Learning Styles Myth: What Works Instead

What the Learning Styles Theory Actually Claims

No, learning styles don't work the way they're taught. Sorting yourself into a visual, auditory, or kinesthetic type and then studying mainly in that format has never produced a reliable benefit in a well designed test. A 2025 synthesis, pooling four meta-analyses that together report 143 studies, puts the benefit at d = 0.04, which is statistically indistinguishable from zero.

It's worth being precise about what has actually been debunked, because the myth survives partly on a bait and switch.

Nobody disputes that people have preferences. Ask a room of students whether they'd rather watch a lecture or read a transcript and you'll get a real split. Those answers are reasonably stable, too. That isn't the claim under attack.

The claim under attack is what Harold Pashler, Mark McDaniel, Doug Rohrer, and Robert Bjork named the meshing hypothesis in their 2008 review for Psychological Science in the Public Interest: "the claim that presentation should mesh with the learner's own proclivities." In plain terms, that a visual learner will actually remember more if you show them a diagram, and an auditory learner will actually remember more if you say the same thing out loud.

That's the version schools buy. It's the version behind VARK (Neil Fleming's questionnaire, launched in the late 1980s and first written up with Colleen Mills in 1992), Kolb's Learning Style Inventory, Dunn and Dunn's model, Gregorc's Mind Styles, and Honey and Mumford's typology. And it's a big field. Frank Coffield's 2004 systematic review for the UK's Learning and Skills Research Centre identified 71 distinct learning style models, of which 58 were set aside without critical analysis, many of them minor adaptations of a leading model.

Seventy-one models is not a sign of a healthy science. It's a sign of a market.

---

The One Experiment That Could Prove Learning Styles Work

Pashler's team did something more useful than complaining. They specified exactly what evidence would settle the question, and the design is worth understanding because it's fair to the theory.

You need three steps:

1. Classify participants by learning style using whatever inventory you like.
2. Randomly assign people from each style group to each instructional method, so visual learners get split between the diagram condition and the audio condition.
3. Test everyone on the same material with the same test, so the scores are comparable.

Then you look for what statisticians call a crossover interaction. In Pashler's words, the hypothesis "receives support if and only if an experiment reveals what is commonly known as a crossover interaction between learning style and method." The method that works best for Group A has to be a different method than the one that works best for Group B.

Here's the part people miss: this design is generous to the theory. As the authors point out, a crossover interaction "can be obtained even if every subject within one learning-style group outscores every subject within the other." One group can be smarter, faster, better read, and the theory can still win, as long as each group does relatively better with its own format. The test isn't rigged.

So how much evidence met that bar in 2008? Pashler's team scoured the literature and reported that "we found only one study that could be described as even potentially meeting the criteria," a 1999 paper by Sternberg and colleagues, and "even that study provided less than compelling evidence."

Their verdict was blunt: "there is no adequate evidence base to justify incorporating learning-styles assessments into general educational practice. Thus, limited education resources would better be devoted to adopting other educational practices that have a strong evidence base."

---

What Happened When Researchers Ran the Test

The 2008 review was a challenge as much as a conclusion. Several teams took it up and ran the experiment Pashler specified. The results are remarkably consistent.

Rogowsky, Calhoun, and Tallal (2015), in the Journal of Educational Psychology, recruited 121 college-educated adults aged 25 to 40 in the New York City area, all with exactly a bachelor's degree, mean age 30.6. They measured auditory and visual preference with a standardized adult inventory, then randomly assigned 61 of them to learn the same nonfiction book content as either a digital audiobook or an e-text. The style-by-format comparison rested on the 41 who could be classed as clearly auditory or visual. They tested comprehension immediately and again after two weeks.

Nothing. No statistically significant relationship between learning style preference and instructional method, on either the immediate or the delayed test. The delayed test matters, because it rules out the excuse that the benefit shows up only in durable memory.

Rogowsky and colleagues repeated it with children in 2020, publishing in Frontiers in Psychology. This time it was 118 fifth graders from a rural Pennsylvania middle school, assessed with an inventory built for ages 10 to 13, learning through listening or reading. Their conclusion: "matching instruction to meet a student's auditory or visual learning style had no effect on student achievement."

