“I’m a Visual Learner.” What Should a Designer Do With That?

Imagine a restaurant manager discussing a new ordering system with the person responsible for training the staff. The previous rollout did not go particularly well. Employees completed the course, passed the questions, and then struggled to use the system during a busy shift. Orders went to the wrong place. Modifications disappeared. Someone discovered a workaround that worked beautifully until it created a different problem for the kitchen.

The manager has an explanation. “Our people are hands-on learners. They need to do it.”

There is probably something useful inside that observation. Employees may need practice entering realistic orders, correcting mistakes, and making decisions while a customer waits. The course may have explained the system without preparing anyone to operate it. Those are worthwhile possibilities to investigate. But the conversation can take a shortcut. Someone writes “kinesthetic learners” in the analysis document, and a description of the audience begins doing the work of an explanation.

We have identified a learning style. We have not yet established why the training failed.

This is where the familiar debate about learning styles becomes interesting to me. Declaring them a myth is relatively easy. Deciding what to do with the experiences people describe through that language requires considerably more judgment. An adult who says, “I’m a visual learner,” may be giving us useful information about something that helped them. They may also be offering an explanation that the evidence does not support. A designer has to be able to hear the first without automatically accepting the second.

01The claim hiding inside a familiar phrase

“Learning styles” can refer to a sprawling collection of models, so an article about the subject needs to be clear about its target. Here, I am concerned primarily with the familiar practice of classifying people as visual, auditory, reading/writing, or kinesthetic learners and then choosing instruction to match that classification.

The important claim is that matching produces better learning. It goes beyond observing that someone enjoys diagrams or prefers listening to an explanation. It proposes that identifying a person's style gives us a reliable basis for selecting the method through which that person will learn most effectively.

The influential review by Harold Pashler and colleagues explained why that claim needs a particular kind of evidence. If one group is classified as visual and another as auditory, the strongest method should differ between those groups in the predicted direction. A visual presentation helping everyone would support that presentation for the material being taught. It would not, by itself, establish the value of sorting people into styles. Their review found insufficient evidence to justify applying learning-style assessments in general educational practice.^[1]

Research involving adults has also tested the proposition directly. Beth Rogowsky, Barbara Calhoun, and Paula Tallal compared learning from an audiobook and electronic text among college-educated adults. They found no statistically significant matching advantage on comprehension tests administered immediately and two weeks later. That study addressed a particular population, set of materials, and comparison. It gives us relevant evidence about adult learning without settling every question about every learner or format.^[2]

More recent findings deserve careful wording, too. A 2024 meta-analysis by Virginia Clinton-Lisell and Christine Litzinger found a small average benefit for matched instruction across 21 studies. However, the pattern needed to support matching across different styles appeared in only about a quarter of the learning outcome measures examined. The authors judged the benefits too small and infrequent, alongside concerns about study quality and implementation costs, to warrant widespread adoption.^[3]

That is a more demanding conclusion than either side of a social media argument usually wants. We should avoid claiming that no study has ever produced a favorable result. We should also avoid turning a small, inconsistent finding into permission to build an entire design process around learning-style inventories. The practical question is whether the classification earns the decisions and resources we attach to it.

02Why the idea feels respectful

There is an understandable appeal to telling an adult that their difficulty with training reflects a mismatch between the instruction and the way they learn. It acknowledges that the experience was difficult. It offers an alternative to assuming that the learner was incapable or unwilling to try. For someone who has spent years feeling unsuccessful in formal education, that explanation may feel like relief.

The difficulty begins when reassurance becomes a design prescription. “I found that diagram helpful” is an observation worth exploring. “Therefore, future instruction should be designed around my visual learning style” adds a conclusion that needs its own justification. We can respect the observation while asking what made the diagram useful. Perhaps it made a relationship visible, reduced the need to hold several steps in memory, or gave unfamiliar terminology a concrete reference. Those possibilities give a designer something to examine and improve.

I would be reluctant to respond to that learner by announcing that their experience has been debunked. They did find something helpful. The question is what we can reasonably infer from it. A person should not have to defend a theory of cognition before we take their frustration seriously.

The same care should extend to designers who have used learning styles. Many were taught these models as established professional knowledge. They may have built thoughtful, effective learning experiences while using an explanation that did not account for the results. We can examine the explanation without treating their entire practice as worthless. The more useful conversation asks which decisions improved the experience and what evidence would support making those decisions again.

03What “hands-on” leaves unexplained

Return to the restaurant. Suppose the designer watches an employee enter a straightforward order. The employee completes it without difficulty. Then a customer asks for two substitutions, a split payment, and an item that is unavailable. The employee hesitates, opens the wrong menu, and calls the manager.

That observation changes the assignment. The employee can operate the basic interface. The difficulty appears when an order departs from the example shown in training. A reasonable next step would be to investigate those variations and give employees opportunities to work through them, with guidance and feedback. Calling the audience kinesthetic does little to identify which situations deserve practice or what the feedback should address.

Now imagine a different finding. Employees know how to enter modified orders, but the training environment uses a different screen configuration from the live system. Or they understand the process but cannot remember an exception that occurs twice a month. Or nobody agrees on who is authorized to approve a particular change. Each finding would lead the designer toward a different response.

