Is Most Published Research Wrong?

Caption notes

Derek Muller asks in this Veritasium video why statistically significant findings can disappear when a study is repeated. He uses a hypothetical example to show how false positives among published findings depend on the hypotheses being tested, a study’s ability to detect a real effect, and which results make it into print. He then uses a chocolate and weight loss headline to illustrate how looking at many outcomes and highlighting one significant result can mislead readers. He closes by discussing review of research plans before an experiment and publication based on following the agreed procedure.