Counterintuitive scientific findings, verified and well-sourced. No clickbait. No sensationalism. Just the research.
Using a novel statistical method that tested every plausible analytic path through three massive datasets, Oxford researchers found that digital technology use accounts for at most 0.4% of the variation in adolescent well-being, an effect so small that potatoes and corrective lenses have comparable or larger negative associations.
The belief that learning piano or violin boosts children's IQ, reading, and math skills has driven school policy and parenting decisions for three decades. A comprehensive meta-analysis found that when studies use proper controls, music training produces no cognitive or academic benefit whatsoever.
The idea that self-control is a finite resource, backed by 600+ published studies and taught in every psychology textbook, collapsed when 59 laboratories ran coordinated replications. The observed effect was an order of magnitude smaller than the original meta-analytic estimate.
A nationally representative study found that patients rating their healthcare highest had greater hospital admissions, 9% higher drug expenditures, and 26% increased mortality risk. A 2019 replication with nearly 93,000 participants found an even stronger effect.
A national analysis of Medicare hospital admissions found that 30-day mortality rose steadily with physician age, from 10.8% for doctors under 40 to 12.1% for those over 60. The effect disappeared for high-volume physicians.
A systematic review of 25 studies across six countries found that recreational runners develop knee and hip osteoarthritis at barely one-third the rate of sedentary people. Competitive athletes do face higher risk. But for the tens of millions who jog a few times a week, the couch appears to be the bigger threat to their joints.
Acetaminophen is the most widely used drug ingredient in the United States. A pair of randomized controlled trials at Ohio State found it doesn't just numb your own pain. It numbs your response to everyone else's.
For two decades, organizations mandated that survivors of disasters, accidents, and violence sit down with a counselor within hours and relive what happened. Fifteen randomized trials found no benefit. Two found it caused harm.
Sixty randomized controlled trials. 4,554 participants. Patients told the pills were sugar pills. They improved anyway — especially the sickest ones.
The Hospital Readmissions Reduction Program fined hospitals nearly $2 billion for readmitting patients within 30 days. A JAMA study of 8.3 million Medicare hospitalizations found the policy was associated with roughly 10,000 excess deaths from heart failure and pneumonia.
A randomized controlled trial of 21 strength-trained men found that 10 minutes of cold water immersion after every workout reduced muscle mass gains by 67%. A 2021 meta-analysis of eight studies confirmed the damage.
An original analysis of 345 best-of-7 series finds Game 7s occur 15% less often than probability models predict. The conspiracy theory relies on a decade-old mathematical error.
A 2024 meta-analysis of 10,189 participants found that venting anger has an effect size of essentially zero. Deep breathing and meditation cut anger by a medium-large margin.
Across 26 articles, more than 400 experiments, and over 5 million research subjects, a single structural failure keeps disproved beliefs alive.
A Finnish RCT of 186 workers found that those told to stay in bed recovered slowest. A Cochrane review confirmed it.
A study of 33 surgical residents found that those who played video games more than three hours a week committed 37% fewer errors and completed laparoscopic procedures 27% faster than their non-gaming colleagues.
An analysis of S&P 500 firms found no improvement in profits or stock value after return-to-office mandates. A separate study tracking 3 million workers found the best employees left first. A randomized trial found hybrid work cut attrition by a third with zero productivity loss.
A meta-analysis of 107 cohort studies tracking 4.8 million people found that the apparent longevity benefit of one or two drinks a day vanishes entirely once you stop counting sick former drinkers as "abstainers."
A special issue of the American Economic Journal brought together six independent randomized controlled trials spanning India, Morocco, Ethiopia, Bosnia-Herzegovina, Mongolia, and Mexico. The combined evidence: microcredit modestly increases business activity but does not raise incomes, reduce poverty, or empower women.
A Monash University meta-analysis of 13 randomized controlled trials found no evidence that breakfast aids weight loss. Breakfast eaters consumed 260 extra calories per day and weighed slightly more than those who skipped it entirely.
The Finnish FIDELITY trial randomized 146 patients with degenerative meniscal tears to arthroscopic partial meniscectomy or sham surgery. At 12 months, outcomes were identical. At 10 years, the surgery group had more osteoarthritis and three times the rate of knee replacement.
A JAMA meta-analysis pooling 23 double-blind, placebo-controlled experiments found that sugar has no measurable effect on children's behavior or cognitive performance. The "sugar high" persists as a cultural belief driven almost entirely by parental expectation.
A randomized trial of 1,440 households in rural Kenya found that unconditional cash transfers raised consumption by 22%, increased assets by 61%, and improved psychological well-being, while spending on alcohol and tobacco did not increase.
A landmark review examined 50 years of learning-styles research and found virtually no evidence that matching instruction to a student's preferred style improves outcomes.
A landmark review of 10 study techniques across hundreds of experiments rated highlighting, rereading, and summarization as low utility. Only practice testing and distributed practice earned the highest rating.
A meta-analysis of 66 studies spanning 45 years found that acute sleep deprivation produces rapid antidepressant effects in roughly half of depressed patients.
A study of 118 burn patients found nighttime injuries took 28 days to heal versus 17 for daytime burns, driven by circadian rhythms in individual skin cells.
A Cochrane systematic review of 946 juveniles across nine experiments found that prison-visit deterrence programs raised the odds of reoffending by 68%.
A Drexel University review of 196 experiments across 12 studies found that potted plants remove VOCs so slowly that you'd need 10 to 1,000 plants per square meter to match basic ventilation.
A natural experiment exploiting Indiana's 2006 statewide adoption of DST found it increased residential electricity consumption by 1–4%, costing households $9 million per year.
The most comprehensive meta-analysis of financial education research found that teaching people about compound interest, budgeting, and debt management accounts for essentially none of their actual financial decisions.
Two Fortune 500 companies outfitted employees with sociometric sensors before and after converting cubicles to open plans. Face-to-face conversation plummeted while email and instant messaging surged.
A Cochrane review of 78 trials and 296,707 participants found that beta-carotene and vitamin E supplements increase all-cause mortality.
Experiments at the University of Waterloo found that positive self-statements backfired for those who needed them most.
A meta-analysis of 20 studies found that brainstorming groups generate fewer ideas than the same number of people working independently.
A systematic review of 25 RCTs found stretching had no injury prevention benefit, while strength training cut risk by over a third.
A study of 332 adults across five populations found total energy expenditure plateaus above moderate activity. Your body compensates.
Consumers were 10× more likely to buy jam when offered 6 options instead of 24. More choice led to worse decisions and lower satisfaction.
A pre-registered Oxford study of 1,004 adolescents using objective violence ratings found zero relationship with aggressive behavior.