The science of learning: a working map for teachers

A century of research on human learning has produced a small number of findings sturdy enough to build a school day on, a larger number of promising-but-contested ideas, and a stubborn collection of myths that outpoll the real findings in staffrooms. This pillar is our working map of all three, one evidence-reviewed article at a time. Everything Wiz Kids builds traces back to pages in this section — and every page grades its evidence, because a claim's popularity and its support are different things.

The findings strong enough to build on

Four results have the pedigree — replication across decades, ages, and subjects — to justify structural decisions:

Spaced repetition. Practice distributed over time massively outperforms the same practice massed together, on any test that matters. The oldest and most replicated finding in the field, and the cheapest to apply: move minutes, not budgets.

Retrieval practice. Pulling knowledge out of memory strengthens it more than putting it back in. Tests are learning events. The operational rule that follows — a review must never contain its own answer — quietly indicts most worksheets ever printed.

Mastery learning. Hold learning constant and let time vary, and the students conventional pacing abandons stop being abandoned. Bloom's two-sigma challenge, the meta-analytic record, and why cumulative subjects demand prerequisite graphs.

Feedback, immediate and specific (article coming). The tightest loops teach fastest — with real boundary conditions about praise, grades, and what "feedback" even means (Hattie & Timperley's model; the surprising evidence on grades crowding out comments).

Strong, with judgment required

Cognitive load theory (coming) — working memory is the bottleneck; worked examples, split attention, and why "discovery" fails novices (Sweller; Kirschner, Sweller & Clark, 2006). Interleaving (coming) — mixing problem types beats blocking, especially for discrimination between methods (Rohrer & Taylor). Desirable difficulties (coming) — the umbrella idea: conditions that slow visible performance often deepen learning (Bjork & Bjork). Motivation and self-determination (coming) — autonomy, competence, relatedness; what reward systems do to intrinsic interest, and the overjustification effect every points system should fear (Deci & Ryan).

Contested — handle honestly

Growth mindset (coming): the original findings, the large-scale replications with much smaller effects (Sisk et al., 2018; the national studies), and what survives: don't praise innate ability. Productive failure (coming): real but bounded. We grade these MIXED and say so — a reference source that can't say "contested" can't be trusted saying "strong."

The myths (staffroom edition)

Believed by majorities of teachers, supported by approximately nothing:

• Learning styles — the matching hypothesis fails every fair test. Match modality to content, not to children.
• Digital natives — exposure teaches swiping, not skills; the international assessments are brutal on this one. The founding myth this whole product exists to correct.
• Left brain/right brain, 10,000 hours, "goldfish attention spans," multitasking (articles coming) — each with its origin story, its refutation, and its kernel of misread truth.

How to read this pillar

Every article follows the same contract: an evidence grade up top (STRONG / PROMISING / MIXED / THIN, criteria here); a plain-language account of the key studies with citations; a "What the evidence doesn't say" section, because overclaiming is how good findings become next decade's myths; a practical "In the classroom" section; and — only where genuinely true — a note on how Wiz Kids applies it. When the literature is thin, we write that. When we change our reading, the article changes.

© Glu IO Pty. Ltd. — Wiz Kids (wiz.kids). Link freely; republication requires permission — see terms. Found an error in our reading of the research? We correct fast: tell any teacher piloting Wiz Kids.