What the Evidence Says
Famous claims weighed against the peer-reviewed literature — a verdict, a plain summary, and real, verifiable sources.
Showing 56 of 56
Working from home increases productivity.
It helped in a controlled call-centre trial, but the effect depends heavily on the job and the design.
Open →RefutedMoney stops buying happiness above about $75,000 a year.
The famous plateau was a measurement artefact; for most people, well-being keeps rising with income.
Open →ContestedAbout 10,000 hours of practice makes anyone an expert.
Practice matters a great deal, but it explains only part of the gap between experts — and the '10,000-hour rule' is a popularisation, not a finding.
Open →MixedTeaching a 'growth mindset' reliably raises student achievement.
Effects are real but small on average, and concentrated among lower-achieving and disadvantaged students.
Open →RefutedHolding a 'power pose' for two minutes changes your hormones and makes you bolder.
A large replication found no hormonal or risk-taking effect; only a subjective feeling of power survived.
Open →ContestedRaising the minimum wage always destroys jobs.
Moderate increases have shown little to no employment loss in the best studies; the textbook prediction is not automatic.
Open →RefutedListening to Mozart makes you smarter.
The original study found a brief spatial-task bump, not a lasting IQ gain; meta-analysis shows the effect is tiny and probably just arousal.
Open →RefutedTeaching to a student's 'learning style' (visual, auditory, kinaesthetic) improves learning.
People have preferences, but matching instruction to them does not improve outcomes in controlled tests.
Open →RefutedPeople are 'left-brained' (logical) or 'right-brained' (creative).
Brain imaging shows no dominant hemisphere that defines personality; both sides work together.
Open →ContestedWillpower is a limited resource that gets 'used up' during the day (ego depletion).
A famous early effect largely failed a big pre-registered replication; the resource model is now seriously doubted.
Open →ContestedA child who resists one marshmallow now is destined for success later.
The link to later outcomes shrinks sharply once family background is accounted for.
Open →ContestedThe least competent people are the most overconfident (Dunning–Kruger effect).
Almost everyone misjudges themselves toward the average; much of the classic pattern is a statistical artefact.
Open →ContestedForcing a smile makes you feel happier (facial feedback).
A famous pen-in-mouth study failed a large registered replication; later work suggests a small effect at best.
Open →MixedThe more bystanders present, the less likely anyone helps in an emergency.
The lab effect is real, but real-world CCTV shows someone intervenes in the large majority of public conflicts.
Open →ContestedThe Stanford Prison Experiment proved good people turn cruel when given power.
Archival evidence shows guards were coached and results shaped; it functions more as a demonstration than proof.
Open →MixedOrdinary people will deliver lethal electric shocks if an authority tells them to.
High obedience is real and has been partly replicated, but the shocks were fake and interpretations are debated.
Open →Contested'Grit' predicts success better than talent or IQ.
Grit is barely distinguishable from ordinary conscientiousness and predicts performance only weakly.
Open →ContestedReminding people of a negative stereotype reliably lowers their test scores.
The original effect is real in some settings but small, inconsistent, and shadowed by publication bias in replications.
Open →ContestedThe Implicit Association Test measures hidden prejudice that predicts discrimination.
The test is unstable over time and only weakly predicts real behaviour.
Open →RefutedYou flourish when positive-to-negative emotions cross a precise 2.9-to-1 ratio.
The exact 'tipping point' number came from misapplied fluid-dynamics maths and was retracted.
Open →MixedTeacher expectations alone can dramatically raise a student's IQ.
Expectations do influence outcomes, but the effect is modest and the dramatic IQ jumps did not hold up.
Open →MixedHeavy media multitaskers are better at juggling information.
An influential study found the opposite, but replications and meta-analysis show the effect is small and inconsistent.
Open →SupportedSpreading study out over time beats cramming.
One of the most robust findings in learning science: spaced practice produces far better long-term retention.
Open →SupportedTesting yourself is a better way to learn than re-reading.
Retrieval practice consistently improves long-term retention more than restudying the same material.
Open →SupportedWriting about your feelings improves mental and physical health.
Brief expressive writing produces small but reliable benefits across many outcomes.
Open →SupportedSleep helps consolidate what you learned during the day.
Strong, converging evidence shows sleep actively strengthens and reorganises new memories.
Open →MixedThe 'hot hand' in basketball is a myth — streaks are just chance.
