One statistical analysis must not rule them all

Any single analysis hides an iceberg of uncertainty. Multi-team analysis can reveal it.

COMMENT 17 May 2022 Any single analysis hides an iceberg of uncertainty. Multi-team analysis can reveal it. By Eric-Jan Wagenmakers0, Alexandra Sarafoglou1 & Balazs Aczel2 Eric-Jan Wagenmakers Eric-Jan Wagenmakers is a methodologist at the University of Amsterdam, the Netherlands. Alexandra Sarafoglou Alexandra Sarafoglou is a postdoctoral fellow at the University of Amsterdam, the Netherlands. Balazs Aczel Balazs Aczel is vice-dean for science at Eötvös Loránd University in Budapest, Hungary. Illustration by David Parkins A typical journal article contains the results of only…

saved by

• Michelle Wang

related reading

• Beware the Casual Polymath | Applied Divinity Studiesapplieddivinitystudies.com
• A statistical approach to model evaluations \ Anthropicanthropic.com
• 2012.00174arxiv.org
• Unblock research bottlenecks with non-profit start-upsnature.com
• The trouble in comparing different approaches to science fundingscienceplusplus.org
• How Should We Critique Research? · Gwern.netgwern.net
• Too Much Information - by Ben Recht - arg minargmin.net
• Your intuitions are not magic — LessWronglesswrong.com
• Quantifying the effect of experimental perturbations at single-cell resolutionnature.com
• Embrace teams large and small to foster the health of researchnature.com
• “Antedisciplinary” Science | PLOS Computational Biologyjournals.plos.org
• Why an overreliance on AI-driven modelling is bad for sciencenature.com