Learning styles: does teaching to visual or auditory learners work?

You may have heard someone say, “I’m a visual learner, so I need to see it,” or “I’m an auditory learner, so I learn from lectures.” Learning styles are usually split into visual, auditory and kinesthetic (VAK), sometimes with read/write added as a fourth (VARK). Quizzes tell you your type and often say you learn better when taught that way. Researchers call that last claim the meshing hypothesis (Pashler et al., 2008).

The short answer is that people really do have preferences about how they take in information. Children and adults will state preferences if asked, and people also differ in specific abilities, such as picturing shapes in space (Pashler et al., 2008). But experiments designed to test whether teaching or studying in your preferred mode helps you learn more have mostly not found that benefit (Pashler et al., 2008; Aslaksen & Lorås, 2018). A 2024 meta-analysis reported a small average effect in favour of matched teaching (g = 0.31), but the pattern that would support matching appeared in only about a quarter of the outcome measures, and its authors judged the effect too small and too infrequent for wide adoption (Clinton-Lisell & Litzinger, 2024).

Meanwhile, some study techniques have helped learners of different ages and abilities: testing yourself with the book closed, and spreading reviews over several days (Dunlosky et al., 2013). That review did not look at learning styles separately, but the 2008 review noted that the memory benefit of testing applies, as far as can be told, to all learners rather than a small subset (Pashler et al., 2008). So a learning-style result is best used as a clue to what feels comfortable, alongside techniques whose effects have been checked.

This guide covers learning styles that sort people by the senses (seeing, hearing, doing) and the claim that teaching should match them. Tests sold as “learning type” assessments come in many kinds, and this guide does not assess them one by one. mind.io has no learning-style test. Nor is this a guide to blame people who enjoy such quizzes or teachers who have used learning styles in class. It covers what learning styles are, how the matching claim can be tested, the experiments that tested it, the debate over a 2024 meta-analysis, why so many people believe it, and study techniques worth recommending to anyone.

Learning styles are the idea that people differ in which kind of teaching or study works best for them. Learning-style questionnaires usually ask what kind of presentation you prefer, such as words, pictures or speech, and which activities feel most congenial (Pashler et al., 2008). The best known is VAK, for visual, auditory and kinesthetic; VARK adds read/write. The verbalizer–visualizer distinction, between thinking mainly in words or in pictures, is a similar idea (Clinton-Lisell & Litzinger, 2024).

There are far more models than most people realise. A systematic review commissioned in the UK identified 71 models and examined 13 influential ones in detail. It checked each against four minimum standards: internal consistency, test–retest reliability (similar results when tested again), construct validity (measuring what it claims to measure) and predictive validity (predicting actual performance). Only 3 of the 13 came close; the Dunn and Dunn model, which includes the sensory modalities, met one of the four. The authors advised against changing teaching on the basis of any learning-style instrument alone (Coffield et al., 2004). VARK was not one of the 13.

Few people fit one type neatly. In a study of fifth-graders at a public school in rural Pennsylvania, 11% of the 107 pupils analysed had a clear auditory preference and 21% a clear visual one, while 68% had similar auditory and visual preferences (Rogowsky et al., 2020). The VARK website itself says many people are “multimodal”, combining several approaches depending on the situation.

The VARK website describes VARK as a questionnaire to help you think about your learning preferences, and says it does not measure intelligence or predict success. It also says VARK does not claim that people can only learn when information is presented in their preferred modality, and that it acknowledges the debate over whether matching teaching to preferences improves outcomes (VARK website, accessed October 2026).

In 2008, four cognitive psychologists commissioned by a journal reviewed the evidence for using learning styles in education. They first set out what would count as evidence. People must be sorted by learning style; within each style, people must be randomly assigned to different teaching methods; everyone must take the same test; and the method that works best for one style group must differ from the method that works best for another. For example, visual learners would have to learn best from pictures and auditory learners from listening (Pashler et al., 2008).

By this standard, if pictures work better for everyone, that is not evidence for matching, because one method is best for all. On the other hand, one style group can score higher overall and the claim can still be supported, as long as the best method differs between the groups (Pashler et al., 2008). Researchers call this pattern a crossover interaction.

The review found only one study, from 1999, that could even potentially meet the standard, and judged it weak evidence: among other problems, only about a third of the students were classified into the groups. Several well-designed studies, by contrast, contradicted the matching idea. In one, help screens tailored to visual or verbal learners in an electronics lesson showed no matching benefit across nearly 20 measures of individual differences; in another, words shown as line drawings were remembered better than spoken words, regardless of a visual preference (studies described in Pashler et al., 2008).

