Card Sorting

Anonymous contributor

Published Oct 28, 2023

Card sorting is a user research method where researchers record keywords on cards and participants organize these cards into groups. This method is often used during the early stages of the research phase and is frequently used for designing information architecture.

There are many tools that can be used to conduct a card sorting session. In-person sessions can be done using pen and paper, index cards, or sticky notes. Card sorting can also be facilitated virtually using online tools, such as Miro, Optimal Sort, and FigJam.

A GIF of someone doing a card sorting activity in Miro. The mouse is moving cards labeled "Platforms", "Heels", "Sandals", and "Sneakers" into a group called "Shoes".

Open vs. Closed

Open card sorting is when participants are given a stack of cards with keywords and asked to categorize them in any number of groups. Once they’ve sorted the cards, participants name each group.

Diagram illustrating three steps of open card sorting. Step one: Participant receives a stack of cards. Step 2: Participant divides cards into groups. Step 3: Participant labels groups.

By contrast, closed card sorting is when researchers create labels for categories and ask participants to sort the individual cards into these predetermined categories. Researchers evaluate if participants place the cards into the expected category.

Diagram illustrating three steps of closed card sorting. Step one: Participant receives a stack of cards. Step 2: Participant sorts cards into groups the researcher has made.

Moderated vs. Unmoderated

Moderated card sorting is when the researcher conducts a one-on-one interview with a participant to probe why they categorized the cards in one way or another. This provides qualitative data because researchers can ask follow-up questions and discover why participants believe certain terms belong in certain categories.

Unmoderated card sorting requires participants to organize cards into their groups without a researcher overseeing the session, usually via an online tool. This method generates quantitative data and is good for conducting large-scale research. One downside of unmoderated card sorting is that researchers are not able to ask follow-up questions, making it difficult to understand the reasoning behind why a participant placed a card in a certain category.