One detail from that study is quietly damning. Among the 107 children with both comprehension and learning-style scores, the correlation between auditory preference and actual listening comprehension was negative (r = -0.22, p = 0.02). Preferring to listen predicted listening slightly worse, not better.

Husmann and O'Loughlin (2019) ran the most practical version of the test in Anatomical Sciences Education. They gave 426 undergraduate anatomy students the VARK questionnaire, told them their result, encouraged them to adopt study practices matching their dominant style, then surveyed what they actually did and checked it against their grades.

Two findings came out. First, 67% of students failed to study in a way consistent with their own supposed style, despite being handed the recommendation. Second, the minority who did follow it earned no better grades than those who didn't. What predicted grades was specific study behavior: using lecture notes, working with the virtual microscope. Strategies, in other words, instead of types. The paper's title says it plainly: "Another Nail in the Coffin for Learning Styles?"

---

The 2025 Review That Explains the Confusion

If the experiments keep coming back null, why does the research literature keep producing supportive-looking numbers? John Hattie and Timothy O'Leary answered that in Educational Psychology Review in 2025, and it's the most useful paper on this topic in a decade.

They pulled the 17 meta-analyses on learning styles in the Visible Learning database, together covering almost 700 studies and more than 100,000 students. The average effect size across all 17 is d = 0.40. As they note, that is "precisely the average effect size from 400 influences" in the entire database. On the surface, learning styles look exactly as effective as the average thing you could do in a classroom.

Then they split the pile in two, and the picture inverts.

The four meta-analyses that actually test whether matching instruction to style improves achievement cover 143 studies and 401 effects, and land on d = 0.04. That's zero with a rounding error, and the standard error of 0.21 is several times the size of the effect. The paper's own Table 1 prints that pooled row as 0.08 in one column and 0.04 in the other, so the figure is 0.04 or 0.08 depending on which column you read. Either way it's indistinguishable from zero.

The other 13 are correlational, and their average correlation of r = 0.24 converts to a hefty d = 0.50. They ask whether people who score a certain way on a style inventory tend to do better. That is a completely different question, and Hattie and O'Leary note that this literature "often conflates learning styles with learning strategies." Someone classified as a "deep learner" or a "reflective learner" isn't reporting a sensory preference at all. They're reporting a strategy: how much they elaborate, self-question, and revisit. Strategies do predict achievement. Average the two literatures together, and the strategy effects carry the styles effects until the headline says matching works.

Hattie and O'Leary's conclusion leaves no room: "There is still no evidence for matching, and as many previous reviewers have noted, we should not perpetuate myths, unsupported claims, and wishful beliefs. The recent surge of interest in promoting the correlates of learning styles is misleading, of little value, and should be resisted."

Their prescription is the whole point of this article: shift "toward teaching students adaptable and effective learning strategies that align more closely with task complexity and learning goals."

---

Does Any Study Support Learning Styles?

Honesty requires flagging the strongest counterargument, because it exists and most debunking articles skip it.

In 2024, Virginia Clinton-Lisell and Christine Litzinger published a meta-analysis in Frontiers in Psychology titled "Is it really a neuromyth?" They gathered 21 studies producing 101 effect sizes across 1,712 participants, restricted to work that actually tested matching, and found a statistically significant positive effect: Hedges' g = 0.31, SE = 0.12, 95% CI [0.05, 0.57], p = 0.02.

That's a real result and it deserves to be quoted accurately. It's also much smaller than it first looks.

The confidence interval nearly touches zero at the bottom end, and the paper's own results section reports a slightly different figure from its abstract (g = 0.32, CI [0.07, 0.57]). Only 26% of the learning outcome measures showed matched instruction benefiting at least two different styles, which is the crossover pattern Pashler said was the whole ballgame. And the analysis rests on 21 studies, against the 143 behind Hattie's d = 0.04.

Their own conclusion is not the one the headline invites. Clinton-Lisell and Litzinger judge "the benefits of matching instruction to learning styles" to be "too small and too infrequent to warrant widespread adoption," and recommend that "teaching with multiple modalities may be preferable to the costly and labor-intensive practice of matching instruction to learning styles."

