Inspired loosely by an article by Dr. Shilpa Kapur Singh, this blog post revisits the widely accepted idea that students possess fixed learning styles such as visual, auditory, reading or writing, or kinesthetic.
Although learners clearly differ in their preferences, abilities, prior knowledge, motivation, and learning needs, research doesn’t provide convincing evidence that matching instruction to a presumed learning style improves learning outcomes (Pashler et al., 2008; Papadatou-Pastou et al., 2021; Clinton-Lisell & Litzinger, 2024).
Introduction
Learning styles have remained remarkably influential in education. The concept gained popularity because it offered a simple explanation for individual differences among learners.
Students were frequently classified according to preferred sensory modalities, with the VARK framework becoming particularly familiar in educational settings (Coffield et al., 2004; Papadatou-Pastou et al., 2021).
The VARK model, developed by Neil Fleming, describes four learning preferences: Visual, Aural, Read or Write, and Kinesthetic (Fleming & Mills, 1992). It suggests that learners may prefer diagrams, discussion, written material, or hands-on experiences. However, research hasn’t established that matching teaching to these preferences improves learning outcomes (Pashler et al., 2008; Papadatou-Pastou et al., 2021).
However, the simplicity of the concept doesn’t reflect the complexity of learning. Coffield et al. (2004) identified 71 different learning-style models, demonstrating considerable variation in how the term has been defined and measured.
Papadatou-Pastou et al. (2021) similarly found that education professionals frequently used the term learning styles to describe very different concepts, including VARK categories, multiple intelligences, learning theories, and teaching approaches.
The evidence behind the myth
Learning styles become problematic when preferences are treated as fixed characteristics, and instruction is deliberately matched to those preferences.
Pashler et al. (2008) conducted a critical review of the available evidence and found insufficient experimental support for the learning-styles matching hypothesis. Similar conclusions have been reported in subsequent reviews and empirical studies (Rohrer & Pashler, 2012; Cuevas, 2015; Willingham et al., 2015).
Rogowsky et al. (2015) examined whether participants’ preferred learning modalities predicted their performance when information was presented through different modes. No statistically significant relationship was found between learning-style preference, instructional mode, and performance.
These findings challenge the assumption that learners necessarily benefit when instruction is aligned with a self-identified learning style (Rogowsky et al., 2015).
More recently, Clinton-Lisell and Litzinger (2024) conducted a meta-analysis of 21 studies involving 1,712 participants. Although a small overall effect was identified, only 26% of learning-outcome measures demonstrated the crossover pattern required to support the matching hypothesis.
The authors therefore concluded that the evidence remained insufficient to justify widespread instructional matching based on learning styles.
Neuroscience and learning
The learning-styles concept has often been presented using neuroscientific language, but contemporary understanding of the brain doesn’t support the idea that learning occurs through isolated visual, auditory, or kinesthetic systems corresponding to fixed learner types.
Learning involves interconnected cognitive processes, including attention, working memory, prior knowledge, encoding, retrieval, motivation, and metacognition (Mayer, 2003; Howard-Jones, 2014). Multimodal learning can be beneficial when different forms of representation complement one another and reduce unnecessary cognitive demands (Mayer, 2003).
This distinction is important. Using visual material, discussion, demonstrations, writing, movement, or practical activities can strengthen learning, but their value depends on the learning objective and the characteristics of the content rather than on assigning students to permanent sensory categories (Mayer, 2003; Papadatou-Pastou et al., 2021).
From labelling learners to designing learning
One of the significant concerns associated with learning styles involves the labelling of students. Once learners are categorised as visual, auditory, or kinesthetic, those categories can become perceived limitations rather than flexible preferences.
Such labelling can reduce opportunities for learners to develop less-preferred strategies and may encourage fixed perceptions of ability (Papadatou-Pastou et al., 2021).
Research instead supports approaches that encourage metacognition, self-regulation, retrieval practice, meaningful encoding, and the flexible use of learning strategies (Dinsmore et al., 2008; Dunlosky et al., 2013).
Effective teaching therefore involves selecting methods according to the nature of the content, learning objective, learner characteristics, and context. A diagram may be appropriate for understanding a complex spatial relationship, while retrieval practice may be more appropriate for strengthening factual knowledge. A demonstration may be essential for developing a practical skill, while discussion may support interpretation and conceptual understanding.
Moving beyond learning styles
Dr. Pooja Agarwal’s work focuses on retrieval practice, an evidence-based learning strategy in which learners actively recall information rather than simply rereading it. Her research has demonstrated that brief retrieval activities, when incorporated into everyday teaching, can strengthen long-term retention and learning across different educational settings (Agarwal et al., 2021).
Her work has helped translate cognitive science research into practical classroom strategies (Agarwal & Bain, 2019).
Dr. Punya Mishra has written critically about learning styles while also highlighting the importance of cognitive diversity. He distinguishes between rejecting fixed learning-style categories and recognizing genuine differences in how people think, represent information, and approach problems (Mishra, 2009, 2023, 2025).
This distinction moves the discussion beyond labelling learners and towards recognising diverse ways of thinking and meaning-making in education.
AI, and the risk of automated myths
The learning-styles myth becomes particularly relevant in the age of AI. It can reproduce widely circulated educational claims without adequately distinguishing established evidence from popular misconceptions.
Dr. Mishra demonstrated this in 2023 when Bing Chat generated a seemingly credible explanation supporting learning-style matching, despite the evidence contradicting the claim (Mishra, 2023).
This highlights an important issue for AI-supported education — personalization shouldn’t mean placing learners into fixed categories.
AI can instead support flexible learning by adapting explanations, representations, feedback, pacing, and opportunities for retrieval according to the task and learner context. Such use requires informed teacher judgement and critical evaluation of AI-generated content (Mishra et al., 2023).
Conclusion
The persistence of learning styles demonstrates how easily an intuitive educational idea can become established as scientific knowledge. Learners are undoubtedly different, and individual preferences deserve recognition. However, preference shouldn’t be confused with evidence of a fixed neurological learning style.
The contemporary evidence supports a more flexible understanding of learning. Rather than categorizing students according to presumed learning styles, educators can design varied learning experiences, select representations according to the demands of the material, and help students develop metacognitive awareness and adaptable learning strategies (Papadatou-Pastou et al., 2021; Clinton-Lisell & Litzinger, 2024).
The movement away from learning styles isn’t a movement away from individualized education. Instead, it represents a shift from labelling learners to understanding learning.
That shift places greater emphasis on evidence, cognitive science, flexibility, and the remarkable capacity of learners to develop new ways of thinking and learning.