“They need to do it” may still be a sensible starting observation. It becomes more useful when we ask what they need to do, under which conditions, and where their current performance breaks down. Those questions help us choose between practice, a reference, a clearer interface, a management decision, or some combination. The style label cannot make those choices for us.

04Choice still has a place

None of this requires a designer to become indifferent to how adults want to engage with material. If someone wants a transcript because they can search it quickly, that is a practical reason to provide one. If a demonstration makes a complicated sequence easier to inspect, it may earn its place in the design. If an employee needs to review information in a noisy environment, audio alone creates an obvious problem. We can make these decisions by examining the use case.

Accessibility deserves particular care in this conversation. Captions, readable text, keyboard access, and information that does not depend on color alone address barriers to participation. A person who cannot access an explanation needs an accessible explanation. Debating whether they possess a learning style contributes nothing to resolving that barrier.

There is also room to offer a choice when the alternatives are suitable for the purpose. An adult may reasonably prefer to read an explanation rather than watch it unfold at a speaker's pace. Honoring that preference does not require promising a superior learning outcome. We should simply be clear about what the choice accomplishes and what still needs to be demonstrated afterward.

John Hattie and Timothy O’Leary’s 2025 examination of the literature highlights a related source of confusion: learning styles and learning strategies can become tangled together. A relationship involving how people study does not establish that matching instruction to a style causes better learning. Their argument favors helping learners select adaptable strategies suited to the task and goal.^[4] For a designer, that suggests a more useful ambition than finding the right permanent category for each person. We can help adults develop ways to approach unfamiliar work and judge whether those approaches are helping.

05AI makes the shortcut easier to scale

This debate has a contemporary consequence. An AI system can generate a visual version, an audio script, and a hands-on activity from the same source material. Producing several versions may take little enough time that questioning the premise starts to feel unnecessarily fussy. Why argue about styles when we can accommodate all of them?

My concern is that faster production makes an unsupported explanation easier to operationalize. A platform could ask three questions, assign a learner type, and generate an experience that appears impressively personal. The learner receives a label. The organization receives a dashboard. Neither tells us whether that classification improved the person's ability to do the work.

Lower production costs also leave other costs intact. Someone still needs to check the accuracy, accessibility, and suitability of each version. A change in format may change what the learner can inspect or practice. Generating an activity is easy to confuse with designing a useful opportunity to learn, particularly when the activity arrives with a confident explanation of why it suits the learner.

I would rather use that capacity to respond to evidence from the work. If someone repeatedly misses a particular exception, we can investigate the misunderstanding and provide a relevant example. If someone can explain a process but struggles to perform it, we can examine the gap between explanation and execution. Those are concrete starting points for personalization, and the resulting choices can be checked against subsequent performance. A preference questionnaire should not substitute for that investigation simply because a system can act on its results immediately.

06Listening without labeling

When an adult says, “I'm a visual learner,” I think a useful response begins with curiosity. What were they trying to understand? What helped? What happened when they tried to use it? The conversation may reveal a useful representation, a confusing explanation, an access barrier, or a missing opportunity to practice. It may reveal a preference that we can reasonably accommodate. We learn more by asking than by accepting or rejecting the label on the spot.

For the restaurant manager, this means taking the concern about hands-on learning seriously enough to investigate it. Watch an employee enter a difficult order. Find out which decision creates hesitation. Check whether the training examples resemble the conditions of a busy shift. Then build a response whose purpose can be explained in terms of what employees need to accomplish.

The appeal of learning styles is that they seem to make understanding people easier. The responsibility of design is to remain interested when understanding people becomes more complicated. Adults bring experiences, preferences, abilities, and constraints that deserve attention. Putting those differences into a few familiar categories can end the inquiry before we discover the detail that would actually change the design.

The next time someone tells us how they learn, we have an opportunity to ask them to show us. There may be a much better design waiting inside the answer.

07Research and references

1. Pashler, H., McDaniel, M., Rohrer, D., & Bjork, R. (2008). Learning styles: Concepts and evidence. Psychological Science in the Public Interest, 9(3), 105–119. Read the research ↩

2. Rogowsky, B. A., Calhoun, B. M., & Tallal, P. (2015). Matching learning style to instructional method: Effects on comprehension. Journal of Educational Psychology, 107(1), 64–78. Read the research ↩

3. Clinton-Lisell, V., & Litzinger, C. (2024). Is it really a neuromyth? A meta-analysis of the learning styles matching hypothesis. Frontiers in Psychology, 15, 1428732. Read the research ↩

4. Hattie, J., & O’Leary, T. (2025). Learning styles, preferences, or strategies? An explanation for the resurgence of styles across many meta-analyses. Educational Psychology Review, 37, Article 31. Read the research ↩

Decision Workbench

Make the questions visible before you build.

Use Decision Workbench to examine a training request, surface assumptions about the audience, and identify the questions and facts that should shape your design.

Try Decision Workbench Read more perspectives

Use the output to prepare your investigation. Learner observations and credible evidence still need to inform the decision.

Bring a real request

• Identify questions and assumptions
• Name the facts that need checking
• Bring questions to the right people
• Develop recommendations to discuss