Forty Years Later, and We Are Still Talking About Learning Styles

Photo by Google DeepMind on

In 1983, Howard Gardner introduced his theory of multiple intelligences in his landmark book Frames of Mind: The Theory of Multiple Intelligences. Gardner argued that intelligence is not a single general factor. Rather, human beings possess a plurality of “intelligences” or modalities (e.g., linguistic-verbal, logical-mathematical, spatial, musical, bodily-kinesthetic, interpersonal, intrapersonal, naturalistic). His theory advanced four core claims:

1. intelligence is not one thing,

2. these intelligences are shaped by biology and culture,

3. schools traditionally overvalue only two intelligences (linguistic and logical-mathematical), and

4. education should broaden its view of what counts as being “smart.”

Since its publication more than forty years ago, MI theory has been heavily criticized and widely rejected in mainstream psychology due to its lack of empirical evidence. Research has repeatedly found strong overlap among Gardner’s “intelligences,” implying they are better understood as abilities or personality tendencies—not separate neural structures. Gardner himself has repeatedly emphasized that MI theory is not the same as the popular “learning styles” idea, though the two have been conflated for decades.

But like many theories in education, MI theory produced unintended—and remarkably durable—offspring. It merged with other trends, became simplified into “visual/auditory/kinesthetic” learning styles, and eventually found new life inside other instructional frameworks. 

The Myth of Differentiation: Why Good Intentions Aren’t Enough

Beginning in the 1990s and early 2000s, Tomlinson popularized the concept of differentiated instruction, arguing that teachers should adjust content, process, product, and environment based on student readiness, interest, and learning profile. Her idea of “learning profiles” was influenced by MI theory and other learning-styles frameworks circulating in schools of education at the time. In her own writing, Tomlinson references multiple intelligences, learning styles, best modality, and preference-based grouping—positioning them as legitimate dimensions of differentiation even though they lacked empirical support.

Tomlinson’s work unintentionally fertilized and perpetuated the learning-styles myth. Differentiation promised personalization, inclusivity, and honoring each learner’s uniqueness. On paper, it sounded like the golden ticket to Willy Wonka’s factory of cognitive sweetness. In practice, however, differentiation often legitimized the idea that matching instruction to student type boosts learning—a claim that research overwhelmingly contradicts.

Even though Tomlinson never explicitly framed differentiation as an extension of MI theory, the two became deeply intertwined in teacher-preparation programs. Differentiation—as commonly taught and implemented—rests on several problematic assumptions:

1. that teachers can accurately diagnose learning profiles,

2. that matching instruction to those profiles improves learning,

3. that teachers can deliver multiple lesson versions simultaneously without compromising rigor, and

4. that tailoring tasks “up and down” leads to equitable outcomes.

The truth about these theories is that they are just that. There is no empirical evidence that teaching to a person’s learning style improves learning (National Council on Teacher Quality, 2021). Cognitive psychologists have repeatedly shown no reliable benefit from matching instruction to self-identified styles (Pashler et al., 2008; Willingham, 2018). Differentiation itself shows highly inconsistent effects. John Hattie (2009) found that while differentiation can work, its effect size varies widely and depends heavily on teacher expertise and classroom conditions–often making it overrated and difficult to implement well.

Perhaps most troubling is that when teachers differentiate by readiness, struggling students frequently receive simpler tasks, fewer opportunities for challenge, and less demanding texts. Research on leveled literacy programs, tiered assignments, and ability grouping repeatedly shows the same pattern: lowered expectations, lowered opportunities, and widened gaps. This outcome is the opposite of Tomlinson’s intention, yet it is the most common result in real classrooms.

The Rise of “Learner Variability”: The New Branding of an Old Idea

In recent years, “learner variability” has emerged as the newest well-intentioned but often misunderstood framework in education. The term appears everywhere—SEL, culturally responsive teaching, personalized learning, and digital platform marketing. On the surface, the concept seems self-evident: students differ. Their backgrounds, emotions, experiences, and needs matter. None of that is controversial—and none of it is new.

But the instructional interpretations of learner variability often echo the same problematic assumptions that fueled Multiple Intelligences and differentiation: that instruction must be tailored to student “profiles”, and that matching teaching to these profiles will unlock deeper learning.

Yet like learning styles and differentiation, there is no empirical evidence that designing instruction around preferred “learning pathways,” “modality strengths,” “profiles,” or “unique learner maps” improves academic outcomes. Just like learning styles, these constructs suggest that effective teaching means customizing how students receive content rather than strengthening what and how well they learn.

Instead, the research is clear: students benefit most when instruction is structured, explicit, knowledge-rich, sequenced, and grounded in cognitive principles—not when teachers attempt to engineer bespoke paths for every learner.

And ironically, when “learner variability” gets framed as a call for personalized pathways, it too often becomes a justification for lowering rigor, narrowing expectations, or fragmenting whole-class instruction in ways that widen rather than close gaps.