The authors concluded: “at present, there is no adequate evidence base to justify incorporating learning-styles assessments into general educational practice.” Because sound studies were so few, they added, it would also be an error to conclude that every possible version of learning styles had been tested and found wanting; many had simply not been tested at all (Pashler et al., 2008).

The key distinction is between preference and benefit. The authors said the existence of fairly stable preferences is not in dispute, yet a student’s having a particular preference does not by itself mean that teaching should follow it. The same goes for abilities: differences such as spatial visualisation clearly exist, but they do not on their own support the learning-styles hypothesis (Pashler et al., 2008).

More experiments built to this standard followed. A study of college-educated adults measured auditory and visual-word preferences with a standardised learning-style inventory, then randomly assigned participants to study the same nonfiction material as a digital audiobook or an e-text. On a written test straight away and two weeks later, there was no statistically significant relationship between learning style and teaching method (from the abstract; Rogowsky et al., 2015). A later review, after searching six databases and screening more than 1,000 records, named it the only study fully aligned with the 2008 review’s criteria (Aslaksen & Lorås, 2018).

In 2020 the same team ran the design with fifth-graders at a public school in rural Pennsylvania. The pupils read some passages themselves and listened to recordings of others at the same level, then answered the same written questions. Of the 34 who could be classified (12 auditory and 22 visual learners), the visual group scored higher on both reading and listening, and both groups did slightly better when listening; the best method did not differ between the groups. Across the 107 pupils with both sets of scores, visual preference was not significantly correlated with reading scores (r = 0.07), and auditory preference was weakly negatively correlated with listening scores (r = −0.22) (Rogowsky et al., 2020).

The authors concluded that matching teaching to pupils’ auditory or visual style had no effect on their achievement. They also noted that the study took place in one school and used only written questions, so applying the results to other pupils or teaching methods calls for caution (Rogowsky et al., 2020).

Another study looked at how university students actually study. Of 426 undergraduates in an anatomy course who completed the VARK questionnaire and a study-strategy survey, most did not report strategies that matched their VARK result, and course performance was not correlated with any VARK score. Some specific strategies, such as using the virtual microscope, were linked to better grades whatever the VARK result, but whether strategies matched the VARK result made no difference (from the abstract; Husmann & O’Loughlin, 2019). It is a correlational study of one course.

When memory was measured directly in each sense, how well people remembered pictures, sounds and things learned by touch did not correlate with their self-reported learning style. People turned out to answer learning-style questionnaires from general memories and beliefs about themselves rather than specific experiences of learning (from the abstract; Krätzig & Arbuthnott, 2006).

A 2018 review found 10 studies that used a suitable method (only the 2015 study met the criteria fully). Pooling 22 effect sizes, matching visual teaching to visual learners (g = −0.09, n = 484) and auditory teaching to auditory learners (g = −0.27, n = 356) both had small, non-significant effects; the authors concluded that there is still no replicable statistical evidence for a matching benefit (Aslaksen & Lorås, 2018). A 2015 review likewise found that the more methodologically sound studies tended to refute the hypothesis (from the abstract; Cuevas, 2015).

In 2024 a meta-analysis pooled studies that compared teaching matched and unmatched to modality learning styles. Across 21 studies published in English (1,712 participants, 101 effect sizes), matched teaching came out slightly ahead on average, with a small but statistically significant effect of g = 0.31 (the figure in the paper’s abstract). The authors judged that publication bias did not appear to explain the effect (Clinton-Lisell & Litzinger, 2024).

They also warned against taking that number at face value. Support for matching requires both style groups to do better with teaching matched to their own style, and that happened in only 11 of 42 outcome measures (26%). Only 5 of the 21 studies met the What Works Clearinghouse quality standards, and most of the studies showing the crossover pattern did not. Every study covered a single session, and in 85% of them people whose learning style could not be classified confidently were left out (Clinton-Lisell & Litzinger, 2024).

The authors therefore wrote that it would be “too much of an overreach” to insist on building learning styles into teaching, and concluded that the benefits are “too small and too infrequent to warrant widespread adoption”. They advised extreme caution in using their result to justify matching, and recommended that anyone who uses learning styles do so within teaching that combines several modalities. For comparison, they noted a design change applied to all students whose effect is about twice as large (Clinton-Lisell & Litzinger, 2024).

Another group read the evidence differently. Looking at 17 meta-analyses on learning styles in their own database of education meta-analyses (Visible Learning MetaX), they found that the four that tested matching (143 studies) showed almost no effect (d = 0.04), while the correlational ones, which link learning-style scores with grades (average r = 0.24), say nothing about matching and often mix up styles with study strategies (Hattie & O’Leary, 2025). Two of those four date from the 1980s, not all of them concern the senses, and the 2024 meta-analysis is not among them.

