User Experience Research – A Card Sort
User-centered design techniques are used by UX researchers for several reasons. This article will help you understand how to manage and structure information by using a method called card sorting.
What is a card sort?
A card sort is a way for your team to get your user’s hands on the information and will help you structure the hierarchy of a website, web application, or mobile app. There are two methods you can use to conduct the card sort. Running a card sort is a great way to learn about and ultimately structure for your websites’ information. It can be used to help you set up the site’s navigation and will give more meaning to the words you use in your menus, headings, and will help you make the most sense to your users.
There are two kinds of card sorts. You can use an open card sort if you want your users to define the categories. For instance, this can be used to help you learn more about your sites navigation system. However, if you have some set terms or baseline structure from an existing web application it might be better to run a closed card sort. Some researchers may elect to blend these two methods, but I’ve traditionally kept with either an open or closed sort based on what I’m trying to solve.
Card Sort as User Research
If you’re working on a new website and have already identified information that needs to be structured you may need help aligning that information to the user’s mental model. You can think of user research as a way to understand your user’s motivations, behaviors, and needs, through observation. I like to refer to user research as a hypothesis, an idea which will need to be tested to prove or disprove the theory. There are dozens of research methods. However, card sorting is recognized as a great way to learn about how your users think about the data. It’s also a great way to build a structure for your data.
Setting up a Card Sort
In my career, I’ve led internal and external (consumer facing) application redesigns, but have mostly led the UX research, design, and testing efforts for large internal enterprise systems. Several of the systems I’ve redesigned have had 5 or 6 levels of information each with 10 to 20 available functions for the users to navigate. With each redesign, I have always conducted a card sort to help the project team understand how users think about the information. This will help you understand how information is grouped and may unveil findings that you wouldn’t uncover with web analytics.
When choosing your participants, try to include users with no prior understanding of the project. Bias is not your friend. Don’t let it ruin your study. In my opinion, you’ll only want to include 3 to 5 users in each session. However, try adding or subtracting participants to get a feeling of what works for your team. You can also try running the card sort with everyone at once or separately to see if one method works better for you than the other.
You’ll need to provide the participants with your baseline information that you plan to restructure if you’re doing a closed card sort. If you’re running an open card sort then make sure you provide the participants with paper or note cards and markers so they can create the naming conventions and align like terms to create the hierarchy.
There are also card sorting websites or applications that allow you to take your study to the digital arena. This is a great way to save time and will most likely give you the ability to get in front of more users. Some card sorting websites or applications will also give you some data automatically. This will save you a lot of time and give you better findings. However, if you don’t have money or means to use online card sorting applications, don’t hesitate to do it manually. There are pros and cons to each approach. I’m a big fan of getting in the room with users as this allows me to dig deeper into certain areas of the conversation.
An Open Card Sort
An open card sort is a great way to give freedom and control to your users. This method will allow you, as a research, to understand why your users chose to align certain information together. In this type of card sort, you’re allowing the users to set up the categories as well as aligning the remaining cards within these categories.
A Closed Card Sort
A closed card sort will give you more control over your study. A closed sort will allow you to provide a starting point for your users. Again, let’s assume you have an existing menu system with 15 or 20 functions. You’ll want to write a single function name or menu options on each of the note cards.
Decide on which cards will be the categories and allow the users to separate out the remaining cards into each of the established categories. If you are running a study with multiple users let them discuss and debate how the cards will be grouped as a team or on their own. However, if they are together and working on their own you’ll need cards for each participant. Be sure to leave some time in the end to ask questions and understand why they’ve decided to group the information in this way.
Once you’ve finished the card sort and have some findings, take the time to analyze the information. Use a spreadsheet if you’re dealing with a large data set. A couple things to consider are how you capture the card sort information during or just after the exercise. Make sure to take a picture of the actual cards laid out on the table. Doing this will help you fill the information into the spreadsheet. The spreadsheet should include the participant number or name, category listing and a breakdown of the updated categories. Also, keep your notes and the answers to the followup questions in the same spreadsheet.
Take your time and use the findings to properly illustrate to senior management and executive team. You’ll want to transfer your findings and notes into a high-level and consumable presentation. Make sure you illustrate what worked as well as what didn’t work.
Card sorting is very simple to set up and run and provides consistent and reliable results. The research can be done with very little effort, time or money. If done properly, it will save you time and money. You can learn a lot about an existing web application’s structure as well as how that relates to the user’s overall mental model.
While this is a great method and will provide you with a lot of great information, it will not be a one and done effort. You should use this as a supplement to other research techniques during the discovery phase. This exercise, mixed with other user-centered design methods, will help you create a better product. It will ultimately make your product easier to use and help satisfy user and business needs.