.
Rote Frequencies and Decision Making
Let’s take a look at some additional research that can help us understand the nuances of how data format affects judgment and decision making.
In one study, researchers gave a group of people a chance to win a prize by drawing a red bean from a bowl containing red and white beans.  There were two bowls from which they could choose. In one bowl, there were 100 beans, 7 of which were red. In the other bowl, there were 10 beans, one of which was red. A significant number of people wanted to choose from one bowl rather than the other. Can you guess which bowl was most popular?
Although they actually had a better probability of winning if they chose from the bowl with fewer beans, many people preferred to choose from the bowl with more beans. Why? Because the image of seven winning beans was more attractive than the image of one winning bean. The larger bowl looked more inviting because it contained more red beans and seemed to give people more chances of winning.
Notice what people are doing here. They are focusing only on the number of winning beans and ignoring the rest of the context—the remaining white beans in each bowl—as they think about which bowl from which to choose. Gut—which makes an immediate decision prior to Head’s weighing in—strongly influences this mental shortcut. Remember, from my previous columns, that Gut is attuned to basic elements of contrast—such as the difference between big and small or light and dark.
Another study found a similar effect when researchers asked people to judge which disease is more dangerous—one that kills 1,286 out of every 10,000 people or one that kills 24 out of every 100 people.  Can you predict how most people responded? Results of the study showed that the degree of perceived risk varied according to the number of deaths researchers presented—irrespective of the total number of possible deaths.
One practical application of this finding is that, since many people judge the level of risk according to rote frequency, the format in which you present numbers can affect people’s perception of the level of risk. In general, higher rote frequencies result in higher perceived risk and vice versa.