User Experience Design with Mental Models

Have you ever tried a new product expecting it to be difficult but found it simple? Or how about the opposite? — using something that should be easy but isn’t. In the latter case, the product wasn’t designed with an understanding of how you think. Design intuitiveness is determined by the difference between how you expect something to work and how it actually works.

Understanding new concepts is largely based on relating the concept to what you already know. The first time you read an ebook you probably compared the experience to reading a physical book. In both versions, content is segmented into pages, they both have covers, and you can “bookmark” important pages.

The way you understand books is called a mental model. How the ebook actually functions is a conceptual model. An ebook doesn’t need to have pages; there is no limitation that requires them. Simulating pages makes ebooks easier to understand because you already know how pages work. The closer a conceptual model matches the mental model, the more intuitive it becomes. Ebooks use physical book conventions for precisely that reason.

Anything designed has a conceptual model, including websites. People form mental models about everything and use them to understand new concepts; therefore, intuitive design is a matter of matching your conceptual model to the user’s mental model.

Users Work from Mental Models

Mental models are conceptual “maps” that allow a user to predict an object’s behavior based on past experiences with the same (or similar) objects, and they’re a great addition to your design toolkit.

For example, predicting how a faucet works is based on a mental model–specifically your own unique model of faucets which was formed by all your past interactions with faucets. People use mental models to make sense of the world around them, but mental models are slippery creatures. Not only do they differ from person to person (based on knowledge and experience), they’re also highly dependent on culture. One example is that in Japan book pages are turned from the right side to the left due to the way the language is written, whereas in English-speaking countries books are read from left to right.

The key point to note here is that mental models are beliefs not facts. A faucet is designed to operate in a specific manner — fact. If I’m accustomed to using a faucet by turning a handle and run into a design that does not operate this way, my mental model is shattered; my belief in how it should operate no longer works. Depending on my mental constitution (experience, curiosity, perceived importance of water at that moment in time, etc.), I’ll either experiment in an attempt to adjust my existing model or I’ll give up and look for something more familiar.

Mental models shape perception which ultimately influences your behavior. If you need to write something down and see something that looks like a pen you will likely try to write with it. Your mental model tells you that objects shaped like pens can be used to write on paper. If the “pen” were actually a digital stylus without ink or lead you would be frustrated at the wasted effort. Likewise, if you came across a pen that didn’t conform to your mental model of a pen, you would pass it up–unaware you could use it for writing.

On the web, experienced and novice users have differing mental models. Experienced users include established conventions in their mental models. Underlined text is clickable, the button at the bottom of a form submits, and site logos typically link to the homepage. Inexperience users on the other hand rely on mental models to determine how to accomplish a given task — it’s just that their models are not as deep or informed.

When design complements a user’s mental model the experience is positive, memorable, and efficient which is the goal. Like the unusual faucet which is turned on via motion sensor rather than handle, a website doesn’t always operate as expected. The way the site actually operates is known as the conceptual model.

Uncovering the User’s Mental model

Everyone’s mental models differ, so the goal is not to unify them all. Rather, you’re looking to understand commonalities in the way users make sense of things. Since every target user is unique, you can’t guess mental models or base them on previous projects. Each one must be developed from scratch. This can be done through conventions, task analysis, and an assortment of user-research techniques.

Conventions as Mental Models

Users who have even the most rudimentary understanding of websites share a basic expectation of where components should be located and how they should function. Some common conventions are:

• An input box in the top right is a search box
• The logo is in the top left
• Links are underlined
• The text of a link describes where the link takes you
• Beveled objects are clickable buttons
• Page headlines match the navigation label
• You can scroll to see more of a page if it extends past your screen

Conventions–while important–aren’t enough to guarantee an intuitive website. You still need to understand exactly who’s using your website and how they think. User discovery is the ideal way to uncover these insights and there are six common methods of going about it.

Task Analysis

Task analysis is the interpretation of users’ common tasks and goals. If you developed a list of user tasks for the functionality layer, task analysis will seem much more straightforward. Though task analysis can be speculative, it’s an easier way to suss out the user’s mental model. One thing to note is since it does not involve actual interaction with the target user its accuracy is low.

Surveys and Questionnaires

Surveys are a quick and low-cost way to find out what users are currently thinking and doing. They can be administered via traditional methods (phone, in person, and mailings) or digitally via email or websites. Some websites present randomly selected users with surveys–an effective way to gain insight from actual users without having to recruit and contact them individually.

The quality of the information gained is directly related to the quality of questions asked. Great survey design is a delicate science and should be well researched and executed.

Focus Groups and Interviews

Surveys give you direct yet sometimes shallow and hasty insight from actual users. In-person research allows for more nuance. Interview moderators can better communicate by asking follow up questions, seek additional explanations and pay attention to non-verbal communication.

As with surveys, you need a large enough sample size for this approach to be effective. Everyone has different opinions and one person’s thoughts in isolation are irrelevant. Your goal is to identify trends and commonalities to provide support for design objectives.

Contextual Inquiry

Contextual inquiry is observing user behavior in their familiar or typical environment. The problem with interviews, focus groups, and surveys is that data quality is dependent on good questions and accurate answers. Some participants alter answers in fear of judgment; others don’t actually understand why they do something and develop false explanations, and a bad question can be misleading.

By observing a user in their own environment you can see how they react and assess relevant behaviors and motivations. You may be surprised to find some users are able to sit uninterrupted for hours at a time, where others need to move around every fifteen minutes. A site for fidgety users should be designed differently than one used in deep focus.

Participatory Design

Participatory design is an emerging method of mental-model development. In this practice, target users inform design sessions by providing designers with a running commentary on their needs and expectations. In a sense, the users become the designers.

The most well-known method of this is paper prototyping. Paper prototypes are hand-drawn interface mockups using cutouts to demonstrate page changes (like a modal windows, hover states or error messages). Participants pretend as if they are using the site, attempting to complete tasks, all while discussing what they’re doing and why. If a participant stumbles, the designer can quickly iterate and adjust the experience to a more optimal one.

The practice is not without pitfalls. Like surveys, focus groups, and interviews, people are not always aware of why they behave the way they do.

Usability Testing

Usability testing is typically done on an existing interface, but it doesn’t have to be. You can test earlier to inform better mental models. In usability testing, users are observed while navigating an interface (existing or prototype), as they attempt to complete predefined tasks. Much like paper prototyping, they are asked to explain what they’re doing throughout the process.

Observing users behavior and hearing how they think, gives valuable insight into how they solve problems. Usability testing is more expensive, as it requires that you build a website of some sort. The benefit over paper prototyping is improved data, as the environment better represents an actual experience.

Because a website or prototype is required, usability tests tend to be the most expensive method. For many projects it isn’t feasible due to timeline and budget. When it is possible, usability testing provides the best result.

Regardless of what method(s) you use, usable interfaces for the web or applications require consideration for the user’s mental model. If you can identify their motivations and accommodate their reactions, you can then harness the power of conceptual design principles to create an engaging and intuitive site that just makes sense.

Conventions Aren’t Everything

Websites are not intuitive because they include a standard set of “usability conventions.” While conventions rarely hurt they don’t address how unique user groups perceive, interpret, and make sense of new concepts.

Intuitive design is more effectively achieved by matching the way a website functions to the way users think. More specifically, the system (how the site actually functions) is translated into the conceptual model (how the function is displayed to the user) which matches the users mental model (how users make sense of the world).

Save