한국어

Are Learning Styles Real? What the Experiments Found

At a glance
  • Preferences are real. The claim that you learn more when taught in your preferred style has no good evidence.
  • 93 to 96% of teachers in one survey believed it, yet experiments built to test it found no benefit.
  • In one study of 107 fifth graders, 68% had about the same visual and auditory preference scores.
  • Your style predicts how well you expect to remember something, not how well you actually remember it.
  • Instead, pick the format the material needs, then close the book and recall what you learned.

Learning styles are real as preferences, but not as a way to learn more. Most people can say whether they like pictures, words or listening. What has not held up is the claim that you learn more when teaching matches that preference. Experiments built to test it found no such benefit. What helps is picking the format the material needs, then closing the book and recalling what you learned.

What does a learning styles test measure?

Mostly what you like, not how well you learn. A typical quiz asks how you prefer to take in information.

Pashler and colleagues (2008) reviewed the field. They wrote that learning style tests usually ask people what kind of presentation they prefer (words, pictures, speech) or what kind of mental activity they find most congenial. The most common claim built on them is the meshing hypothesis: teach a "visual learner" with visuals and an "auditory learner" by listening, and each learns more. There are many such systems. Newton (2015), citing an earlier review, counts more than 70. VARK (visual, aural, read/write, kinesthetic) and Kolb's experiential types are among the best known.

Do you learn better in your learning style?

There is almost no evidence that you do. Of the few well-designed experiments that tested it, several found results that run against it.

Pashler and colleagues (2008) first set out what a fair test would need.

StepWhat the experiment must do
1Sort people by learning style
2Randomly assign the people in each style group to one of several methods
3Give everyone the same final test
4Show that the best method for style A is not the best method for style B

Source: Pashler et al. (2008), abstract. They found virtually no studies showing that crossover. Several studies with a fitting design found results that flatly contradict the meshing hypothesis. Their conclusion was that there was no adequate evidence for using learning style assessments in general education. They also noted that many style systems have never been tested at all, so not every possible version has been ruled out.

Rogowsky and colleagues (2020) followed this design with fifth graders, giving every child both methods. On one day they read passages, on another they listened to recorded ones, and after each passage they answered the same kind of written comprehension questions. On the third day they took a standardized learning style inventory. Comparing clearly visual learners (22) with clearly auditory learners (12), the visual group scored higher both when reading and when listening, and both groups scored slightly, but not significantly, higher on listening than on reading. The crossover the hypothesis predicts, with auditory learners ahead on listening and visual learners ahead on reading, did not appear.

Visual learners scored higher both ways, and the predicted crossover did not appear

20304050AuditorylisteningAuditory listening · 24.1~34.824.1~34.8AuditoryreadingAuditory reading · 22.6~32.922.6~32.9VisuallisteningVisual listening · 35.7~43.635.7~43.6VisualreadingVisual reading · 34.7~42.334.7~42.3
95% confidence intervals for mean comprehension scores of fifth graders. 12 clearly auditory and 22 clearly visual learners each read and listened to passages, then answered the same kind of questions.Source: Rogowsky et al. (2020), Results and Figure 2 legend
See the numbers
FromTo
Auditory listening24.134.8
Auditory reading22.632.9
Visual listening35.743.6
Visual reading34.742.3

The visual group may simply have been stronger readers and listeners to begin with. Either way, their edge did not come from visual material, since it held when they listened too. The same team's 2015 experiment with adults also failed to support the meshing hypothesis, the paper reports.

If it is a myth, why does almost everyone believe it?

Because two claims get blended into one: that people have preferences, which is true, and that teaching to them works better, which has no good evidence.

Dekker and colleagues (2012) surveyed 242 teachers in the UK and the Netherlands who were interested in brain science and learning. They showed them 32 statements about the brain and learning. Most agreed with the true one, "Individual learners show preferences for the mode in which they receive information." But 93 to 96% also agreed with the myth, "Individuals learn better when they receive information in their preferred learning style." It topped the paper's list of the most widely believed of the 15 myths.