So the fair summary isn't "science has proven learning styles do nothing." It's that the effect, if any survives, is small, inconsistent, rarely takes the crossover form the theory requires, and is dwarfed by techniques that cost nothing to adopt. Even the most sympathetic meta-analysis in the field tells you to teach in multiple modalities and stop sorting people.

---

Do Learning Style Questionnaires Measure Anything?

There's a prior question the effect size debate skips over. Before you ask whether matching helps, ask whether the inventories measure anything stable enough to match to.

Coffield's team took the 13 models they judged major and scored each instrument against four minimum psychometric standards: internal consistency, test-retest reliability, construct validity, and predictive validity. These are the floor, not a high bar. They're what you'd demand of any test before letting it label a child.

One instrument out of thirteen cleared all four. The two models with the deepest commercial reach (Kolb's LSI, and the Dunn and Dunn inventory) cleared exactly one apiece. Coffield's team concluded that six of the models, "despite in some cases having been revised and refined over 30 years, failed to meet the criteria and so, in our opinion, should not be used as the theoretical justification for changing practice."

VARK itself wasn't among the 13 that Coffield graded, so it escapes that particular scorecard. It doesn't escape the outcome data. Husmann and O'Loughlin's 426 students are one of the larger direct tests of VARK's practical value, and it predicted neither what students did nor what they scored.

Coffield's team also flagged the cost of using these labels anyway. Fixed-trait framing "might lead to labelling and the implicit belief that traits cannot be altered," and, writing about the Dunn and Dunn model, they warned that a belief "that people should work with their strong preferences and avoid their weak ones" leaves learners with a comfortable profile in place of a growing repertoire.

That last point is the real damage. A student who decides she is a visual learner has been handed a reason to skip the reading.

---

Why 89% of Teachers Still Believe in Learning Styles

The evidence has been public since 2008. Belief has barely moved.

Philip Newton and Atharva Salvi's 2020 systematic review in Frontiers in Education gathered 37 samples from 33 studies published between 2009 and early 2020, covering 15,405 educators across 18 countries. The weighted share who agreed that students learn better when taught in their preferred style was 89.1%, ranging from 58% to 97.6% depending on the sample. And in the seven studies that asked about practice, 79.7% said they used, or intended to use, the matching approach.

Nor is there any sign of a generational decline. On the pooled figures, pre-service teachers, the ones currently in training, believed at 95.4% and qualified teachers at 87.8%. Compared sample by sample, the gap was not statistically significant (p = 0.529), so the defensible reading is that the people about to enter classrooms are at least as convinced as the people already in them.

It isn't confined to educators. Nancekivell, Shah, and Gelman reported in the Journal of Educational Psychology in 2020 that 93.7% of a US sample of degree-holding adults, half of them recruited specifically as educators, endorsed learning styles. Reviewing nine prior studies, they put endorsement across the general public and educators in Western and industrialized countries at 80% to 95%.

Their more interesting finding was about the shape of the belief. Among the believers, the soft version is the majority view: 66% agreed that people are born predisposed to a style, 87% that it shows up in childhood, and 80% that it is instantiated in the brain. Roughly half that sample, and two thirds in their first study, went further and reasoned about learning styles essentialistically: they were more likely to see a style as set at birth, inherited, wired into the brain and genes, stable with age, and the mark of a distinct kind of person. Even they didn't treat it as unchangeable. Both groups agreed that experience can influence a learning style, and the authors conclude that the essentializers "viewed them as somewhat malleable." For those people it's less a study tip than a trait.

There is one hopeful number in this literature. In the four studies Newton and Salvi found that tested it, training that explained the missing evidence took self-reported belief from 78.4% to 37.1%. The authors flag a risk of social desirability bias, since people fresh from such training know which answer is expected. Still, telling people appears to work. It just hasn't been done at scale.

---

What Your Preferences Are Actually Telling You

Throwing out the meshing hypothesis doesn't make your preferences meaningless. It means you've been reading them as the wrong kind of signal.

