How We're Fooled By Statistics | Veritasium
Appreciation
7
Importance
6
Date Added
9.5.25
TLDR
Regression to the mean: extreme results (good or bad) tend to be followed by results closer to the average. So very poor performing baselines tend toward improving the next time you measure.
2 Cents
Very neat 7 minute video and I think of the studies he mentions (particularly fighter pilots) sometimes when reading medical studies!
Tags
Cool studies mentioned:
- In Israeli fighter pilot training, they thought that negative reinforcement worked better than positive reinforcement, because chastising the particularly poor performances always led to improvement while encouraging particularly good performances was ineffective or even detrimental. It turns out regression to the mean was the phenomenon in effect here.
- Teacher gives positive feedback → student spends more time on task.
- Rugby player film review. If given positive feedback, blood tests revealed that a week later testosterone was higher than those who received negative feedback. Not just psychological effect, actual phsyiological differences.