Tree testing vs. card sorting. What one should you use?
We’ve written a guide that dives into the specifics of both research tools, their similarities, their differences, and why our UserQ card sorting and tree testing tools are handy when worked in unison.
On paper, there are clear similarities between card sorting and tree testing…they’re both super useful tools for validating your information architecture. But today we’re clarifying the key differences once and for all.
If the question is: When should you use each one? Our answer is you might actually need both…
What is card sorting?
Card sorting is a tool that’s used to help you create an IA in a democratic way. It involves creating a set of cards based on your IA content and asking participants to sort and group them together in a way that makes most sense to them.
{image of example card sorting test}
There are two types of card sorting:
Open card sorting –the tester is asked to create, name, sort, and categorise the cards themselves.
Closed card sorting – the researcher provides a set of cards and the tester sorts the cards in into predefined categories. In some cases, the researcher may decide to give the tester the option to add and name any missing category.
Closed card sorting is most suitable for evaluating and validating an already established IA, whereas open card sorting is great for earlier development stages – when you’re looking for ideas or are open to modifying the existing structure.
And what does card sorting tell you?
- provides insights into user expectations and how they tend to think
- how certain products or content should be categorised
- what the most logical labels and comprehensive structure might be
- whether your existing categorization and labelling is user-friendly and easy to follow (closed card sorting only).
Check out our full card sorting guide which sets out how you might create, run, and analyse a card sorting test.
What is tree testing?
Like card sorting, it’s a tool used to improve your website or app’s proposed information architecture (IA) by testing the findability. In other words, it helps you understand how easy it is to navigate a website and identifies if there’s any confusion within specific sections of the IA.
The test involves setting users a task to find certain information on a text-only version of your site structure, through hierarchical categories (ie. the ‘tree’). For example, an online supermarket might ask participants where to find a specific grocery product from a text list of your named product categories (dairy, meats, tinned goods, etc.).
{image of example tree test}
And what does tree testing tell you?
- whether the information content is grouped together logically and in a simple-to-understand way
- how quickly and easily certain info can be found
- what might be preventing users from finding what they’re looking for (i.e. misleading page names or confusing categories).
Check out our full tree testing guide which sets out how you might create, run, and analyse a tree test.
Key differences between card sorting and tree testing
Card sorting and tree testing both help you categorise content effectively and test the coherence of an IA, but they take different approaches and can be used at different research stages.
While tree testing typically evaluates an existing IA, open card sorting is often used to generate ideas and gather insights into how your content might be best structured.
It’s useful to think of tree testing and card sorting as a reverse of each other. Card sorting informs you about a user’s intuitive expectations on things like which content might go best together or under what category labels you might expect to find certain things…tree testing informs you on what already does and doesn’t work from the labels and categories you’ve come up with.
When is it best to use each method?
We recommend using both in tandem!
They complement each other perfectly and work best together to provide the most well-rounded view into the efficiency of your IA. But which one comes first? Well, it all depends on your research and what you’re trying to find out…
Luckily card sorting is a versatile tool that’s useful at any stage of the design process, including at the very beginning when you want some base ideas on how to structure your information and later down the line when you’ve already got a proposed IA and need some improvements. You might even want to retest a new reorganisation with card sorting to ensure it’s still aligned with user expectations. Best of all, card sorting can be carried out remotely as a digital tool AND in person with physical cards laid out on a table.
As a remote testing tool, tree testing is also pretty flexible – especially when you’ve got a first draft of the IA and you don’t want to proceed with testing the IA with a prototype since the interface may produce biases. It’ll show you if there are any major issues where users get confused or lost along the navigational findability tasks.
Become a UserQ researcher to use our card sorting and tree testing tools
No matter what stage you’re at, you can use our UserQ platform for card sorting and tree testing.
Both tests are simple to set up and easy to build, plus you can carry out the tests in English or Arabic, depending on who you’re targeting and what language your audience speaks. We recommend carrying out each method with 20 – 30 participants for both.
Try out a card sorting demo test here. Or check out a tree test demo.
Conclusion: tree testing vs. card sorting
Card sorting and tree testing can wonderfully work in unison, so pitting tree testing vs. card sorting against each other isn’t really the right way to think about it.
We advise carrying out both hand in hand. Card sorting can be used to generate ideas (particularly open card sorting), and it can be closely followed by tree testing to test those ideas.
Related Post
Subscribe to our newsletter
COMPANY