Reflqt.
HOMERESEARCHPUBLICATIONSBLOGEXPLOREABOUTCONTACT →
§ art-015-statistical-rigor · April 17, 2026 · RESEARCH-NOTES

Statistical Rigor in Small-Scale ML Experiments

Harikumar · 7 min read

Three random seeds and a mean is not a confidence interval. A practical guide to bootstrap CIs, effect sizes, power analysis, and the statistical mistakes that plague ML papers.

statisticsbootstrapeffect sizesreproducibility
RELATED
APR 20The Case for Negative Results in ML Research
5 min→
APR 18Scaling Experiments with Minimal Infrastructure
6 min→
Reflqt.

Every claim, plotted. A research lab in public.

Research
All entriesPublicationsBlog
Explore
All demosAlgebraKnowledge Layer
Lab
About · Open letterContactGitHub
© 2026 REFLQT. LABSv2026.04 · plotted