Most teachers agreed that people learn better in their preferred style

  • UK
  • Netherlands
0%20%40%60%80%100%UK · People have preferences (true) · 95%95%Netherlands · People have preferences (true) · 82%82%People havepreferences (true)UK · Matching helps learning (myth) · 93%93%Netherlands · Matching helps learning (myth) · 96%96%Matching helpslearning (myth)Over 9 in 10 agreed
Share of teachers who said each statement was correct (%). UK n = 137, Netherlands n = 105. The first statement is true. The second is a myth.Source: Dekker et al. (2012), Table 1
See the numbers
UKNetherlands
People have preferences (true)95%82%
Matching helps learning (myth)93%96%

The research literature looks similar. Newton (2015) analyzed 109 papers on learning styles in higher education from 2013 to 2015. 94% started from a positive view of learning styles and 89% reached a positive conclusion. Kolb's model featured in 34% and VAK-type models in 33%. Only one paper tested matching, and it found no effect.

Why does my learning style feel so right?

Because it goes with how well you expect to remember, not with how well you do.

Knoll and colleagues (2017) measured people's preference for words or pictures, then had them study a list of word pairs and a list of picture pairs. While studying, they rated how likely they were to remember each pair. People who preferred words expected to remember the words better, and people who preferred pictures expected to remember the pictures better. But preference was unrelated to what they actually recalled, and unrelated to how accurate their predictions were.

Learning style was linked to how well people expected to remember, not to how much stuck.

That may be part of why the idea survives. Studying in your favorite format feels easy, and easy feels like learning. It is the same trap as the feeling of having studied, covered in the note on how to improve metacognition.

Do students actually study in their style?

Most do not, and the ones who did were not better off.

Husmann and O'Loughlin (2019) gave 426 undergraduate anatomy students the VARK questionnaire and a survey of their study habits. Most did not report study strategies that matched their VARK results. Scores in no VARK category correlated with course performance. What did correlate with final grades were particular strategies, regardless of style. Using the virtual microscope was one. Whether students' strategies lined up with their VARK results had no correlation with how they did.

One more number from Rogowsky and colleagues (2020). Of 107 fifth graders, only 21% had a clearly visual style and 11% a clearly auditory one. The other 68% had identical visual and auditory scores or scores one or two points apart.

Seven in ten children had no clear visual or auditory style

0%20%40%60%80%100%Students · Clearly visual · 21%21%ClearlyvisualStudents · About equal · 68%68%AboutequalStudents · Clearly auditory · 11%11%Clearlyauditory73 of 107
Share of 107 fifth graders with both learning style and comprehension scores (%), grouped by their visual and auditory preference scores.Source: Rogowsky et al. (2020), Results
See the numbers
Students
Clearly visual21%
About equal68%
Clearly auditory11%

A quiz that hands you one label makes it feel like a fixed trait. In this sample, mixed preferences were far more common.

What to do instead, starting today

Choose by two things other than your style. You need only a sheet of paper.

  1. Match the format to the material. Shapes like maps and anatomy are usually easier to show in pictures, pronunciation in sound, and arguments in text. The material decides, not the learner.
  2. Spend three minutes recalling. After a video, a chapter or a lecture, close it and write three lines of what you remember. Then open it again and mark only what you missed.
  3. Test your own method by results, not feelings. Study half a chapter from a video and half from text, then recall both the next day. The format that felt easier may not be the one that stuck.

Why recall makes things last is covered in the note on spaced repetition schedules. Another classroom myth is taken apart in is the learning pyramid accurate?

In Aloom

The point of this note is that whether you recall what you learned matters more than which sense it came through. In Aloom you ask AI a question and keep only the answers you want. Each kept answer links to what you already know. Each day you recall 1, 3 or 5 of them as questions before you see the answer, the same way whatever your style. The more you recall, the darker the ink. See how Aloom works

Frequently asked questions

Is the VARK learning styles model valid?

VARK describes how you say you like to take in information. As a guide to how you should study or be taught, it has not held up. In one study of 426 anatomy students, no VARK score correlated with grades, and students whose study habits matched their VARK results did no better.

Are learning styles completely debunked?

The meshing hypothesis has no good support, and the experiments built to test it found none. Reviewers note that many style systems have never been properly tested, so "no evidence" is more accurate than "proven false" for every version. Either way, there is no reason to choose how you study based on a style quiz.

I'm a visual learner. Should I only use videos and diagrams?

There's no need. In the fifth-grade study, children with a visual style did just as well when they listened as when they read. Use pictures when the material is visual, and whatever the format, close it afterwards and recall what you learned.

References