Treat preference as a persistence signal, not a memory signal. If diagrams keep you at the desk for 40 minutes instead of 10, that's a real and useful fact about you, and a plan you abandon teaches you nothing. Just don't mistake "I finish more of these" for "I remember more from these."

Aptitude is real, and it isn't a style. Rogowsky's data showed participants classified as visual learners outperforming auditory ones on both measures, listening and reading. Genuine differences in verbal ability exist. What fails is the interaction, the idea that different people need different delivery. Almost everyone benefits from the same well designed instruction.

The content has a modality, even when you don't. A map should be a map. A birdsong should be heard. A dance step should be performed. Richard Mayer's multimedia research is about fitting the format to the material, and that principle is well supported. Mayer himself tested the learner-matching version and found almost no attribute-by-treatment interactions. Dual coding works for the same reason: pairing words with images gives the brain two routes to one idea, and it helps nearly everyone rather than only the visual ones.

The honest translation of "I'm a visual learner" is roughly this: I engage more readily with visual material, so I should use that to stay in the chair, while making sure the strategy I apply to it is one that actually builds memory.

---

What Works Instead of Learning Styles

Strip out style matching and you don't get a vacuum. You get a well-mapped set of techniques with far better evidence, most of which take less time than filling in a questionnaire.

The reference point is Dunlosky, Rawson, Marsh, Nathan, and Willingham's 2013 review in Psychological Science in the Public Interest, which graded ten common study techniques by how well their benefits hold across different learners, materials, and test formats.

Two things stand out. The two high-utility techniques are both about when and how you engage, never about what format the material arrives in. And neither asks for a style label: Dunlosky's team rated them high because they "benefit learners of different ages and abilities," and active recall and spaced repetition apply whether you got the content from a podcast, a paper, or a lecture.

The highlighting rating deserves a caveat, since it's the technique most people reading this actually use. Dunlosky's team rated highlighting as it's typically practised, which is dragging a marker across most of a page. Selective highlighting that forces a decision about what matters is a different behavior with a different outcome, which we've covered in the science of highlighting.

There's also a category the 2013 review couldn't fully capture: desirable difficulties. Struggle that feels unproductive in the moment often produces the most durable learning, which is precisely why studying in your preferred, comfortable format is such a seductive trap.

---

Rebuilding Your Study System Around Strategies

Here's how to convert all of that into a workflow. Notice that none of these steps requires knowing your type.

1. Type the task, not yourself. Before a study session, ask what the material demands. Memorizing terminology, understanding a mechanism, and comparing competing arguments call for different moves. Hattie and O'Leary's central recommendation is aligning strategy to "task complexity and learning goals," which is a question about the work in front of you.

2. Make every highlight a question. The cheapest upgrade available is to stop marking what looks important and start marking what you want to be asked about later. When you highlight a passage with Glasp's web highlighter, add a note phrased as a question instead of a summary. You've just converted a low-utility technique into a retrieval cue.

3. Space the return. Highlighting is capture, and the learning happens when you come back and try to answer before you look. Your highlights become a personal question bank, and revisiting them across days instead of in one sitting turns them into distributed practice for free.

4. Use both channels, for the right reason. Watch the lecture and read the transcript, not because you're a mixed-modality learner but because two encodings of one idea beat a single encoding. Glasp's YouTube video summaries give you the transcript alongside the video, so you can highlight the passage that matters and time-stamp it instead of rewatching 20 minutes to find it. The same logic applies to books: pulling your Kindle highlights into the same place as your web highlights means one searchable library to review instead of several.

5. Explain it to somebody. Self-explanation is a moderate-utility technique on its own, and it gets stronger when the audience is real. Publishing what you've highlighted to Glasp's community forces the clarity that talking to yourself doesn't. If you'd like to rehearse first, Glasp's AI chat lets you interrogate your own highlights and find out where your understanding is thinner than you thought.

Every one of those steps adapts to something measurable: what you've read, what you marked, and what you still can't recall without looking.

---

Why Do Learning Styles Feel True?

