Variations on Practiced Patterns Cause Mistakes

“You can’t teach an old dog new tricks.” It’s debatable whether this adage is entirely accurate, but it certainly is difficult to unlearn a long-practiced habit.

Recently, I began to attend dance fitness classes by a new instructor to mix things up. And boy, did I get mixed up! While I was prepared to learn entirely new dance routines — I greatly enjoy this dance equivalent to sight-reading music — I was not expecting the struggle of learning new routines for familiar songs. When the new instructor played a song that my old instructor had used before, I was forced to focus, fight my habitual routine, and learn the new steps. To make matters worse, the new routine had several steps in common with my old one here and there, but then deviated. It took great concentration to unlearn my usual pattern and follow the new one. This experience reminded me of how human memory works, and the dangers of varying a well-practiced pattern.

Familiarity and Learning

Familiarity is most often a good thing — it allows us to draw on past experiences to inform current interactions, and make fast decisions without much effort: a well-practiced task requires fewer cognitive resources to be completed successfully. That’s how expertise works: studies with chess masters have shown that they are really good at recognizing successful patterns they had encountered before and planning their movements based on these patterns, whereas a novice will need a lot more effort to figure out the consequences of each potential move.

Like chess masters, we are experts at using the web: we’ve seen many patterns repeat and we’ve learned to recognize them. Whenever we encounter a familiar design, we can free some of our cognitive resources toward more important activities. That’s why users are averse to changes in interface designs: it is more efficient to use a system we already know.

Familiarity is acquired through repetition. Like all human learning, learning a new interface or a new interaction pattern follows a power law: it requires many repetitions before people will become as fast and efficient as with an old, much practiced interface.

Jakob’s Law of Web User Experience says that users spend most of their time on other websites, not on yours. They have countless opportunities to learn and practice established patterns. When you depart from those well-known standards, you force them to spend effort to understand the interface and how it is different from what they’ve seen before.

Expectations Drive Attention

When people think they recognized a pattern they’ve encountered elsewhere on the web, their mental model for that interaction drives their behavior. If your particular implementation of the pattern is different in any way from the standard, people will often be lost or confused. It’s as if they have “tunnel vision”: they are highly focused on the areas of the screen that they think, according to their experience-based mental model, should contain those pieces of information relevant to their current task. Exhaustive review, in which the eyes return repeatedly to an area of the screen where users expect the right content to be, is an example of this behavior. The opposite behavior is called the hot-potato scanning patter — the eyes purposefully avoid a screen area and whenever they accidentally fixate on element inside that area immediately move away. The hot potato is also an effect of mental models informing users’ actions: users expect that no interesting information will be located in that area.

Banner blindness is an example of hot-potato behavior — anything that remotely looks like an ad will be ignored by users, and items in a web page’s right rail will be avoided simply because ads are usually placed there. Users’ expectations are informed by their mental models, which are created through repetition. Thus, when you design an interface, it is important to understand users’ mental models and expectations and support them, so people could rely on those shortcuts that they’ve formed in their memory and perform their tasks efficiently.

Violated Assumptions Lead to Mistakes

When users dismiss areas of the screen because — based on prior mental models — they have the wrong assumption about the functionality in those areas, major usability issues can arise.

In our recent usability studies, we asked participants to use the Starbucks site or mobile app to find the nearest Starbucks that had certain amenities, such as a drive-through, or that offered handcrafted sodas. While completing this task, 3 out of 5 users did not notice the available filter on the store-locator screen, and instead felt that they were forced to navigate to each store detail page to review its list of amenities. Needless to say, this process was time-consuming and not very pleasant. Those participants complained that there was no way to narrow down their results: “I can’t figure out a way to do this except look at the actual stores to see if they do anything … I’m not finding it, other than doing this and looking through each one.”

Although these users complained that there was no way to narrow down store results, a Filter function was actually available, but was hidden in plain sight. In fact, users had looked at and even used the Search Stores field immediately next to the Filter button, but still did not notice that the filter was available. The design and placement of the Filter button caused users to assume it was a search-submit button, and so they never attended to the actual button to realize that their assumption was wrong.

In most digital interfaces, a button to the right of a search field is the Submit button. Because this design mimicked the dance they had often performed, people assumed the steps would be the same.

Repeated Routines Cause Repeated Mistakes

It is very difficult to alter a frequently practiced routine even after a mistake has been made once — or even multiple times.

One participant in our mobile eyetracking study repeatedly made the same error: she accidentally closed her browser tab because her Chrome app had recently changed its design, and now displayed a prominent Close icon that looked like a typical Close button for a modal window. The placement of the Close icon in the top left corner of the screen is also often used for back buttons in many mobile apps, and so the participant ended up losing her place multiple times as she continued to close her browser tab when attempting to go back to an earlier screen.

Although she understood her mistake, and realized how the new design was intended to function, tapping the top left area of the browser chrome to move back to a previous page is such a strong, standard interaction pattern that she continued to repeat this action multiple times even though it did not yield the results she intended. The more we have practiced a certain process, the more difficult it becomes to change — we become the old dog that can’t learn new tricks.

Conclusion

When designs deviate from convention, users may not be able to bridge the Gulf of Execution to see how to approach their tasks in the interface, and thus will make mistakes. Especially during well-practiced processes, they allocate fewer attentional resources and can easily slip and perform the wrong step without realizing it.

The biggest danger exists when users make mistakes or overlook (through selective attention) critical aspects of the interface and do not realize it — and hence have zero chance of correcting the mistake or altering their behavior. It requires too much effort to thoroughly assess every design encountered throughout the day to determine whether it indeed follows our known, practiced pattern or if it deviates from the norm — which is why it is up to us, the designers, to be aware of users’ mental models and design thoughtfully, to understand what people have practiced and either follow it to its entirety or be sure to not mimic it in any way to prevent confusion. Because, just like fighting against following a known dance routine to accommodate new steps, focus and effort are required to recognize deviations and conform to a new path.

Learn more about memory, learning, and their effects on design in our full-day training course The Human Mind and Usability.