The Education of Statistical Intuition

Thinking, Fast and Slow

This flow traces the repeated confrontation between vivid causal stories and the impersonal logic of statistics.

increasing

Bearing strength in each event

decreasing

Small Sample Surprise Small Sample Surprise from "Thinking, Fast and Slow". Kidney Cancer Counties Opposite Extremes Both the highest and lowest cancer rates cluster in small rural counties.

Small Sample Surprise from "Thinking, Fast and Slow". Kidney Cancer Counties Opposite Extremes Both the highest and lowest cancer rates cluster in small rural counties.

Kidney Cancer Counties Opposite Extremes Both the highest and lowest cancer rates cluster in small rural counties. Statistical Explanation Sampling variation, not rural virtue or vice, explains the pattern.

The Law of Small Numbers The Law of Small Numbers from "Thinking, Fast and Slow". Small Samples Misread Research Risk Psychologists often choose samples too small to reliably confirm true hypotheses.

The Law of Small Numbers from "Thinking, Fast and Slow". Small Samples Misread Research Risk Psychologists often choose samples too small to reliably confirm true hypotheses.

Small Samples Misread Research Risk Psychologists often choose samples too small to reliably confirm true hypotheses. Intuitive Overreach Researchers expect short sequences to resemble the population too closely.

Randomness as Pattern Randomness as Pattern from "Thinking, Fast and Slow". Chance Resists Stories Apparent Regularity Birth sequences and bombing patterns look intentional even when random.

Randomness as Pattern from "Thinking, Fast and Slow". Chance Resists Stories Apparent Regularity Birth sequences and bombing patterns look intentional even when random.

Chance Resists Stories Apparent Regularity Birth sequences and bombing patterns look intentional even when random. Causal Hunger System 1 searches for causes where sampling variation is sufficient.

Regression Misunderstood Regression Misunderstood from "Thinking, Fast and Slow". The Return to Average Performance Fluctuation Exceptional outcomes tend to be followed by more moderate ones when correlations are imperfect.

Regression Misunderstood from "Thinking, Fast and Slow". The Return to Average Performance Fluctuation Exceptional outcomes tend to be followed by more moderate ones when correlations are imperfect.

The Return to Average Performance Fluctuation Exceptional outcomes tend to be followed by more moderate ones when correlations are imperfect. False Causation Observers mistake natural regression for the effect of praise or punishment.

Statistical Humility Statistical Humility from "Thinking, Fast and Slow". Correcting Intuition Baselines Prediction should begin from average outcomes and plausible base rates.

Statistical Humility from "Thinking, Fast and Slow". Correcting Intuition Baselines Prediction should begin from average outcomes and plausible base rates.

Correcting Intuition Baselines Prediction should begin from average outcomes and plausible base rates. Moderation Specific evidence deserves influence only in proportion to its predictive validity.

At the beginning of Part 2

Early in Part 2

After the kidney cancer discussion

Middle of Part 2

Near the end of Part 2