John B. Holbein @JohnHolbein1
John B. Holbein @JohnHolbein1
Look at the results of clinical trials before and after the preregistration scholars' study design became a requirement.
What do you notice?
[Embedded scatter plot: "Relative risk of primary outcome" (y-axis, 0 to 1.6) vs "Publication year" (x-axis, 1974-2014). Vertical line at year 2000 labeled "Year 2000: Registration of primary outcomes required on ClinicalTrials.gov". Points are marked as harm (red no-entry symbol), null (teal filled circle), or benefit (circled plus). Before 2000, many points show "benefit" (relative risk well below 1, down to ~0.15). After 2000, almost all points cluster near 1.0 (null), with far fewer benefit points and one harm point.]
Note from Claude Sonnet 5
A tweet illustrating the effect of clinical-trial preregistration on reported effect sizes — dramatic drop in "benefit" findings after 2000 when outcome registration became mandatory, a classic illustration of publication bias / p-hacking mitigation. General science-methodology content (relevant to Nathan's epistemics/replication interests), not AI-specific.
twitterscience-methodologypreregistrationpublication-biasclinical-trialsreplication