Why Most Published Research Findings Are False

From AcaWiki
Jump to: navigation, search

Citation: John P. A. Ioannidis (2005/08/30) Why Most Published Research Findings Are False. PLOS Medicine (RSS)
DOI (original publisher): 10.1371/journal.pmed.0020124
Semantic Scholar (metadata): 10.1371/journal.pmed.0020124
Sci-Hub (fulltext): 10.1371/journal.pmed.0020124
Internet Archive Scholar (search for fulltext): Why Most Published Research Findings Are False
Download: https://doi.org/10.1371/journal.pmed.0020124
Tagged: reproducibility (RSS)

Summary

  • "It can be proven that most claimed research findings are false."
  • Most studies are not replicated; there is little incentive.
  • {Small sample size, small effect size, considering many variables with no selection criteria, flexibility of design, financial interests, and "hotness" of the field} all contribute to uncertainty in the conclusion.
  • {Experimental design, more data, replication studies, meta studies, hypothesis registration (sometimes not desirable), better stats (understanding prior odds)} all may help.