The learning styles myth is the belief that people learn better when instruction matches their preferred style, such as visual, auditory or kinesthetic. People do have preferences. What has not been shown is that teaching to those preferences improves learning, and the best-known reviews found little rigorous evidence that it does.
If you design courses, you have probably been asked for “content for visual learners” anyway. The request usually comes from a leader or subject matter expert (SME) who is sincere and has seen a VARK report or a workshop slide. A flat “that’s a myth” rarely changes the brief.
This post separates what the research shows from what it does not, explains why VARK keeps circulating, and gives you two practical tools: a myth-to-practice table for your next course build and a short script for the stakeholder conversation.
What is the learning styles myth, exactly?
The learning styles myth is the claim that matching instruction to a learner’s preferred format, such as visual or auditory, improves learning. Researchers call this the meshing hypothesis. The myth is not that people have preferences. It is the extra step from “I prefer diagrams” to “I will learn more from diagrams than from text.”
Three separate claims often get bundled together, and only the third is in dispute:
- Preferences exist: people say they like some formats more than others. The 2008 review discussed below found plenty of evidence for this.
- Abilities differ: people vary in aptitude. The same review found evidence for this too.
- Matching improves outcomes: teaching to the preference produces better learning. This is the claim that has not been shown to hold.
Keeping the three apart matters in a stakeholder meeting. If you deny that preferences exist, you will be wrong and lose credibility. If you say that matching has not been shown to help, you are on solid ground.
What does the research say about matching instruction to learning styles?
Reviewers have found almost no rigorous evidence for matching. In their 2008 review in Psychological Science in the Public Interest, Pashler, McDaniel, Rohrer and Bjork said a fair test must group learners by style, randomly assign them to methods and give everyone the same test. Few studies met that bar, and some contradicted the hypothesis.
The authors concluded there was not enough evidence to justify using learning-style assessments in general educational practice, as summarized on the Pashler et al. learning styles review record. They also noted that many possible versions of the idea have simply never been tested.
A later study followed the design they proposed. Rogowsky, Calhoun and Tallal (2015) measured college-educated adults’ auditory and visual preferences, then randomly assigned them to hear a nonfiction book as an audiobook or read it as e-text. Comprehension tests came immediately and again after two weeks. Preference did not significantly predict results on either test. A 2020 follow-up with fifth-graders also found no significant relationship between preference and comprehension.
Read the evidence precisely. It supports “matching has not been shown to improve outcomes.” It does not support “every possible style theory has been disproven.” That narrower statement is the one to use at work.
Why does the learning styles myth persist in L&D?
The myth persists because it feels true, it is simple to explain, and belief is common among educators. A 2020 review of 37 studies covering 15,405 educators in 18 countries found a weighted average of 89.1% belief in matching, with individual studies ranging from 58% to 97.6%, according to Newton and Salvi’s systematic review.
The authors caution that most of those studies used convenience samples and small samples, and that it is unclear whether respondents understood they were being asked about matching rather than preferences in general. Still, belief did not appear to be falling. Self-reported use or planned use was also high.
Debunking alone may not move people. In a 2017 survey of 114 UK academics, Newton and Miah found 58% believed in learning styles and 33% reported using them. Of the respondents, 32% said they would keep using them even when shown the lack of evidence.
In corporate learning, a few practical forces probably add to this. Learners can describe their preferences, so the theory seems confirmed by experience. A questionnaire gives a personal result that feels useful. And “visual learners” is easy to write into a course brief, even though it multiplies the production work.
Does VARK claim that teaching should match your style?
No. As of November 2026, VARK’s own website states that it “has never claimed that teaching should be matched to a learner’s ‘style’ in order to improve outcomes.” It presents VARK as a reflection tool covering visual, aural, read/write and kinesthetic preferences, per its What VARK Is (and Isn’t) page.
The vendor’s FAQ adds that knowing your preferences “does not automatically improve learning,” that most people are multimodal, and that VARK is not designed to measure intelligence, ability or academic potential. Its validity and reliability page reports coefficients of about 0.73 to 0.78 for version 9.2 and does not address whether matching improves learning.
