Thinking Fast and Slow - Study Notes - Version 2: 17 - Chapter 17: Regression to the Mean

Chapter 17: Regression to the Mean

e.g. story of Israeli pilots, praise/condemnation, afterwards better/worse performance

great success = a little more talent + a lot of luck

e.g. pro golfer who does v. well on Day 1 has above-avg talent + above-avg luck, prob’ly less well on Day 2 — regression to the mean
more extreme original score, more regression expected (extremely good score suggests very lucky)
regression has no causality, random fluctuations

correlation coefficient (0 to 1) betw. 2 measures desc.’s relative weight of factors they share —
e.g. correlation betw. size of objects measured English or in metric units is 1 (any factor that infl’s one measure also influences the other; 100% of determinants are shared) — e.g. corr. betw. family income & last four digits of phone number is 0
correlation & regression are the same thing, viewed diff’ly
if corr. betw. two scores is imperfect, will be regression to mean — e.g. highly intelligent women tend to marry men less intelligent than they are, but this is not a causal story — corr. betw. intelligence scores of spouses is less than perfect
regression has an explan’n but not a cause — S1 confuses corr. with cause
regression a difficult concept, our minds are biased toward causal explan’ns, not stats — so much that S1 nearly blocks S2’s attempt to underst.

“She says experience has taught her that criticism is more effective than praise. What she doesn’t understand is that it’s all due to regression to the mean.”
“Perhaps his second interview was less impressive than the first because he was afraid of disappointing us, but more likely it was his first that was unusually good.”
“Our screening procedure is good but not perfect, so we should anticipate regression. We shouldn’t be surprised that the very best candidates often fail to meet our expectations.”

^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-^-

Intuitive predict’ns need to be corrected because they are not regressive and therefore are biased.