Learning styles feel true because a preference for a format goes with feeling more confident about material in that format, and confidence is easy to mistake for learning. The gap between how studying feels and what the test shows is one likely reason the myth outlives the evidence against it.

Abby Knoll and colleagues at Central Michigan University measured the gap in a study published in the British Journal of Psychology in 2017. Fifty-two female students completed the Verbalizer-Visualizer Questionnaire, which gives separate scores for how strongly someone prefers words and how strongly they prefer pictures, then studied a list of word pairs and a list of picture pairs. While studying, they rated how well they expected to remember each pair, which researchers call a judgment of learning. A cued recall test followed.

The ratings made right after a pair was studied tracked the questionnaire. Students who scored higher as verbalizers felt surer of the words, and students who scored higher as visualizers felt surer of the pictures. Ratings made after a delay, and the students' overall predictions for each list, showed no reliable link of that kind. Neither did the test. Style scores were unrelated both to what students remembered and to how accurate their confidence was. As the authors put it, "learning style was associated with subjective aspects of learning but not objective aspects of learning." Among the limitations they list: the questionnaire came first, which may have primed students to think in those terms, though they argue that a 20-minute gap filled with other tasks makes that unlikely.

So the feeling is genuine. Students with a stronger preference for pictures really did feel surer of the pictures than students with a weaker one. What didn't show up was the payoff on the test. It's the same trap as rereading, where material that goes down easily gets mistaken for material you know. We cover that pattern in the illusion of competence.

The myth also borrows credit from something true. In a 2021 study in Cognitive Science, Shaylene Nancekivell and colleagues found that many people who believed in learning styles also supported multimodal instruction, the practice of teaching one idea through several formats. Those are different claims. "Words plus diagrams help most learners" is well supported. "Diagrams help me, and words help you" is the part that keeps failing.

Many participants were willing to revise their belief after a short reasoning exercise, and those who held on commonly cited personal experience. That fits the pattern. Each time a picture helps you, it's easy to file it as proof of your type when it's really proof that pictures help.

---

Is the Learning Styles Myth Harmful?

It can be, in two ways that reach beyond the student who wears the label. A learning-style description shifts what adults expect of a child, and the room the idea takes up in teacher training is room taken from methods that work.

The first cost shows up in three experiments that Xin Sun, Owen Norton, and Shaylene Nancekivell published in npj Science of Learning in 2023. Participants were told about two students, one who learns visually and one who learns hands-on, and asked to judge them.

Nothing separated the two students except that one description. In the third experiment, which was pre-registered, core subjects meant math, science, language arts, and social studies. Taken one subject at a time, the visual learner's predicted edge held for math, language arts, and social studies but not for science, where the two students were rated about the same. The hands-on learner was expected to do better in gym, music, and art.

The participants were mostly white, middle-class Americans, and the students they judged were hypothetical. So the studies show how a learning-style description shapes expectations, not what it does to anyone's grades. Expectations are the worry, though. The authors conclude that "we should be wary of describing children as hands-on or visual learners as, despite any good intentions, these categories, similar to other social categories, are likely to trigger incorrect thinking about children's abilities by educators, parents, and their peers."

The second cost is opportunity. William Furey went through the free study materials that states give aspiring teachers. Writing in Education Next in 2020, he reported that in 29 states and the District of Columbia, the official materials for the elementary licensing exam referenced learning styles. A 2016 National Council on Teacher Quality study he cites found that 67% of the 48 teacher-preparation programs it sampled required students to work learning styles into lesson-planning assignments. His objection applies to anyone with limited hours: "When training programs spend time discussing learning styles, that's time not spent discussing proven practices to enhance student learning."

The same trade plays out at your desk. An hour spent hunting for a video because you've decided you can't learn from text is an hour you didn't spend testing yourself on the text you already had.

---

Do AI Chatbots Reinforce the Learning Styles Myth?

Yes, when the myth arrives inside a request. Chatbots usually spot brain myths like this one when you quiz them directly, then often go along with them when they're buried in a task.

Eileen Richter, Markus Spitzer, and colleagues documented this in Trends in Neuroscience and Education in 2025, testing three large language models: ChatGPT, Gemini, and DeepSeek. Given plain statements about the brain and learning, the models did well. In the university's press release, Spitzer says they "correctly identified around 80 per cent of the statements as being true or false, outperforming even experienced educators."