The famous debunking contained a subtle statistical bias; corrected, a small real hot hand reappears.
Open →RefutedThe MMR vaccine causes autism.
The original paper was fraudulent and retracted; large studies of millions of children show no link.
Open →RefutedAntioxidant supplements help you live longer.
Pooled trials show no mortality benefit — and beta-carotene and vitamin E may slightly increase death.
Open →MixedVitamin C prevents the common cold.
It does not prevent colds for most people, but regular intake modestly shortens them.
Open →RefutedA daily aspirin protects healthy older adults from heart disease.
In healthy elderly people it did not extend life and raised the risk of serious bleeding.
Open →MixedFish-oil (omega-3) supplements prevent heart disease.
General supplements show little benefit, though high-dose prescription formulas help specific patients.
Open →RefutedSugar makes children hyperactive.
Controlled trials find no effect on behaviour; the belief is largely in the parent's expectation.
Open →RefutedYou must drink eight glasses of water a day.
There is no scientific basis for the '8×8' rule; food and other drinks count, and thirst is a decent guide.
Open →ContestedBreakfast is the most important meal — skipping it makes you gain weight.
Randomized trials find eating breakfast does not aid weight loss and may slightly increase daily calories.
Open →ContestedFlossing daily clearly prevents gum disease and cavities.
Trials show flossing plus brushing reduces gum inflammation, but the evidence is weak and low-quality.
Open →MixedMindfulness meditation is a powerful treatment for a wide range of problems.
Meditation modestly helps anxiety, depression and pain, but is not a cure-all and evidence quality varies.
Open →SupportedExercise helps treat depression.
Multiple meta-analyses find a real antidepressant effect, especially for more vigorous activity.
Open →SupportedThe placebo effect can genuinely reduce pain.
Placebos produce real, measurable changes in pain processing — though they do not cure underlying disease.
Open →RefutedCoffee is bad for your health.
Moderate coffee drinking is linked to lower, not higher, risk of death and several diseases.
Open →MixedA low-fat diet prevents heart disease.
A huge randomized trial found no reduction in heart disease from cutting overall dietary fat.
Open →RefutedResveratrol in red wine helps you live longer.
In people, dietary resveratrol levels showed no link to inflammation, disease, or lifespan.
Open →RefutedEating chocolate makes a country produce more Nobel laureates.
A real journal article showed the correlation — as a deliberate lesson that correlation is not causation.
Open →MixedVitamin D supplements prevent cancer and heart disease.
A large trial found no reduction in cancer or cardiovascular events overall, with only subtle possible signals.
Open →Refuted'Scared Straight' prison-visit programmes keep at-risk teens out of crime.
Randomized evidence shows these programmes increase, not decrease, later offending.
Open →RefutedThe D.A.R.E. programme stops kids from using drugs.
The original school-based D.A.R.E. curriculum had essentially no effect on drug use.
Open →MixedSmaller class sizes improve student achievement.
A landmark experiment found real gains in early grades, but effects are modest and costly to scale.
Open →ContestedDeworming pills dramatically improve children's schooling in poor countries.
An influential study found big attendance gains; Cochrane reviews of broader trials find little effect on health or learning.
Open →MixedMicrocredit lifts poor households out of poverty.
Six randomized trials found modest business effects but no transformative rise in income or well-being.
Open →SupportedMaking organ donation the default (opt-out) hugely raises consent.
Default settings powerfully shape stated consent, though actual transplant rates depend on more than defaults.
Open →ContestedCracking down on minor disorder ('broken windows') sharply cuts serious crime.
Visible disorder can spread disorder, but the leap to aggressive policing cutting serious crime is not well established.
Open →ContestedTaking notes by hand beats typing them on a laptop.
A famous study said yes, but a large replication failed to reproduce the advantage.
Open →ContestedBeing bilingual gives you a lasting boost in executive control.
Early studies suggested a cognitive edge, but larger, publication-bias-aware analyses find little or none.
Open →ContestedViolent video games make players violent.
Meta-analyses disagree sharply; any link to real-world aggression is small and hotly debated.
Open →MixedSocial media is destroying teenagers' mental health.
Large datasets show only a tiny average association; effects likely depend on the person and type of use.
Open →RefutedThe full moon triggers strange behaviour, crime, and hospital chaos.
A meta-analysis of lunar studies found no reliable link between the moon's phase and human behaviour.
Open →