When people look at interfaces, they want to grasp the gist of something, they don’t want to do a statistical analysis. Always keep values as imprecise as possible.
As a designer, it is my job to communicate information in a straightforward and concise way. However this is sometimes in conflict with what other team members, e.g. project managers, seem to want.
For instance, we recently designed a screen to configure the heaters in your smart home. At one point we needed to decide how accurate the information needed to be and as always there was this one person who brought up a really specific edge case, arguing for the information to be super precise. Now as designers, it is your job to intervene and defend the user’s right to a simple and easy-to-use interface.
So What’s the Problem With Precision?
Somehow we convinced ourselves that high precision is good, while less precise information is sloppy. But is that correct in the context of interface design? While precision provides more information, it also requires people to process the additional information.
In the case of the heating example from above, I believe users do not care about the exact number of degrees, they only care if it is warm or not, where “warm” might mean 21°C to one and 25°C to another. Moreover, when users remember their individual “warmth” temperature, they are unlikely to remember 24.5°C, but rather “close to” 25°C.
Fewer digits are not only easier to remember, they are also easier to adjust. Whole digit increments are fairly easy to process but the more decimal points you add, the harder it is to make a quick decision.
Therefore you should always keep things as simple as possible.
But How Do I Choose the Right Degree of Precision?
Sadly there is no one right way when it comes to precision, as it highly depends on the context, your audience, and the kind of data you display. While there are some cases, for example when designing an atomic clock, the highest level of precision is desirable, but there are many other cases, like heating, where a less precise value is much better suited. The best way to determine the degree of accuracy needed is to put yourself in your user’s position. A user always comes to your interface with a specific goal, which is best described in a user story, for example:
As a user of your product, I want to be able to adjust the temperature of my heater so that I feel comfortable.
Starting from the user’s need we can evaluate the best way to achieve his goal. Since “comfortable” is a very individual degree point, the user needs to be able to adjust the temperature individually. Something like a 5° step solution would make it easy to adjust, but considering preferences and environmental effects, this is not precise enough. The next level of precision is degree steps, which allows people to set a very precise temperature depending on their personal preferences. We can rule out half degree steps because people don’t notice a half degree difference in temperature. Also offering half degree steps would double the necessary steps required, meaning more effort to set a specific value. The least accurate we can be in this situation is using degree steps. We should not get more precise than that.
In short: Never be more precise than necessary
While it is important to provide accurate information, you should always remember that it is equally important to provide an easy-to-use interface. Therefore you should only be as precise as necessary, not as precise as possible.
