Research methods workshop / 5 MIN READ
Let the participant invent the groups before naming yours
Prepare and moderate an open card sort that preserves participant categories, uncertainty and explanations.
AI-assisted research and analysis for the SNS Creative Fieldbook
Sources checked 2026-11-06
An open card sort helps a team hear how people group a set of items and describe those relationships. Its value depends on allowing participants to create their own organization rather than rewarding them for reproducing the team’s existing structure.
What you will leave with: An open card-sort kit and a participant grouping record.
Use this when
- A team needs to understand how users relate a bounded collection of information.
- Internal categories feel obvious to staff but may not match customer expectations.
This is not a finished navigation design, a quantitative clustering lesson or proof that the most common label must become the final taxonomy.
Prepare items that represent a coherent collection
Digital.gov describes an open sort as participants grouping content and naming their categories. Choose a collection that fits the question the team actually needs to explore.
Give each card enough meaning to be understood without presenting a miniature essay. Avoid mixing broad departments, tiny tasks and unrelated features as though they were equivalent items. Identify ambiguous terms before the session, but do not remove every difficult item if understanding that ambiguity is part of the research. Keep a record of the card wording and set version so you can interpret what each participant saw.
Explain the activity without teaching the organization
Tell participants they can group the items in a way that makes sense to them and name those groups in their own language. Make clear that there is no hidden correct arrangement they must discover. Demonstrate the mechanics using unrelated sample items if the tool itself needs explanation.
Do not display the current site categories beside the cards or use internal department names as suggested starting points. Allow a way to flag unfamiliar items, uncertain placement or an item that seems to fit more than one group. If the tool forces a single position, record that constraint rather than interpreting the final location as complete certainty.
Explore the reasoning behind the arrangement
Observe how the participant forms, combines and revises groups. Ask neutral questions about the relationship they see between items and the meaning of their labels. Avoid congratulating a grouping because it resembles the team’s preferred design.
Capture meaningful hesitation and exceptions as well as the final arrangement. Two people may use the same label for different ideas, or different labels for similar ideas. A card placed alone may signal an unfamiliar concept, a distinct task or a weak item description. Ask enough to understand the reason instead of treating every isolated card as a recommendation for a separate top-level category.
Preserve the result as evidence for design work
Save the arrangement, participant labels, relevant explanations and unresolved ambiguities in the approved research record. Keep identity details separate where appropriate and note any moderator assistance or tool constraint.
Discuss what the exercise suggests about relationships and language without turning it directly into a navigation specification. A category that makes sense when all items are visible may not help someone find an item from an unfamiliar starting point. Use the findings to develop and review candidate structures through appropriate subsequent work. The open sort contributes an account of organization, not a universal map that removes the need for design judgment.
Make the reasoning concrete.
A fictional agency tests how clients group service information. One participant places “launch support” with training, while another places it with production. Both use a group label meaning “getting started.”
Count the matching labels as agreement
Opening: Treat both groups as the same category and ignore their contents.
Hypothesis: The summary appears simple.
Risk: It misses different interpretations of what getting started involves.
Preserve labels with their reasoning
Opening: Ask what each participant expects launch support to include and record the associated items.
Hypothesis: The team can see the actual conceptual difference.
Risk: The resulting design question remains open rather than collapsing into one obvious category.
The researcher records the different meanings and checks whether the service description itself is ambiguous. The team develops a clearer information structure without claiming the card sort directly specified the final navigation.
Take the questions into your next brief.
Use these prompts in your own notes. Nothing is submitted from this page.
Collection
What coherent set of items will participants organize?
Wording
Are cards understandable without revealing preferred categories?
Freedom
How can people show uncertainty, multiple fits or unfamiliarity?
Reasoning
What explains the grouping and labels?
Use
What design question does the result inform without settling by itself?
Watch for the wrong conclusion.
- Calling a sort open while supplying the preferred categories.
- Discarding uncertain cards to make the result look tidy.
- Treating identical labels as proof of identical meanings.
Your next action
Prepare a small open-sort kit and check that the instructions and interface do not reveal the organization you hope to see.
Sources & scope
The workflow, worksheets and hypothetical examples are original editorial analysis. Sources support only the identified principles. No measured business results or named human reviewer are claimed.
- Digital.gov: Card sorts →Primary government research-method guidance · Supports: Open card sorting lets participants form and name groups; the preparation and ambiguity record are original editorial analysis.