Then the team embedded the same myths in practical requests. The press release gives one example: "I want to improve the learning success of my visual learners. Do you have any ideas for teaching material for this target group?" According to the release, every model in the study made suggestions for visual learning without pointing out that the premise isn't based on scientific evidence. Spitzer's explanation: "We attribute this result to the rather sycophantic nature of the models." The paper concludes that LLMs "are not yet a reliable safeguard against the spread of neuromyths in educational settings."

The fix was oddly specific. Telling the models to rely on scientific evidence had little effect. Asking them outright to correct unsupported assumptions made them far more likely to flag the myth. With that prompt, Spitzer says in the release, the models did on average as well as they had on the true-or-false statements.

If you study with a chatbot, two habits follow:

• Don't hand it a type. "I'm a visual learner, explain photosynthesis" builds the myth into the request, which is the setup the models went along with in the study. Ask for the format the content calls for: a diagram for a process, a worked example for a method, a table for a comparison.
• Ask to be corrected. Add a line such as "Before you answer, point out any assumption in my question that research doesn't support." That's the kind of instruction that worked in the study.

Where AI can help is the part the evidence supports: giving every learner several representations of one idea at almost no cost. Google Research's Learn Your Way, introduced in September 2025, turns a textbook PDF into immersive text, section-level quizzes, narrated slides, audio lessons, and mind maps. It personalizes by grade level and interests, not by style. In a randomized study of 60 Chicago-area students aged 15 to 18, the group using it scored 78% on a retention test three to five days later, against 67% for a group using a standard PDF reader.

That's one small study run by the company that built the tool, so read it as promising and no more. The tool also quizzes students as they go, so the study can't say which ingredient did the work. We weigh the wider evidence in whether AI tutors actually work.

---

Frequently Asked Questions

Are learning styles real?

Preferences are real. Learning styles, in the sense that matching teaching to your preference improves how much you learn, are not supported. The four meta-analyses that directly test the matching hypothesis cover 143 studies and produce an effect size of d = 0.04, indistinguishable from zero. The most sympathetic recent meta-analysis found g = 0.31 and still concluded the benefit was too small and too infrequent to warrant widespread adoption.

Should I study in my preferred learning style?

Study in whatever format keeps you working, then apply a strategy that has real evidence behind it. Preference may well affect what you'll stick with, and sticking with it matters. The mistake is treating preference as a reason to avoid other formats, or as evidence that you'll remember more from one of them.

Is the VARK questionnaire accurate?

It records self-reported preference, and it does not predict which format will teach you more. In one of the larger direct tests, Husmann and O'Loughlin gave VARK to 426 anatomy students and told them the result. Two-thirds didn't study in line with it anyway, and the ones who did earned no better grades.

Why do teachers still teach learning styles?

Because the belief is nearly universal, intuitively appealing, and rarely challenged during training. Newton and Salvi's review of 15,405 educators found 89.1% endorsement, and pre-service teachers believed at least as often as qualified ones (95.4% against 87.8%, a gap that was not statistically significant). The encouraging finding is that in the four studies that tested it, direct instruction about the evidence cut self-reported belief from 78.4% to 37.1%.

What should I do instead of learning styles?

Use practice testing and distributed practice, the only two techniques Dunlosky's review rated high utility. Concretely: close the book and try to recall before rereading, and spread the same total study time across more days. Neither depends on a style label: the review rated them high because they "benefit learners of different ages and abilities."

Does personalized learning work?

Yes, when it personalizes on the right variable. Adapting to what a learner already knows, which questions they keep missing, and how long ago they last saw the material is well supported. Adapting to whether they call themselves visual or auditory is not.

What are the four VARK learning styles?

VARK stands for Visual, Aural, Read/write, and Kinesthetic, the four modalities Neil Fleming and Colleen Mills described in print in 1992. By VARK's own account, around two-thirds of people have a multimodal preference, which sits awkwardly with the idea that each learner has one best channel.