The disagreement is not new. A 2005 meta-analysis reported that its results “overwhelmingly supported” matching instruction to preferences in the Dunn and Dunn model (from the abstract; Lovelace, 2005); a 2007 critique pointed out that 96% of its studies were dissertations, 70% of which were linked to the model’s authors, raising the possibility of “home-team” bias (from the abstract; Kavale & LeFever, 2007). The 2004 UK review summed up the matching evidence as “equivocal at best and deeply contradictory at worst” (Coffield et al., 2004).

In short, research has not shown that every version of learning styles has exactly zero effect. But the evidence that matching improves learning is weak and inconsistent, and much of it comes from short, low-quality studies. Even the authors of the meta-analysis that found an effect did not see it as grounds for wide adoption. The 2008 review had already noted that such interactions would need to be large and robust, not just statistically significant, to justify the cost of testing pupils and tailoring teaching (Pashler et al., 2008).

Learning styles are an idea that teachers have widely endorsed in surveys. In a survey of 137 teachers in the Dorset region of the UK and 105 teachers around Amsterdam in the Netherlands, 93% of the UK teachers and 96% of the Dutch teachers agreed that “individuals learn better when they receive information in their preferred learning style (e.g., auditory, visual, kinesthetic)”. In the same survey, 95% and 82% correctly agreed that learners show preferences for how they receive information, and 98% and 64% had encountered learning styles in their schools. These were teachers who chose to take part out of an interest in the brain and learning (Dekker et al., 2012).

In a US online study of 3,877 people, the same statement was judged true by 93% of the general participants, 76% of educators and 78% of people who had taken many neuroscience courses; the statement that “children have learning styles that are dominated by particular senses (i.e., seeing, hearing, touch)” was judged true by 88%, 71% and 68%. All were volunteers with at least some college education (Macdonald et al., 2017).

A systematic review of 37 studies covering some 15,000 educators in 18 countries, from 2009 to early 2020, found a weighted 89.1% agreeing that people learn better when taught in their preferred style (58% to 97.6% across studies), and 95.4% among trainee teachers. Most studies used convenience samples, though, and the reviewers were not sure participants understood they were being asked about matching instruction. In four studies that explained the lack of evidence, agreement fell from 78.4% to 37.1%. The reviewers added that concern about this belief may not be fully supported by current evidence (Newton & Salvi, 2020).

Surveys in Korea found similar beliefs (the groups surveyed differ, so the figures cannot be compared directly with those above). Among 521 Korean pre-service teachers (269 elementary and 252 secondary education students), fewer than 1% answered the items on visual, auditory and kinesthetic learning styles correctly (from the abstract; Park et al., 2016). In a survey of 83 Korean pre-service elementary teachers published in 2026, 91.6% accepted the learning-styles neuromyth (from the abstract; Lim, 2026).

Why is it so widely believed? For one thing, it is not made up from nothing: visual, auditory and kinesthetic information is processed in different parts of the brain, though those areas are closely interconnected (Dekker et al., 2012). Preferences are real too, but they can shape how learning feels more than how much is remembered. In one experiment, people with higher visualizer scores felt they were learning pictures better, yet they did not recall more (from the abstract; Knoll et al., 2017).

Our intuitions about how we learn often miss the mark. Methods that feel easy and familiar give a sense of learning, while methods that feel slower, such as spacing reviews or mixing problem types, often last longer (Pashler et al., 2008). The authors of the 2008 review also suggested other reasons for the idea’s appeal: the wish to find out “what type of person one is”, and to be treated as an individual (Pashler et al., 2008).

Some researchers worry that labelling learners can pigeonhole them or waste teachers’ effort, while advocates argue that knowing one’s style can be motivating (Newton & Salvi, 2020). Teachers who mix pictures, talk and activities in one lesson may well be helping students, since combining modalities can benefit learning; one research team described this as good practice “for the wrong reasons” (Macdonald et al., 2017). Having believed in learning styles is nothing to be embarrassed about. The left- and right-brain idea, asked about in the same teacher surveys, is covered in the left-brain/right-brain guide.

A review that rated ten easy-to-use study techniques on the research evidence gave a “high utility” rating to practice testing and distributed practice, because they helped learners of different ages and abilities and improved performance on many kinds of test and even in real classrooms. Elaborative interrogation, self-explanation and interleaved practice were rated “moderate utility”; summarization, highlighting, the keyword mnemonic, imagery for text and rereading were rated “low utility” (Dunlosky et al., 2013). How to use each technique, with a one-week study plan, is in the MBTI study tips guide.

