Chapter 3. Hick’s Law

The time it takes to make a decision increases with the number and complexity of choices available.

Key Takeaways

• Minimize choices when response times are critical to increase decision time.

• Break complex tasks into smaller steps in order to decrease cognitive load.

• Avoid overwhelming users by highlighting recommended options.

• Use progressive onboarding to minimize cognitive load for new users.

• Be careful not to simplify to the point of abstraction.

Overview

One of the primary functions we have as designers is to synthesize information and present it in a way that doesn’t overwhelm the people who use the products and services we design. We do this because we understand, almost instinctively, that redundancy and excessiveness create confusion. This confusion is problematic when it comes to creating products and services that feel intuitive. Instead we should enable people to quickly and easily accomplish their goals. We risk causing confusion when we don’t completely understand the goals and constraints of the people using the product or service. Ultimately, our objective is to understand what the user seeks to accomplish so that we can reduce or eliminate anything that doesn’t contribute to them successfully achieving their goal(s). We in essence strive to simplify complexity through efficiency and elegance.

What is neither efficient nor elegant is when an interface provides too many options. This is a clear indication that those who created the product or service do not entirely understand the needs of the user. Complexity extends beyond just the user interface; it can be applied to processes as well. The absence of a distinctive and clear call to action, unclear information architecture, unnecessary steps, too many choices or too much information—all of these can be obstacles to users seeking to perform a specific task.

This observation directly relates to Hick’s law, which predicts that the time it takes to make a decision increases with the number and complexity of choices available. Not only is this principle fundamental to decision making, but it’s critical to how people perceive and process the user interfaces we create. We’ll look at some examples of how this principle relates to design, but first let’s look at its origins.

Origins

Hick’s law was formulated in 1952 by psychologists William Edmund Hick and Ray Hyman, who were examining the relationship between the number of stimuli present and an individual’s reaction time to any given stimulus. What they found was that increasing the number of choices available logarithmically increases decision time. In other words, people take longer to make a decision when given more options to choose from. It turns out there is an actual formula to represent this relationship: RT = a + b log₂ (n) (Figure 3-1). This formula calculates response time (RT) based on the number of stimuli present (n) and two arbitrary measurable constants that depend on the task (a, b).

Fortunately, we don’t need to understand the math behind this formula to grasp what it means. The concept is straightforward when applied to design: the time it takes for users to interact with an interface directly correlates to the number of options available to interact with. This implies that complex or busy interfaces result in longer decision times for users, because they must first process the options that are available to them and then choose which is the most relevant in relation to their goal. When an interface is too busy, actions are unclear or difficult to identify, and critical information is hard to find, a larger amount of brain power is required to find what we are looking for. This leads us to our key concept for Hick’s law: cognitive load.

Figure 3-1. Diagram representing Hick’s law

PSYCHOLOGY CONCEPTCognitive Load

When engaging with a digital product or service, a user must first learn how it works and then determine how to find the information they are looking for. Understanding how to use the navigation (or sometimes even finding it), processing the page layout, interacting with UI elements, and entering information into forms all require mental resources. While this learning process is happening, the user must also maintain focus on what they intended to do in the first place. Depending on how easy an interface is to use, the latter can be quite a challenge. The amount of mental resources needed to understand and interact with an interface is known as cognitive load.

You can think of it like memory in a phone or laptop: run too many apps and the battery begins to drain and the device slows down, or worst of all, it crashes. The amount of processing power available determines performance, and this depends on memory—a finite resource.

Our brains work similarly: when the amount of information coming in exceeds the space we have available, we struggle mentally to keep up—tasks become more difficult, details are missed, and we begin to feel overwhelmed. Our working memory, the buffer space (Figure 3-2) available for storing information relevant to the current task, has a specific number of slots in which to store information. If the tasks at hand require more space than is available, we begin to lose information from our working memory to accommodate this new information.

Figure 3-2. Working memory buffer illustration

This becomes problematic when the information lost is critical to the task that someone wishes to perform or is related to the information they want to find. Tasks will become more difficult and users might start to feel overwhelmed, ultimately leading to frustration or even task abandonment—both symptoms of a bad user experience.

