meta-analysis
A formal statistical synthesis of results across many studies to estimate an average effect and explain variability among findings.
As defined in ms_152
Other definitions
A statistical method that pools results from many independent studies to estimate an overall effect.
How Games Make Life Better
Where it comes up · 18
- John Ioannidis on Statistical Significance, Economics, and Replication5 claims
- Julia Belluz on Epidemiology, Nutrition, and Metabolism3 claims
- Jacob Stegenga on Medical Nihilism3 claims
- How Games Make Life Better2 claims
- Rob Wiblin and Russ Roberts on Charity, Science, and Utilitarianism2 claims
- 318 | Edward Miguel on the Developing Practice of Development Economics2 claims
- Louis Menand on Psychiatry 05/31/20101 claim
- Intuitive Surgical: Robotic Precision1 claim
and 10 more
Most important claims · 10 of 29Search all claims for “meta-analysis” →
- With about 14 million da Vinci procedures and 38,000 peer-reviewed articles on robotic surgery, 3,000 published in the last year mostly independent of Intuitive, the clinical evidence base is a large barrier to entry that new systems cannot quickly replicate.Intuitive Surgical: Robotic Precision
- Randomized trials and meta-analyses are not infallible truth-tellers; they can be shot through with biases via malleability practices analogous to p-hacking—such as measuring many outcomes and publishing only a subset (one German study found only about 26% of pre-registered outcomes were published) and publication bias of entire trials.Jacob Stegenga on Medical Nihilism
- Methods are improving through pre-registration (publicly posting hypotheses and tests in advance) and through 'mega-studies' that test dozens of hypotheses simultaneously on the same outcome and timeframe, which accelerates science and makes it easier to publish null results alongside positive ones, though they carry risks of outlier skew and misrepresentation.Katy Milkman on How to Change
- The ivermectin-for-COVID evidence rests on small trials totaling about 2,200 participants, with the largest (from Egypt) shown by Gideon Meyerowitz-Katz to be highly irregular and likely fraudulent—including patients already dead before recruitment—so while there may be a weak signal worth pursuing, claims that ivermectin is 99% effective or reduces mortality by 70-90% are impossible and unsupported.256 - A Contagion Of Bad Ideas
- Meta-analysis is not automatically more reliable than a single study; if every underlying study is flawed, the meta-analysis yields a flawed result, but it still has value because it reveals the universe of studies and lets you see comparative patterns and footprints of bias rather than promising a definitive answer.John Ioannidis on Statistical Significance, Economics, and Replication
- A meta-analysis of hundreds of studies is not automatically more reliable; if all the underlying studies share the same flawed methodology, pooling them sheds no more light than a single bad study because they are all making the same error.Julia Belluz on Epidemiology, Nutrition, and Metabolism
- Across a meta-analysis of more than 500 papers from the past decade, the top predictor of whether kids reap gaming benefits such as better grades, less drug use, and more social support, versus harm, is whether they can articulate the real skills they build in games as part of who they are, rather than seeing games as escapist and separate from reality.How Games Make Life Better
- A large meta-analysis combining many studies found the correlation between how empathic a person is and their propensity for physical, verbal, and sexual aggression is virtually zero, so empathy tests are useless for predicting aggressiveness, contradicting the common belief that empathic people are nicer and low-empathy people are terrible.14 - The Virtues Of Cold Blood
- Rosiglitazone, a Type 2 diabetes drug, was on the market for years (and as of last check still available in the US) when a 2007 meta-analysis suggested it had caused something like 70,000 heart attacks; of roughly 45 trials testing it, only about 15 had been published.Jacob Stegenga on Medical Nihilism
- A September 2021 meta-analysis (Fan et al., Vaccines) found no statistical difference in adverse events between vaccine and control groups, with serious-event rates under one in a thousand and death/serious-event rates under one in a million—safer than the conventional vaccine standard of one serious event per 100,000-200,000—so the rational choice is vaccination given infection's ~1% death risk.270 - What Have We Learned From The Pandemic