The Learning Styles Illusion: Debunking Education's Most Persistent Myth (Again)

New paper explores how a thoroughly discredited theory still shapes teacher expectations and student potential

Despite decades of evidence showing that matching instruction to students' "learning styles" has near-zero impact on student achievement, this myth remains remarkably persistent in educational settings. I’ve written about learning styles at length here but recently I've noticed a resurgence of the learning styles myth so this excellent new paper from John Hattie and Tim O’Leary is timely.

The central myth of learning styles is simple: it’s not that students have different learning preferences but that students learn better when taught in their preferred modality, usually visual, auditory, kinesthetic, or otherwise. This “matching hypothesis” has been tested repeatedly and across four meta-analyses* focused specifically on matching styles to teaching methods, the average effect size is d = 0.04 - essentially zero. As the authors make clear, this is not just an underwhelming result; it is definitive: matching instruction to learning styles does not work. Yet the myth persists. Why?

The "Appealing Simplicity" of Learning Styles

The intuitive appeal of learning styles cannot be overstated. The theory aligns with our common-sense observation that people do indeed have preferences for how information is presented. We readily observe that some students appear to engage more with visual materials, while others seem to prefer discussion or hands-on activities. The leap from these observable preferences to the conclusion that teaching should match these preferences feels natural, even though the evidence contradicts this assumption. As the paper notes, the myth offers "appealing simplicity" compared to more complex, evidence-based approaches.

Hattie and O’Leary point to a critical confound: the conflation of learning styles with learning strategies. Correlational studies often show modest relationships between learning preferences and achievement (average r = 0.24), but these studies do not test the core claim of matching instruction to learning styles. Many also fail to account for prior knowledge or strategy use which are much stronger determinants of learning outcomes.

The Dangers of Labelling

This confusion is more than just academic. The most troubling implication of the learning styles myth is that it invites low expectations. Labelling a student a “kinesthetic learner” or a “visual learner” easily becomes shorthand for what they can’t do. The paper cites evidence that teachers disproportionately associate kinesthetic styles with low achievement and visual styles with high ability. In other words, learning styles become a euphemism for ability grouping, cloaked in pseudoscientific legitimacy.

The paper references recent research by Sun, Norton, and Nancekivell (2023) which provides compelling evidence that the harm of learning style labels extends beyond mere ineffectiveness. Their study reveals that parents, teachers, and even children themselves judge students described as "visual learners" as more intelligent than those described as "hands-on learners."

In a series of experiments, they found that these labels trigger significant assumptions about academic potential. Visual learners were expected to excel in traditionally "academic" subjects like mathematics, language arts, and social studies, while hands-on learners were expected to perform better only in subjects often considered less academically rigorous, such as physical education and music.

What makes these findings particularly concerning is how early these biases emerge. Even children as young as 6-12 years old showed clear preferences, judging visual learners as smarter than hands-on learners. The researchers noted that these aren't merely neutral descriptors but value-laden categories that shape expectations and opportunities.

The use of learning styles has led to such adverse effects of labeling. Sun et al. (2023) conducted three studies that showed that “parents, teachers, and children judged children described as visual learners as more intelligent than children described as hands-on learners” (p. 46). Additionally, visual learners were expected to excel in core subjects like mathematics, language arts, and social sciences, whereas hands-on learners were predicted to perform better in more practical subjects such as physical education and music.(p.30)

Perhaps most telling was their finding that when teachers and parents were asked to predict report card grades, they consistently rated visual learners higher in core academic subjects, with the single exception being science, where hands-on learners were judged to be equally capable. This pattern of judgment creates an implicit hierarchy that can impact everything from teacher expectations to student self-perception.

The research suggests that well-intentioned attempts to recognise learning differences through these labels may actually reinforce stereotypes about intellectual ability. When we label a child as a "hands-on learner," we're not merely describing a preference, we might be inadvertently signalling limitations about their academic potential.

Worse, once students are labeled, the label sticks. It reduces flexibility, encourages passive learning, and distracts from what we know actually improves learning: teaching effective, flexible strategies aligned with task demands. Effective learners are adaptive—they don’t stick to one modality; they switch, adjust, and experiment.

The Methodological Disaster of Learning Styles Research

A key point made in the paper is that learning styles research stands uniquely as the most methodologically flawed area in educational psychology, characterised by poor design, calculation errors, theoretical gaps, low validity, and commercial bias rather than scientific rigor. Hattie and O'Leary's comprehensive review of 2,500 meta-analyses reveals that no other educational influence is marked by such consistently poor quality. The research typically lacks even basic methodological standards: studies rarely meet minimum acceptability criteria, calculations contain fundamental errors, theoretical frameworks remain undeveloped, and assessments demonstrate remarkably low validity.

What distinguishes this field from other problematic research areas is the sheer pervasiveness of these flaws combined with aggressive commercial promotion. The learning styles literature fails to explain how these purported styles develop, neglects connections to meta-cognitive processes, ignores personality factors, and presents styles as unchanging, mutually exclusive categories hardwired into the brain. The authors write:

An important observation from this review is that many, if not most, of the studies in these meta-analyses exhibit many methodological weaknesses in research design: especially evaluating the aptitude-treatment interaction expected from the model. They are riddled with errors, and too often they are exemplars of poor-quality meta-analyses. The learning styles literature does not describe how styles develop, barely relate to meta-cognitive, motivation and personality mechanisms, and they seem unchanging, mutually exclusive, implying different kinds of students, and that the styles are hard-wired into the brain (An & Carr, 2017; Nancekivell et al., 2020).

"One of Education's Most Seductive Falsehoods."

The persistence of learning styles theory despite its empirical failures represents what the authors call "one of education's most seductive falsehoods." Its appeal lies precisely in its ability to provide simple, seemingly scientific explanations for complex learning challenges while avoiding more difficult conversations about evidence-based pedagogy, student engagement, and the development of effective learning strategies.

Moving forward requires more than simply abandoning learning styles; it demands embracing what cognitive science actually tells us about effective learning. Rather than labeling students with fixed, modality-based categories, we should be teaching them adaptive strategies that respond to the specific demands of different learning tasks. This approach acknowledges the complexity of human learning while providing students with versatile tools they can deploy across contexts.

---

* From the paper": ”There are many articles, reviews, and empirical studies showing the poverty of the matching hypothesis (e.g., Cuevas, 2015; Fowler & Snyder, 2021; Husmann & O’Loughlin, 2019; Khazan, 2018; Knoll et al., 2017; Kr.tzig & Arbuthnott, 2006; Rogowsky et al., 2015; Riener & Willingham, 2010; Willingham et al., 2015). Among the most cited reviews are Pashler et al. (2008) paper ‘Learning Styles: Concepts and Evidence,’”