That distinction is useful. A questionnaire can produce consistent scores and still tell you nothing about how to teach. The matching claim usually comes from people who adopt VARK, not from VARK’s documentation.
Quote the Source Back
When a stakeholder cites a VARK result as a reason to build separate tracks, read them the vendor’s own sentence that VARK has never claimed matching improves outcomes. It moves the conversation from “your view versus mine” to “what the tool itself says.”
What should course designers do instead of designing for learning styles?
Use techniques that have evidence across learners. Dunlosky and colleagues rated practice testing and distributed practice as high utility and rereading and underlining as low, in their review of learning techniques. They called the two high-utility methods useful because they “benefited students of many different ages and ability levels.”
A US Institute of Education Sciences practice guide on organizing instruction and study rates each recommendation by strength of evidence. Four of its recommendations translate directly into course builds:
- Retrieval practice: use quizzes to re-expose learners to key content (rated strong). Put short low-stakes questions after each module.
- Spacing: space learning over time (rated moderate). Schedule follow-up questions days or weeks later; the forgetting curve explains why.
- Worked examples: have learners alternate between reading worked solutions and solving problems themselves (rated moderate). In practice this fits best when a skill is new to the learner.
- Graphics plus words: combine graphics with verbal descriptions (rated moderate). Mayer’s multimedia principles turn this into specific slide and video rules.
Each of these applies to everyone in the room, so none needs a style diagnosis. They also lower the cost of design: you build one good version instead of three mediocre ones. To keep visuals from backfiring, check them against cognitive load theory, which asks whether each element helps learners or only adds extraneous load.
Treat transfer to your own workforce as something to verify. Run a pilot and measure delayed retention before you assume a classroom result carries over.
How do you translate the learning styles myth into course-build decisions?
Replace each style-based instinct with a content-based or evidence-based decision. The table below pairs six common requests with what the evidence supports and the concrete change to make in your next build. Rows marked as design judgment are practical inferences, not findings from the studies cited above.
| What the myth tells you to do | What the evidence supports instead | What to change in your next course build |
|---|---|---|
| Sort learners into visual, auditory and kinesthetic tracks | Matching to preference has not been shown to improve learning (Pashler et al., 2008; Rogowsky et al., 2015). | Remove the style quiz from onboarding. If you must route learners, use role or prior knowledge (design judgment). |
| Produce every lesson as video, audio and text | In the audiobook versus e-text study, preference did not significantly predict comprehension, immediately or at two weeks. | Build one strong primary version per objective. Add captions and transcripts for access, not for style, and spend the saved hours on practice items. |
| Put “visual learner” images on every slide | Combining graphics with verbal descriptions is rated moderate evidence, for all learners. | Use an image only when it carries content, such as a diagram, a process or a chart. Cut decorative clips. |
| Give “kinesthetic learners” hands-on activities | None of the sources here shows that hands-on work helps only one learner type. Retrieval practice and worked examples are supported for broad groups. | Choose the activity by the performance the job needs: simulate the task for physical skills, quiz recall for knowledge (design judgment). |
| Let each learner pick a preferred format | Preferences exist, but matching has not been shown to improve outcomes. | Offering choice is fine as a courtesy. Do not claim learning gains from it or make it your main investment. |
| Report “percent visual learners” as an audience insight | A self-reported preference is not a learning outcome. Practice testing and distributed practice are rated high utility. | Report a delayed retest score and the share of learners who completed spaced follow-up questions. |
Use the table as a build checklist. Before sign-off on the next course, check each row against the storyboard and remove anything justified only by “learning styles.”
What do you say when a leader asks for content for visual learners?
Treat “content for visual learners” as a request about a problem, not a specification. Agree to visuals where the content is visual, explain once that matching to learner type has not been shown to help, and offer what you will build instead. Keep it to about a minute and skip the lecture.
Because some people will hold the belief even after seeing the evidence, lead with what you will deliver rather than with the debunk. This script is a template, not a transcript of a real conversation.
Step 1: Agree with the need
“Yes, we will make this visual where the content calls for it. Can you tell me what you are seeing now that prompted the request? Is it dense slides, low completion, or a specific group struggling?”