A meta-analysis of 222 classroom studies (48,478 students) also found that quizzing raised achievement to a medium extent (g = 0.499) (from the abstract; Yang et al., 2021). A review of 317 experiments on spaced study in verbal recall tasks found that the longer you need to remember something, the longer the best gap between study sessions (from the abstract; Cepeda et al., 2006).

When choosing a mode, research points to the content rather than the person. The 2008 review stressed that it was not claiming one kind of teaching suits all content and all learners: a writing course probably needs a heavy verbal emphasis, while geometry needs visual–spatial materials (Pashler et al., 2008). Forming mental images of what you read helped mainly with material that is easy to picture and on memory tests (Dunlosky et al., 2013). An overview of 29 reviews (1,189 studies) found that well-designed multimedia, such as placing text next to the picture it describes or signalling the key points, helped learning (from the abstract; Noetel et al., 2022). Those designs give everyone the same materials, which is different from matching teaching to each learner’s style (Clinton-Lisell & Litzinger, 2024).

If you have taken a learning-style quiz, here are some ways to use the result. These are suggestions in light of the research above, not tested methods. First, treat the result as a clue to what feels comfortable and less tiring; the VARK website also presents results as a starting point for reflecting on how you learn rather than a fixed label. Second, whatever format you start in, add a step where you recall with the book closed and a review a few days later: after a lecture, write down what you remember on a blank page; after studying a diagram, cover it and redraw it. Third, don’t avoid the formats you prefer less. As a point quoted in the 2004 UK review puts it, even if matching improves performance, it does nothing to prepare you for later tasks that don’t match your preferred style (Coffield et al., 2004).

Q. I’m a visual learner. Can I study with pictures and videos alone? A. When experiments that matched visual teaching to visual learners were pooled, there was no clear benefit (g = −0.09) (Aslaksen & Lorås, 2018). If pictures feel comfortable, starting with them is fine, but add self-testing, such as covering a diagram and recalling it, and reviews spread over several days (Dunlosky et al., 2013).

Q. Can I trust my VARK result? A. Read it as a picture of what you prefer. The VARK website itself says VARK does not measure intelligence or predict success. In a study of 426 anatomy undergraduates, VARK scores were not related to grades, and most students did not study in ways that matched their result (Husmann & O’Loughlin, 2019). In the UK review, 3 of 13 major learning-style instruments came close to the minimum standards, but VARK was not among those reviewed (Coffield et al., 2004).

Q. My child prefers listening. Should I teach to an auditory style? A. In the study of fifth-graders, listening did not help the children classified as auditory learners any more than the others: both groups did slightly better when listening, and the children classified as visual learners scored higher on both reading and listening. The authors also suggested that children who prefer listening may need more reading practice, not less, but this is their interpretation rather than something the study tested (Rogowsky et al., 2020).

Q. Does using several senses at once help? A. Well-designed materials that combine words and pictures have helped learning (Noetel et al., 2022). That gives everyone the same materials, though, which is different from matching one sense to each learner (Clinton-Lisell & Litzinger, 2024).

Q. Were teachers who used learning styles in class wrong? A. That is not the right way to see it. Many teachers said they believed in learning styles, but it is unclear how they understood the question (Newton & Salvi, 2020), and mixing several approaches in one lesson can help learning for other reasons (Macdonald et al., 2017).

Q. Could learning styles still turn out to be right? A. The door is not completely closed. Not every version of learning styles has been tested (Pashler et al., 2008), and a 2024 meta-analysis reported a small average effect (g = 0.31). Its authors, however, judged the effect too small and too infrequent to justify wide adoption (Clinton-Lisell & Litzinger, 2024).

Q. How is this different from studying by MBTI type? A. The MBTI was not built as a learning-style test. The MBTI study tips guide uses preferences as cues for studying with less strain and more consistency, then adds the same self-testing and spaced review as this guide. To see your personality traits as scores, you can take the Big Five test.

