By Daniel Kahneman, Olivier Sibony, Cass R. Sunstein
Understanding unwanted variability in judgment
Subject
Themes
Unwanted variability in professional judgment poses a serious problem across organisations, yet remains largely unrecognised. This variability, termed noise, differs fundamentally from bias: whilst bias represents systematic errors in one direction, noise reflects random scatter in decisions that should ideally be identical. Understanding noise requires distinguishing it from legitimate diversity in taste or productive competition amongst ideas. System noise specifically describes unwanted variability where outcomes depend arbitrarily on which professional happens to be assigned to a case.
Unwanted variability in professional judgment poses a serious problem across organisations, yet remains largely unrecognised. This variability, termed noise, differs fundamentally from bias: whilst bias represents systematic errors in one direction, noise reflects random scatter in decisions that…
Noise represents the culmination of decades of research on judgment and decision-making, building directly on Daniel Kahneman's earlier work whilst extending it in novel directions. Kahneman's 2011 bestseller Thinking, Fast and Slow popularised the heuristics and biases research programme,…
Human judgment exhibits a fundamental flaw that remains largely invisible yet profoundly consequential: unwanted variability that produces different outcomes for identical situations. This noise, as distinguished from bias, reflects random scatter rather than systematic deviation. Understanding why…
The criminal justice system, designed to deliver equal treatment under law, instead operates like a lottery where outcomes depend arbitrarily on which judge happens to preside. This system noise creates profound injustice, as identical cases receive drastically different punishments without rational…
Sixty years of research demonstrates a counterintuitive finding that challenges assumptions about expertise: simple mechanical rules consistently outperform human judgment in predictive tasks. This superiority stems primarily from noise elimination rather than superior insight. Understanding why…
Reducing noise in professional judgment raises profound moral questions about consistency, flexibility, transparency, and human dignity. Whilst noise demonstrably harms organisations and individuals through arbitrary outcomes, efforts to eliminate it confront legitimate concerns about mechanical…
Despite its ubiquity and profound consequences, noise remains largely invisible in organisations and professional practice. This invisibility puzzle poses a fundamental challenge: if noise causes such serious problems, why do people systematically fail to recognise it? Understanding this blindness…
Two kinds of error affect human judgment: bias represents systematic deviation in one direction, whilst noise reflects random scatter around a target. Both produce inaccuracy, but noise remains far less recognised despite being equally damaging. The shooting range metaphor illustrates this…
Professional judgment functions as a measuring instrument that should produce consistent readings. When physicians diagnose patients, judges evaluate defendants, or teachers grade essays, their assessments should remain stable across occasions and consistent across professionals. Yet reality…
Groups amplify rather than reduce noise when members influence one another. Information cascades occur when early speakers anchor subsequent discussion, causing later participants to suppress private information contradicting the emerging consensus. Group polarisation intensifies initial tendencies…
Psychological mechanisms generating judgment errors extend the heuristics and biases research programme established over fifty years. System 1 thinking relies on simplifying operations called heuristics that generally prove useful but sometimes produce systematic errors. These psychological biases…
Improving judgment requires understanding both what causes errors and what strategies effectively reduce them. Selecting better judges represents one approach: some people consistently make more accurate predictions than others. Research on forecasting tournaments identifies superforecasters whose…
Medical decision-making demonstrates both noise's consequences and effective mitigation strategies. Physicians diagnosing the same patients from identical information frequently reach different conclusions. When the same physician sees the same case twice without recognising it, significant…
Noise reduction involves costs that must be weighed against benefits. Perfect consistency requires resources for developing guidelines, training professionals, monitoring compliance, and updating procedures as circumstances change. These investments prove worthwhile when stakes are high (criminal…
By Daniel Kahneman, Olivier Sibony, Cass R. Sunstein
Subject
Themes
Unwanted variability in professional judgment poses a serious problem across organisations, yet remains largely unrecognised. This variability, termed noise, differs fundamentally from bias: whilst bias represents systematic errors in one direction, noise reflects random scatter in decisions that should ideally be identical. Understanding noise requires distinguishing it from legitimate diversity in taste or productive competition amongst ideas. System noise specifically describes unwanted variability where outcomes depend arbitrarily on which professional happens to be assigned to a case.