Examples

Now that we have an understanding of Hick’s law and cognitive load, let’s take a look at some examples that demonstrate this principle. There are examples of Hick’s law in action everywhere, but we’ll start with a common one: remote controls.

As the number of features available in TVs increased over the decades, so did the options available on their corresponding remotes. Eventually, we ended up with remotes so complex that using them required either muscle memory from repeated use or a significant amount of mental processing. This led to the phenomenon known as “grandparent-friendly remotes.” By taping off everything except for the essential buttons, grandkids were able to improve the usability of remotes for their loved ones, and they also did us all the favor of sharing them online (Figure 3-3).

In contrast, today we have smart TV remotes: the streamlined cousins of the previous examples, simplifying the controls to only those that are absolutely necessary (Figure 3-4). The result is a remote that doesn’t require a substantial amount of working memory and therefore incurs much less cognitive load. The complexity is transferred to the TV interface itself, where information can be effectively organized and progressively disclosed within menus.

Figure 3-3. Modified TV remotes that simplify the “interface” (sources: Sam Weller via Twitter, 2015 [left]; Luke Hannon via Twitter, 2016 [right])

Figure 3-4. A smart TV remote, which simplifies the controls to only those absolutely necessary (source: Digital Trends, 2018)

Now that we’ve seen some examples of Hick’s law at work in the physical world, let’s shift our focus to the digital. As we’ve seen already, the number of choices available can have a direct impact on the time it takes to make a decision. We can ensure better user experiences by providing the right choices at the right time rather than presenting all the possible choices all the time. An excellent example of this can be found with Google Search, which provides the varying means of filtering results by type (all, images, videos, news, etc.) only after you’ve begun your search (Figure 3-5). This helps to keep people focused on the more meaningful task at hand, rather than their being overwhelmed with decisions at the outset.

Figure 3-5. Google simplifies the initial task of searching (left) and provides the ability to filter results only after the search has begun (right) (source: Google, 2020)

Let’s take a look at another example of Hick’s law. Onboarding is a crucial but risky process for new users, and few nail it as well as Slack (Figure 3-6). Instead of dropping users into a fully featured app after subjecting them to a few onboarding slides, a bot (Slackbot) is used to engage users and prompt them to learn about the messaging features in a risk-free way. To prevent new users from feeling overwhelmed, Slack hides all features except for the messaging input. Once users have learned how to message with Slackbot, they are progressively introduced to additional features.

Figure 3-6. Screenshot from Slack’s progressive onboarding experience (source: Slack, 2019)

This is an effective way to onboard users because it mimics the way we actually learn: we build upon each previous step and add to what we already know. By revealing features at just the right time, we can enable our users to adapt to complex workflows and feature sets without feeling overwhelmed.

TECHNIQUECard Sorting

As we’ve seen in the previous examples, the number of choices can have a critical impact on the time it takes for people to make a decision. This is especially important when it comes to enabling users to find the information they need. Too many items can lead to more cognitive load for users, especially if the choices aren’t clear. Conversely, with too few choices it becomes more difficult for them to identify which item is the most likely to lead them to the information they’re looking for. One particularly useful method for identifying users’ expectations when it comes to information architecture is card sorting. This handy research method is great for figuring out how items should be organized according to people’s mental models: simply have the participants organize topics within groups that make the most sense to them (Figure 3-7).

Figure 3-7. Card sorting

The steps required during this exercise are relatively straightforward. While there are a variety of approaches to card sorting (closed versus open, moderated and unmoderated), they all follow the same general process. The following are the steps that make up a moderated open card sorting exercise,1 which is the most common type:

1. Identify topics. The first step is to identify the topics that the participants will be asked to organize. These topics should represent the main content within your information architecture, with each item written on an individual card (the exercise can also be conducted digitally). It’s recommended that you avoid labeling topics with the same words, which can bias participants and lead them to group these items together.2

2. Organize topics. The next step is to have the participants organize the topics, one at a time, into groupings that make sense to them. It’s common to have participants think out loud during this phase, which can provide valuable insight into their thought processes.

