Noise: A Flaw in Human Judgment

Noise: A Flaw in Human Judgment is a nonfiction book by professors Daniel Kahneman, Olivier Sibony and Cass Sunstein. It was first published on May 18, 2021. The book concerns 'noise' in human judgment and decision-making. The authors define noise in human judgment as "undesirable variability in judgments of the same problem" and focus on the statistical properties and psychological perspectives of the issue.Examples they give include their own finding at an insurance company that the median premiums set by underwriters independently for the same five fictive customers varied by 55 %, five times as much as expected by most underwriters and their executives. Another example is that two psychiatrists who independently diagnosed 426 state hospital patients agreed on which mental illness the patient suffered from only in half of the cases and a finding that French court judges were more lenient if it happened to be the defendant’s birthday.Kahneman, Sibony and Sunstein argue that noise in human judgment is a thoroughly prevalent and insufficiently addressed problem in matters of judgment. They write that noise arises because of factors such as cognitive biases, mood, group dynamics and emotional reactions. While contrasting statistical bias to noise, they describe cognitive bias as a significant factor giving rise to both statistical bias and noise. The authors write that noise can lead to gross injustices, unacceptable health hazards, and loss of time and wealth. They argue that organizations should be more committed to reducing noise and promote noise audits and decision hygiene as strategies to detect, measure, and prevent noise. Noise: A Flaw in Human Judgment became a The New York Times Bestseller and received generally positive reviews among critics. Common critiques against efforts to reduce noise are that such efforts dehumanize those affected by the judgments and that it can lead to discrimination. Some commentators also questioned the authors' claims about the novelty of the noise concept.