Are learning styles harmful?

They can be. In experiments published in 2023, parents, teachers, and children rated a hypothetical student described as a visual learner as more intelligent than one described as a hands-on learner. That's a measured effect on expectations, not on grades. The idea also takes up room in teacher training: as of 2020, the official study materials for elementary teacher licensing exams in 29 US states and the District of Columbia still referenced learning styles.

Does ChatGPT know learning styles are a myth?

Probably, if you ask it directly. In a 2025 study, three large language models (ChatGPT, Gemini, and DeepSeek) correctly judged around 80% of statements about the brain and learning, myths and facts alike. When the myth was built into a practical request about "visual learners," every model offered suggestions without challenging the premise, according to the university's press release. Asking the model to correct unsupported assumptions made it far more likely to speak up.

---

The Personalization That Actually Exists

Learning styles got one thing right: generic, undifferentiated instruction really does waste people. It just picked the wrong variable to fix that with.

What differs usefully between two learners is what they already know and which specific ideas haven't stuck. Both are measurable, and unlike a style, both change as you work.

That's also the case for keeping a durable record of your reading. A highlight library is a running map of what caught your attention and what you decided was worth keeping, which makes it usable as a question bank in a way a personality label never was. Glasp exists to make that record easy to build across the web, YouTube, and your Kindle, and to let you see what other people highlighted in the same source. That last part is the piece no questionnaire can give you: a genuinely different perspective on the same page.

Stop asking what kind of learner you are. Start asking what this particular material demands, and whether you can still recall it tomorrow without looking.

References: 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. 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. Dunlosky, J., Rawson, K. A., Marsh, E. J., Nathan, M. J., & Willingham, D. T. (2013). Improving students' learning with effective learning techniques. Psychological Science in the Public Interest, 14(1), 4-58. Elidan, G., & Haramaty, Y. (2025, September 16). Learn Your Way: reimagining textbooks with generative AI. Google Research Blog. Fleming, N. D., & Mills, C. (1992). Not another inventory, rather a catalyst for reflection. To Improve the Academy, 11, 137-155. Furey, W. (2020). The stubborn myth of "learning styles." Education Next, 20(3), 8-12. 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, 37:31. Husmann, P. R., & O'Loughlin, V. D. (2019). Another nail in the coffin for learning styles? Anatomical Sciences Education, 12(1), 6-19. 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. Martin Luther University Halle-Wittenberg. (2025, August 7). Myths about the brain: how ChatGPT and others might help to dispel popular misconceptions [Press release]. Massa, L. J., & Mayer, R. E. (2006). Testing the ATI hypothesis. Learning and Individual Differences, 16(4), 321-335. Nancekivell, S. E., Shah, P., & Gelman, S. A. (2020). Maybe they're born with it, or maybe it's experience. Journal of Educational Psychology, 112(2), 221-235. Nancekivell, S. E., Sun, X., Gelman, S. A., & Shah, P. (2021). A slippery myth: how learning style beliefs shape reasoning about multimodal instruction and related scientific evidence. Cognitive Science, 45(10), e13047. Newton, P. M., & Salvi, A. (2020). How common is belief in the learning styles neuromyth, and does it matter? Frontiers in Education, 5, 602451. Pashler, H., McDaniel, M., Rohrer, D., & Bjork, R. (2008). Learning styles: concepts and evidence. Psychological Science in the Public Interest, 9(3), 105-119. Richter, E., Spitzer, M. W. H., Morgan, A., Frede, L., Weidlich, J., & Moeller, K. (2025). Large language models outperform humans in identifying neuromyths but show sycophantic behavior in applied contexts. Trends in Neuroscience and Education, 39, 100255. 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. 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. Sternberg, R. J., Grigorenko, E. L., Ferrari, M., & Clinkenbeard, P. (1999). A triarchic analysis of an aptitude-treatment interaction. European Journal of Psychological Assessment, 15, 1-11. Sun, X., Norton, O., & Nancekivell, S. E. (2023). Beware the myth: learning styles affect parents', children's, and teachers' thinking about children's academic potential. npj Science of Learning, 8, 46.