Step 2: Reframe in one sentence
“The research has not shown that teaching people in their preferred style improves results, so I don’t want to split the budget across formats. Visuals help when they carry meaning, and I’ll use them for everyone.”
Step 3: Offer a concrete plan
“Here is what I will build: diagrams for the process steps, captions and transcripts for access, short quiz questions after each section, and follow-up questions two weeks later. I’ll measure retention on a delayed check.”
Step 4: Handle “but I am a visual learner”
“You probably do prefer visuals, and that preference is real. What has not been shown is that teaching to it improves results. If the pilot shows a gap, I’ll change the design.”
Pilot the Claim, Don't Argue It
Offer a two-week delayed retest in the pilot, the same design Rogowsky and colleagues used. Compare versions that differ in practice and spacing, not in learner type. Real retention numbers settle the debate better than citations do.
When is designing for visual or auditory needs the right call?
Design for visual or auditory needs when the content demands it or when learners have access needs. A floor-plan diagram for a safety walkthrough or a captioned video for a deaf learner is a content or accessibility decision. Neither depends on learning styles theory, and neither is threatened by the evidence above.
The distinction protects your accessibility work. Access requirements are about whether a learner can reach the material at all. Style claims are about whether matching preference improves learning. Keeping them separate stops a stakeholder from dropping captions because “learning styles are a myth.”
For practical checks on captions, transcripts and audio description, see the WCAG 2.2 and Section 508 accessibility guide.
What should you do next?
Start with one audit. Take your most recent course and mark every decision that was justified by a style: separate tracks, triplicated formats, decorative “visual learner” images, a style quiz. Use the table above to replace each one, and note how many production hours that frees up.
Then rewrite the brief for your next build around learner needs rather than learner types. Frameworks such as andragogy point you to questions you can actually answer: why the learner needs this, what they already know, and which job task it supports. Add a practice and spacing plan to the storyboard before you write a single slide.
Finally, keep the script handy and offer the pilot. Evidence for retrieval practice and spacing is stronger than the evidence for any style-based design, and a delayed retest gives your leaders data from their own workforce.
FAQ
Q1. What is the learning styles myth?
It is the belief that people learn better when instruction matches their preferred style, such as visual, auditory or kinesthetic. People do have format preferences, but the 2008 Pashler, McDaniel, Rohrer and Bjork review found almost no rigorous evidence that teaching to those preferences improves learning, and some studies contradicted the idea.
Q2. Is VARK a learning styles myth?
VARK is a questionnaire about preferences, and as of November 2026 its own site says it has never claimed that teaching should be matched to a learner’s style to improve outcomes. The myth lies in using VARK results to build separate tracks and expecting better learning, which the research has not shown.
Q3. Do students learn better when taught in their preferred style?
Matching has not been shown to help. In a 2015 study, college-educated adults were randomly assigned to hear a book as an audiobook or read it as e-text. Learning style preference did not significantly predict comprehension on an immediate test or on a test two weeks later.
Q4. Why do so many educators still believe in learning styles?
The idea feels true because people can describe their preferences, and it is easy to explain. A 2020 review of 37 studies covering 15,405 educators in 18 countries found a weighted average belief of 89.1%, ranging from 58% to 97.6%. The authors caution that most studies used convenience samples.
Q5. What works better than learning styles in corporate training?
Techniques with evidence across learners. Dunlosky and colleagues rated practice testing and distributed practice as high utility, and rereading and underlining as low. Practical moves include short quizzes after each module, follow-up questions spaced days or weeks later, worked examples for new skills, and graphics paired with words.
Q6. Is it wrong to use visuals in training?
No. Use visuals when they carry content, such as a process diagram or a chart, because combining graphics with verbal descriptions is rated moderate evidence by a US Institute of Education Sciences practice guide. The point is to choose visuals for the content and for all learners, not to assign them to learners labeled visual.
Q7. How should I respond when a stakeholder asks for content for visual learners?
Agree to use visuals where the content calls for them, then say in one sentence that matching to learner type has not been shown to improve results. Offer a concrete plan instead: diagrams, captions, quizzes after each section and a delayed retest in a pilot, so the discussion rests on your own retention data.
