Glossary · 7 · Evaluation methods
Card sorting
Also known as: Card sort, Open card sort, Closed card sort, Hybrid card sort
German: Card Sorting
In information architecture and usability research, card sorting is a method in which participants group cards, each naming a topic or content item, into categories that make sense to them and label the groups, so as to reveal how users mentally organize a subject; in an open sort participants create and name the groups, in a closed sort they use predefined categories, and a hybrid sort combines both.
- Applied cognitive psychology
- Evaluation methods
In one sentence
Card sorting lets users group and label content items to reveal their mental organization of a subject; sorts are open, closed or hybrid.
Example
Twenty service technicians sort 60 topic titles of a palletizer manual. Most group lubrication, belt tension and sensor cleaning together as “weekly checks”, a category the manual did not have; the portal gets a navigation entry for it.
Rests on
- Mental model and Cognitive schema: users bring their own organization of a subject; card sorting makes it visible.
- Prototype theory: people group by resemblance to typical members, not by strict definitions, which explains fuzzy and overlapping groups.
How it applies
- Navigation and portals: Structure the navigation of a documentation portal, the categories of a help center or the facets of search by how readers group the content.
- Taxonomy: Check whether a taxonomy or classification matches readers' expectations (closed sort) or discover new categories (open sort).
- Labels: Open sorts show readers' vocabulary for categories, which feeds information scent; participants' labels are input, not final labels.
- Machinery documentation: Sort tasks of operation, maintenance and troubleshooting with the people who do them; groups often follow intervals, locations or roles.
Card sorting vs. tree testing
Card sorting shows how users would group content. Tree testing checks whether users can find items in a given hierarchy. They are often combined: sort to design, test the tree to verify.
Limits
- Cards are isolated from context, layout and links, and participants usually build one level of groups; card sorting does not show complete navigation paths.
- Results show how users think about the content; they do not decide where safety-related information must appear.