References: Pashler, H., McDaniel, M., Rohrer, D., & Bjork, R. (2008), Learning styles: concepts and evidence, Psychological Science in the Public Interest 9(3), 105–119, https://doi.org/10.1111/j.1539-6053.2009.01038.x; Aslaksen, K., & Lorås, H. (2018), The modality-specific learning style hypothesis: a mini-review, Frontiers in Psychology 9, 1538, https://doi.org/10.3389/fpsyg.2018.01538; 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, https://doi.org/10.3389/fpsyg.2024.1428732; Dunlosky, J., Rawson, K. A., Marsh, E. J., Nathan, M. J., & Willingham, D. T. (2013), Improving students’ learning with effective learning techniques: promising directions from cognitive and educational psychology, Psychological Science in the Public Interest 14(1), 4–58, https://doi.org/10.1177/1529100612453266; Coffield, F., Moseley, D., Hall, E., & Ecclestone, K. (2004), Learning styles and pedagogy in post-16 learning: a systematic and critical review, London: Learning and Skills Research Centre; Rogowsky, B. A., Calhoun, B. M., & Tallal, P. (2020), Providing instruction based on students’ learning style preferences does not improve learning, Frontiers in Psychology 11, 164, https://doi.org/10.3389/fpsyg.2020.00164; VARK, About VARK and Frequently asked questions (web pages, accessed 26 October 2026), https://vark-learn.com/about-vark/ and https://vark-learn.com/about-vark/frequently-asked-questions/; 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, https://doi.org/10.1037/a0037478; Husmann, P. R., & O’Loughlin, V. D. (2019), Another nail in the coffin for learning styles? Disparities among undergraduate anatomy students’ study strategies, class performance, and reported VARK learning styles, Anatomical Sciences Education 12(1), 6–19, https://doi.org/10.1002/ase.1777; Krätzig, G. P., & Arbuthnott, K. D. (2006), Perceptual learning style and learning proficiency: a test of the hypothesis, Journal of Educational Psychology 98(1), 238–246, https://doi.org/10.1037/0022-0663.98.1.238; Cuevas, J. (2015), Is learning styles-based instruction effective? A comprehensive analysis of recent research on learning styles, Theory and Research in Education 13(3), 308–333, https://doi.org/10.1177/1477878515606621; 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(2), 31, https://doi.org/10.1007/s10648-025-10002-w; Lovelace, M. K. (2005), Meta-analysis of experimental research based on the Dunn and Dunn model, The Journal of Educational Research 98(3), 176–183, https://doi.org/10.3200/JOER.98.3.176-183; Kavale, K. A., & LeFever, G. B. (2007), Dunn and Dunn model of learning-style preferences: critique of Lovelace meta-analysis, The Journal of Educational Research 101(2), 94–97, https://doi.org/10.3200/JOER.101.2.94-98; Dekker, S., Lee, N. C., Howard-Jones, P., & Jolles, J. (2012), Neuromyths in education: prevalence and predictors of misconceptions among teachers, Frontiers in Psychology 3, 429, https://doi.org/10.3389/fpsyg.2012.00429; Macdonald, K., Germine, L., Anderson, A., Christodoulou, J., & McGrath, L. M. (2017), Dispelling the myth: training in education or neuroscience decreases but does not eliminate beliefs in neuromyths, Frontiers in Psychology 8, 1314, https://doi.org/10.3389/fpsyg.2017.01314; Newton, P. M., & Salvi, A. (2020), How common is belief in the learning styles neuromyth, and does it matter? A pragmatic systematic review, Frontiers in Education 5, 602451, https://doi.org/10.3389/feduc.2020.602451; Park, S.-W., Park, J.-S., Lee, S.-Y., & Shin, J.-H. (2016), Prevalence and predictors of neuromyth among pre-service teachers [in Korean], Gyowon Gyoyuk 32(2), 185–212, https://www.earticle.net/Article/A270667; Lim, C.-S. (2026), Preservice elementary teachers’ perspectives on neuroscience: analyzing the association between neuromyths, brain knowledge, and science educational application [in Korean], Journal of Korean Elementary Science Education 45(1), https://www.riss.kr/link?id=A110147520; Knoll, A. R., Otani, H., Skeel, R. L., & Van Horn, K. R. (2017), Learning style, judgements of learning, and learning of verbal and visual information, British Journal of Psychology 108(3), 544–563, https://doi.org/10.1111/bjop.12214; Yang, C., Luo, L., Vadillo, M. A., Yu, R., & Shanks, D. R. (2021), Testing (quizzing) boosts classroom learning: a systematic and meta-analytic review, Psychological Bulletin 147(4), 399–435, https://doi.org/10.1037/bul0000309; Cepeda, N. J., Pashler, H., Vul, E., Wixted, J. T., & Rohrer, D. (2006), Distributed practice in verbal recall tasks: a review and quantitative synthesis, Psychological Bulletin 132(3), 354–380, https://doi.org/10.1037/0033-2909.132.3.354; Noetel, M., Griffith, S., Delaney, O., Harris, N. R., Sanders, T., Parker, P., del Pozo Cruz, B., & Lonsdale, C. (2022), Multimedia design for learning: an overview of reviews with meta-meta-analysis, Review of Educational Research 92(3), 413–454, https://doi.org/10.3102/00346543211052329