How We Learn is More Alike Than Different

In reality, how humans learn is far more alike than different. Research overwhelmingly supports explicit, well-sequenced whole-class instruction, scaffolding, feedback, and formative assessment—not micro-differentiation—as the most reliable ways to improve learning. Cognitive architecture is universal (Sweller, Kirschner, & Clark, 2007; Sweller, Ayres, & Kalyuga, 2011). All humans share the same working-memory constraints, encode and retrieve information using the same long-term memory systems, and learn by integrating new information into existing knowledge structures (“schemas”). These constraints of cognition are shared, not individualized.

Decades of research in reading development similarly show that children learn to read in fundamentally the same way: by mapping print to speech through phonological pathways. Variability lies in rate, prior knowledge, and experience, not in the underlying cognitive mechanisms (Seidenberg, 2017; Castles, Rastle, & Nation, 2018). Reading is learned through the same processes across cultures, languages, and demographics.

Memory research also demonstrates striking commonalities: retrieval practice, spaced practice, interleaving, encoding variability, and dual coding consistently benefit all learners—not just certain “types” (Bjork, 1994; Roediger & Karpicke, 2006). Daniel Willingham’s work reinforces this point: knowledge matters for everyone, and expertise develops universally through modeling, practice, feedback, and worked examples, regardless of age or background.

Across domains, the strongest evidence shows that effective learning is grounded in shared cognitive mechanisms, not in idiosyncratic “learning styles” or individualized instructional pathways.

Why These Myths Matter–and What We Must Do Instead

For more than four decades, education has been shaped by well-intentioned ideas that feel compassionate, inclusive, and student-centered. Multiple Intelligences, learning styles, differentiation, and now “learner variability” all arise from the same impulse: to honor student differences. But good intentions are not enough. When these theories are translated into classroom practice, they too often pull teachers away from what reliably strengthens learning—and toward approaches with little to no empirical support.

The through-line across these fads is clear: each suggests that learning improves when instruction is customized to student “profiles.” Yet across cognitive psychology, neuroscience, and discipline-specific research, the evidence tells a different story. Students don’t thrive because instruction is fragmented into individualized pathways; they thrive because instruction is coherent, explicit, knowledge-rich, and designed around universal principles of how the human mind learns.

When schools chase individualized learning profiles in hopes of boosting engagement, we unintentionally shift our attention away from practices that truly matter—teacher clarity, modeling, scaffolding, deliberate practice, repetition with variation, and strong feedback loops. And when we sort students into “preferred modalities” or “readiness levels,” we risk lowering expectations, weakening rigor, and widening the very gaps we intend to close.

What students need is not a personalized playlist. They need powerful, shared learning experiences that build the background knowledge, vocabulary, and cognitive frameworks necessary to understand the world. What teachers need is not another profile-based system to diagnose. They need the time and professional learning to implement evidence-based instruction with precision and confidence. And what schools need is not new terminology for old ideas. They need a clear-eyed commitment to practices grounded in how memory, attention, and learning actually work.

When we strip away the myths, what remains is both simple and profound: humans learn more alike than different. Our shared cognitive architecture is not a limitation—it’s the foundation that allows all students to benefit from clear instruction, well-sequenced curriculum, and meaningful practice. When we honor what is universal about learning, we create classrooms that are not only more equitable, but more effective. This is the path forward—not designing instruction around profiles and preferences, but designing instruction around the science of learning itself.

References

Bjork, R. A. (1994). Memory and metamemory considerations in the training of human beings. In J. Metcalfe & A. Shimamura (Eds.),Metacognition: Knowing about knowing (pp. 185–205). MIT Press.

Castles, A., Rastle, K., & Nation, K. (2018). Ending the reading wars: Reading acquisition from novice to expert. Psychological Science in the Public Interest, 19(1), 5–51. https://doi.org/10.1177/1529100618772271

Gardner, H. (1983). Frames of mind: The theory of multiple intelligences. Basic Books.

Hattie, J. (2009). Visible learning: A synthesis of over 800 meta-analyses relating to achievement. Routledge.

National Council on Teacher Quality. (2021). The stubborn myth of learning styles. NCTQ Research & Policy.https://www.nctq.org/research-insights/what-were-reading-the-stubborn-myth-of-learning-styles

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

Roediger, H. L., & Karpicke, J. D. (2006). Test-enhanced learning: Taking memory tests improves long-term retention. Psychological Science, 17(3), 249–255. https://doi.org/10.1111/j.1467-9280.2006.01693.x

Seidenberg, M. S. (2017). Language at the speed of sight: How we read, why so many can’t, and what can be done about it. Basic Books.

Sweller, J., Kirschner, P. A., & Clark, R. E. (2007). Why minimal guidance during instruction does not work: An analysis of the failure of constructivist, discovery, problem-based, experiential, and inquiry-based teaching. Educational Psychologist, 42(2), 75–86. https://doi.org/10.1207/s15326985ep4102_1

Sweller, J., Ayres, P., & Kalyuga, S. (2011). Cognitive load theory. Springer. https://link.springer.com/book/10.1007/978-1-4419-8126-4

Tomlinson, C. A. (1999). The differentiated classroom: Responding to the needs of all learners. ASCD.

Tomlinson, C. A. (2001). How to differentiate instruction in mixed-ability classrooms (2nd ed.). ASCD.

Willingham, D. T. (2018). The reading mind: A cognitive approach to understanding how the mind reads. Jossey-Bass.