3. Name categories. Once the topics have been organized into groups, ask the participants to name each group they created based on the term they think best describes it. This step is particularly valuable because it reveals what each participant’s mental model is and will be helpful when determining what to eventually label categories within your information architecture.

4. Debrief participants (optional). The optional but recommended final step during an open card sorting exercise is to ask the participants to explain their rationale for each of the groupings they created. This enables you to uncover why each participant made the decisions they did, identify any difficulties they experienced, and gather their thoughts on any topics that might have remained unsorted.

KEY CONSIDERATIONOversimplification

As we’ve seen, simplifying an interface or process helps to reduce the cognitive load for users and increases the likelihood that they’ll complete their task and achieve their goal. But it’s also important to consider when simplification can negatively affect the user experience—more specifically, when we simplify to the point of abstraction, and it’s no longer clear what actions are available, what the next steps are, or where to find specific information.

A common example of this is the use of iconography as a way to communicate critical information about possible actions (Figure 3-8). Using icons has a lot of advantages: they provide visual interest, they save space, they present excellent targets for taps or clicks, and they can provide quick recognition if they hold universal meaning. The challenge is that truly universal icons are rare, and icons often mean different things to different people. While relying on icons to convey information can help to simplify an interface, it can also make it harder to perform tasks or find information. This is especially true if the icons aren’t immediately recognizable to users, who more often than not will have a wide spectrum of knowledge and experience.

Figure 3-8. Screenshot of the app bar from Facebook’s iOS app (source: Facebook, 2019)

Another complicating factor is that similar icons may be used to represent different actions or information, sometimes in complete opposition, from one product or service to another. There is no icon standardization body that regulates what icons can be used where in websites or apps, which means how they are used is left to the discretion of the designers and their teams. We know that an icon can represent different things to different people, but what about when the same icon represents different actions? Since there is no standardization, the functionality attached to an icon can vary from one digital experience to another. Take, for example, the “heart” and “star” icons: they typically indicate the ability to favorite, like, bookmark, or rate an item, but they may sometimes simply indicate a featured item. Not only does the meaning and functionality of these two icons vary across different products and services, but they also often compete with each other. This obviously results in confusion and an increase in the cognitive load on users, because the icons’ meaning is hard to interpret precisely.

Adding contextual clues helps users to identify the options that are open to them and the relevance of the information available to the tasks they wish to perform. In the case of iconography, studies have shown that simply adding text labels to accompany icons will provide clarity and aid users with both discovery and recognition. This practice is even more critical when using icons for important elements such as navigation (Figure 3-9). The addition of text labels effectively reduces the abstraction of the icons alone by including additional information to help convey meaning and increase usability.

Figure 3-9. Text labels accompany icons in the navigation on the Twitter web app (source: Twitter, 2019)

Conclusion

Hick’s law is a key concept in user experience design because it’s an underlying factor in everything we do. When an interface is too busy, actions are unclear or difficult to identify, and critical information is hard to find, a higher cognitive load is placed on users. Simplifying an interface or process helps to reduce the mental strain, but we must be sure to add contextual clues to help users identify the options available and determine the relevance of the information available to the tasks they wish to perform. It’s important to remember that each user has a goal, whether it’s to buy a product, understand something, or simply learn more about the content. I find the process of reduction, or eliminating any element that isn’t helping the user achieve their goal, a critical part of the design process. The less they have to think about what they need to do to reach their goal, the more likely it is they will achieve it.

We touched on the role of memory in user experience design with cognitive load. Next up, we’ll further explore memory and its importance with Miller’s law.

1 In a closed exercise, the groups are predefined by the researcher.

² Jakob Nielsen, “Card Sorting: Pushing Users Beyond Terminology Matches,” Nielsen Norman Group, August 23, 2009, https://www.nngroup.com/articles/card-sorting-terminology-matches.