As a student of critical thinking and decision-making, I was more than surprised while reading Daniel Kahneman (author of Thinking, Fast and Slow), Olivier Sibony, and Cass Sunstein’s book, Noise, A Flaw in Human Judgment. The authors highlight the importance of a specific type of error in human judgment: noise, or the random variability of decisions amongst a group of people presented with the same information. They painstakingly explain why noise matters, why there’s so much more of it than we realize, and how to reduce it.
Within the first few pages, the authors clarified the distinction between noise and bias, I was shocked at the effects and magnitude of noise in the decision-making process! Prison sentences that vary widely from different judges, contradictory diagnoses from different physicians, and different evaluations of workplace performance are just a few examples of decisions that are rife with noise.
The book is divided into six parts:
- Finding Noise
- Your Mind is a Measuring Instrument
- Noise in Predictive Judgments
- How Noise Happens
- Improving Judgments
- Optimal Noise
The chapters take you through identifying noise in different sectors, comparing noise in human decisions to AI algorithms, determining the causes of noise, and ultimately how to reduce noise through noise audits and other “decision hygiene” practices.
The final (and best) section deals with how much noise is tolerable, and what levels can be realistically reached. They also share ways to reduce noise in different settings: personal methods, such as training people to continually update their views with contrary perspectives; and collective approaches such as how to best leverage the expertise of a team charged with an important business decision.
Warning: The book can be a bit pedantic (did the authors get paid by the word?), but well worth the read!
Related Articles:
Our Decisions are Influenced by the Way the Choices are Posed
Butterfly Choices: Making Decisions That Matter
Reduce The “Illusion of Causality” – Improve Critical Thinking
