free consultation
Card sorting is a simple and inexpensive technique for understanding how users organize and structure content in the way that feels most intuitive to them. Card sorting can be applied in a variety of ways - on paper, using Post-It cards, or with online tools that allow such studies to be conducted remotely.
table of contents:
Card sorting is a tool to explore users' ways of thinking. It is a way to understand their expectations about the layout of content on the site and how they themselves would prefer to group the information. Although not a research method, it is often used by user experience (UX) designers to create the final structure of information. In addition, it provides important clues for creating categories and helps identify trends in user preferences.
There are several ways to classify cards, such as:
Users are asked to group cards according to their criteria and assign them matching names. This method is mainly used for already existing information structures, arranging products on a website or creating completely new information layouts.
Test subjects are given content cards and a set of cards with established categories. Their task is to sort the content cards into the appropriate categories. This is the method used when adding new content to an existing website.
Free-list sorting involves each user working individually using their own computer, where they sort cards online. A variety of test tools are available and can be distributed to any number of users. This type of card sorting is done remotely, without direct contact with the user.
The tests are conducted in person with an observer, where users are given paper cards and asked to justify their choices. The observer works with the user to clarify any ambiguities and discuss decisions made during the test.
Find out what Design Thinking is
In group studies, the involvement of a dozen or so participants is preferred to achieve significant correlation. Tasks can be assigned both individually and in a group format. In remote research, a one-on-one approach is often used, especially when researching a large group of users. This type of research offers comfort to participants, eliminates pressure and promotes free self-expression, which has a positive impact on the quality of the results obtained.
In traditional research, a larger number of participants is also important. This approach increases the level of creativity, encourages brainstorming, which often leads to unique ideas. Joint discussions promote effective analysis of findings, allow for free exchange of ideas and joint creation of innovative solutions.
When sorting cards, it is important to follow a set plan of action.
When you are conducting card sorting, a few tips can help you improve your research process:
Before you start official testing, it's a good idea to do some preliminary card sorting, possible with friends or colleagues, for example. This will catch and correct any errors or typos that may interfere with the smoothness of the tests.
The number of cards matters, but it is worth sticking to moderation. For open sorting, it is best to focus on about 60 cards. This is a sufficient number so that the user does not feel overwhelmed by an excess of information.
To make analysis easier, it's a good idea to number the cards. This is a practical approach that makes it easier to organize data. Using a spreadsheet or table when displaying a list of topics makes them much easier to understand.
Abbreviated and closely related labels are a key issue. They are more readable and understandable.
Starting a research session with randomly selected content guarantees an objective grouping, unencumbered by any bias.
It is useful to determine the amount of time participants need to complete a task, even before the task begins.
The optimal approach is to put together two types of tests: open as the first, and closed as the second. The former allows for user analysis, while the latter allows for testing existing labels and their intuitiveness.
How to determine the right number of cards?The selection of the number of cards should correlate with the purpose of the site, not the other way around. Defining the goals and then adjusting the number of cards is key. Other research can help you understand what to focus on.
Evaluation of card sorting resultsIf programs were used, the analysis is usually automatic. With traditional methods, in-house analysis is necessary.
Test results are often presented using dendrograms, especially when the number of participants was large. However, not everyone prefers this method, as dendrograms can be difficult to interpret. Using a spreadsheet, you can draw your own conclusions from the tests, which can be easier, especially with a smaller study sample.
See our offer:
Often, studies use closed sorting to verify that the categories in a site match previous assumptions. Nevertheless, this is not a method of definitively confirming their existence.
When conducting research, it is important to keep in mind the goal, which is to understand user feedback on both the category and the overall content of the site. A good practice is to leave the user with a challenge and observe his reactions. Taking careful notes of his comments and trying to understand his thinking can yield valuable insights.
In the case of imposing categories, the user is given a suggestion, which can influence his choice, even if he does not fully agree with the given arguments. On the other hand, when the user selects the categories himself, unexpected possibilities can emerge, which is valuable even for specialists.
Card sorting can introduce pitfalls, especially if you have a lot of cards, resulting in the creation of excessively diverse categories. It is important to maintain balance - try to keep the number of cards proportional for all categories, according to the method chosen.
For example, if we have products such as a remote-controlled car, a doll, plasticine and crayons, the first two might go into the toys category, and the others into art supplies. Now, by adding a dollhouse and soldiers, our categorization may change - new categories are created, such as art supplies, toys for girls and toys for boys. This may raise questions about incorrect gender distinctions.
So it's worth doing a test with a few people to see if they create categories that are too detailed. If so, it's worth revising our approach and revisiting the categories.
See also:
The final stage of the study is the analysis of the collected data. It is important to find recurring group names, frequent subject combinations and cards that are most often left unattributed.
These patterns, along with participants' feedback after the sorting phase, reveal where difficulties arise and their sources. It may turn out that some topics are not connected because of users' lack of knowledge, or specific names are not understood because users do not use them in their own language, and the card set lacks matching synonyms.
Backward card sorting analyzes search paths and looks for patterns to better align the structure with user thinking and behavior in the future.
Digital Agency Hauerpower
Copyrights 2010-2022
