Card sorting – sorting through the options

Which one of these shapes doesn’t belong?

• Shape two, because it is the only one that isn’t blue?
• Shape three, because it is the only one that isn’t a square?
• Shape four, because it is smaller than all the others?
• Shape five, because it is the only one that doesn’t have an outline?
• Or is it really shape one, because it is the one that can’t be called different?

We all come from different backgrounds, and so we all have different responses to problems like this one.

So how do we group information if we all see things a little differently? The answer is card sorting.

What is card sorting?

Card sorting is an activity used to organise content in a way that is logical and intuitive. It is often used to create and verify an information architecture (IA) or taxonomy. It should always be conducted with real users.

The basics of card sorting are pretty straightforward: participants are given a stack of cards with content topics or titles on them, and asked to put the cards into groups. However, card sorting –  like ice cream – comes in several different flavours that can be used for different purposes.

Much like the puzzle above, there are many different right answers to card sorting. To pick the best approach for your situation, you need to consider:

• What am I going to put in?
• What do I want to get out?
• Who do I need to access?

What are the types of card sorting?

Open card sort

In open card sorts participants are given no structure, and are traditionally asked to name the groups they create. This style is great for creating an initial taxonomy or IA.

Closed card sort

In closed card sorts participants are given headings for predefined groups, and asked to make their cards fit within those groups. They may also be asked to identify where certain cards don’t belong in any group, or if they feel the group should be renamed. This style is better for validating existing taxonomies or IAs.

Independent card sort

Independent card sorts are where participants complete the card sort individually, often being asked to think aloud about the decisions they make. This approach works best with small numbers of cards, and it allows shy or soft-spoken participants to make an equal contribution to the results. This style of card sort also generates more in-depth data about groupings and user thought processes.

Group card sort

Group card sorts are where participants complete the card sort in groups so that an open dialogue is used to determine where cards are placed. This approach is better for large numbers of cards. There is a risk that groupthink or a dominant personality may override the accuracy of the activity and this style produces more generalised data, but it also reduces the number of refinements that are needed later.

In-person card sort

In-person card sorts are where participants complete the activity with physical cards on a table or wall, with a moderator observing them. It is the more common approach to card sorting and is most suitable when you have participants who can dedicate time to the activity. As with all user research activities, there is no way to get data as rich as being in a room talking to actual users while they make decisions.

Online card sort

Online card sorts are where participants complete the activity using a web-based tool that enables them to drag and drop cards into groups. This is a better option when you cannot access user representatives because of time or location. While less common, because it removes the observational aspect of the research, it is a great mechanism to gather content from otherwise inaccessible users. You can also invite a wider group of people to participate and the analytics are insightful.

Need some help with a card sorting or planning for an intranet or website? We can talk to you about some of the best practice we’ve seen with our clients across the public and private sectors.