The Evidence on Teaching

Explore the evidence

7 of 88 decisions · clear everything

Survives the gold standardIndependent, durable, rigorousThe debunked shelfCheap winsFor parents at homeFamous numbers, deflatedWhere we admit ignoranceThe terrain
evidencealldrop DA–BA only
evaluatoranyno dev-ledindependent
durabilityanyfollow-up≥1yr≥2yr
what are these? →
sortconfidencerobustfragileA–Z
domain
verdict
confidence
layer
survival prior
Your situation · from the applicability layer — absence means “not yet assessed,” never “doesn’t apply”
who's delivering
group / setting
cost
outcome class

Teaching coding teaches coding: the one clean school RCT gives g=0.47-0.68. Which approach you pick barely matters, and almost every effect size rests on an instrument the developers built.

moderate supportconf: mediumgc: low
22/2225/48connected:Does learning to code improve geneTalent and trainability

Does teaching programming actually teach programming — and does the method matter? · computer-science · ages 5–18

Mixing problem types helps where confusion is the enemy — discriminating similar categories, mixed math practice — and is useless or worse for facts and prose.

moderate supportconf: mediumgc: low
7/718/48connected:Spaced (distributed) practiceScience content sequencing

Interleaving (mixing problem types vs blocking) · practice · ages 10–18 · method

Manipulatives help when bland, guided, and aged ~7–11 — a guided-representation effect, not 'hands-on learning.' Rich, toy-like objects hurt transfer.

moderate supportconf: mediumgc: low
3/36/48

Manipulatives and concrete-representational-abstract (CRA) sequences · math · ages 5–14

Word-problem solving is its own skill: computation fluency doesn't produce it. Teaching problem schemas explicitly does (~0.25–0.45 SD on trained content).

moderate supportconf: mediumgc: low
7/78/48connected:Math-fact fluency and timed practiWorked examples (studying solution

Word problems, schema instruction, and the conceptual-vs-procedural question · math · ages 6–14

Sort by what the software replaces: adaptive drill inside the school day buys +0.05-0.20 SD; a device, a connection or the teacher buys zero to negative; LLM tutors have no usable evidence at all.

mixedconf: mediumgc: low
53/5339/48connected:Placement and mastery diagnosisMastery learning (teach → test → rTutoringClass-size reduction

Does educational technology raise learning — CAI, adaptive software, devices, screens, and AI tutors? · edtech · ages 5–18 · structure

Near-zero in primary and almost entirely correlational in secondary. The causal base is three small trials; what the homework IS beats how much of it there is.

mixedconf: mediumgc: medium
16/1629/48

Homework — effects by age and by dosage · homework · ages 5–18 · structure

Classroom order is causally worth a great deal — one disruptive peer costs classmates 3% of adult earnings — yet no branded behaviour programme reliably delivers it and exclusion makes things worse.

mixedconf: mediumgc: low
25/2536/48connected:Talent and trainabilityCan executive function be trained,Should a school buy a social and eAre self-control and grit teachabl

How should a school run its classrooms and its discipline system? · behavior · ages 4–18 · structure

The URL is the state — any view you build here is a permalink you can hand to someone mid-argument.