Learning Myths: What the Evidence Actually Supports

A myth is an overreach, not merely an idea we dislike

A neuromyth often begins with a genuine finding and then stretches it into a claim the research did not test. Functional differences between brain regions become personality types; a preferred presentation becomes a prescription for instruction. The useful question is not whether a slogan sounds scientific, but what exact prediction it makes.^[1]

These beliefs have persisted in educational settings, but prevalence surveys do not by themselves prove that a believer teaches poorly or harms achievement. The direct classroom consequences of neuromyth belief remain under-studied.^[2]

Learning styles: preferences are not a matching effect

The testable “meshing” hypothesis says learners first classified into styles should learn better when instruction matches that classification. A proper test assigns learners to different methods, gives everyone a common outcome measure, and looks for a crossover: one method should help one style while a different method helps another.^[3]

A major evidence review found no adequate basis for putting learning-style assessments into general educational practice. In a later experiment, auditory and visual-word preferences did not interact with audiobook or text instruction to improve immediate or two-week comprehension.^[3][4]

This does not mean preferences, aptitudes, disabilities, language needs, or prior knowledge are imaginary. It means a style questionnaire has not been shown to prescribe the presentation that will produce better learning. Nor does this evidence prove that every learner responds identically to every method.^[3][4]

Choose representations for the content and goal

A diagram can display spatial relationships, audio is necessary when the sound itself is being studied, and a demonstration can show a physical procedure. Those are matches between representation and task, not between a learner’s identity and a fixed channel.

More formats are not automatically better. Kaboosh’s practical rule is to give each representation a clear purpose and explain how the formats relate; if one adds neither useful information nor access, consider removing it.

Accessibility is a separate requirement. Captions, descriptions, tactile materials, assistive technology, and agreed accommodations should respond to access needs, not to a learning-style label.

The 10% brain claim uses the wrong kind of meter

There is no evidence that 90% of healthy brain tissue normally lies functionless, waiting to be unlocked. But replacing the claim with “we use 100% of the brain” can be misleading too: different systems contribute at different times, and not every region is maximally active during every task.^[1][5]

Brain activity, cognitive performance, learning potential, and untapped skill are different concepts. None is meaningfully summarized by one percentage. Research can show that learning changes performance; it cannot convert human potential into a brain-use gauge.^[5]

Lateralization is real; hemisphere personalities are not

Some functions and connections are more lateralized than others, and lateralization can vary across people. That real specialization does not establish that analytical people use one hemisphere globally while creative people use the other.^[5][6]

A resting-state connectivity study of 1,011 people aged 7 to 29 found local patterns of lateralization but not a whole-brain left-dominant or right-dominant network phenotype. Resting connectivity is not a direct test of every personality trait, but it provides no support for assigning students a hemisphere personality.^[6]

A better description is specialization plus cooperation. Complex thought depends on distributed, interacting systems, so a learner should not be routed toward art or algebra on the basis of a left/right label.^[1][5]

Multitasking: possible is not the same as cost-free

The slogan “humans cannot multitask” is also too absolute. Some routine or highly practiced activities can coexist. The more defensible finding is that switching between demanding task sets commonly brings interference and performance costs, and those costs vary with the tasks and preparation.^[7]

In two small simulated-lecture experiments, undergraduates assigned unrelated laptop tasks performed worse on an immediate test, and nearby peers were also affected. That does not prove that every device use, music choice, or classroom arrangement harms learning.^[8]

Multitasking is not the same as interleaving. Interleaved practice deliberately alternates related items in sequence so the learner must choose an approach; it does not ask the learner to answer two demanding prompts at the same instant.

Read the focused evidence in the interleaved practice guide.

A practical claim-check

Kaboosh’s view: a good debunk is specific. It should identify the promised outcome, the study design needed to test it, and the evidence that is still missing. These questions are a starting checklist, not a formal quality-rating tool.

• What exact learning outcome is promised—recall, transfer, grades, preference, or enjoyment?
• Were learners assigned to different methods and assessed with a common test?
• Does the study test the classroom claim, or only an adjacent brain phenomenon?
• Are the participants, task, and delay relevant to the conclusion being made?
• Has the result been reviewed or replicated, and are important limitations visible?

For teachers: design for the task, not a type

A teacher introducing cell structure could use a labelled diagram for spatial relationships, a concise explanation for function, and one common retrieval task asking students to reconstruct and explain the cell. Access supports can be added without assigning every learner a type.

For demanding independent work, the teacher might set a clear device-use window and a notification-free explanation or retrieval window. That is practical classroom design informed by switching evidence, not a universal device ban.

For a classroom-ready study activity, browse Kaboosh’s public flashcard decks.

For independent learners: test methods, not identities

A learner who prefers diagrams can compare methods using the same delayed outcome: study matched material with a useful diagram and explanation, study another set with explanation alone, then retrieve both later without notes. The question is what helped this content and goal, not which label felt right.

For attention-demanding work, complete one defined retrieval set before checking messages. That is a reasonable response to task-switching evidence, not a guarantee of higher grades or proof that every paired activity is impossible.

Use the active recall guide to measure delayed retrieval, and the

spaced repetition guide to choose when to check again.

The realistic conclusion

The evidence does not support routine instruction matched to diagnosed learning styles, a dormant 90% of the brain, global hemisphere personality types, or assuming that switching between demanding tasks is cost-free. It also does not prove that all learners are the same or that one method is universally best.^[3][5][6][7]

Preferences, prior knowledge, accessibility needs, and the content itself still matter. Clearing away an overclaim should make room for better questions, not another absolute slogan.