Decision hygiene

Decision hygiene is a family of process safeguards designed to reduce unwanted variability, or noise, and sometimes bias before judgments are combined or outcomes are known. The term was popularized by Daniel Kahneman, Olivier Sibony, and Cass Sunstein in their 2021 book Noise.

The hygiene metaphor is preventive: a process is designed before anyone knows which judge, case, anchor, or moment will create an error. Typical examples include independent judgments, pre-specified criteria, structured interviews, decomposed assessments, base-rate checks, and noise audits.

Decision hygiene is not one standardized treatment with a single effect size. It is an umbrella framework. Evidence must be examined at the level of the particular practice, decision domain, comparator, and outcome.

Two qualified judges can reach different conclusions from the same case, and the same judge can vary across time or context. When that variation is unrelated to relevant facts, it is noise. Noise can reduce consistency and fairness even when average judgment is not biased in one direction.

Organizations often focus on bias because its direction is easier to describe. A noise audit asks a different question: how much do judgments spread when qualified people assess comparable cases independently? That question is useful in hiring, forecasting, underwriting, diagnosis, review, and other repeated judgments.

The practical case is strongest when a decision recurs, several judgments can be compared, and the relevant criteria can be specified. One-off personal choices may still benefit from structure, but the empirical base is thinner and the overhead may outweigh the benefit.

Noise brought together research and examples on judgment variability and gave the preventive practices a memorable label. The framework builds on older literatures: actuarial versus clinical prediction, structured interviews, forecasting and aggregation, base-rate use, blinding, checklists, and Gary Klein’s premortem.

Those components do not all have the same evidence. Mechanical combination has been studied in defined prediction and selection settings; structured interviews have their own personnel-research literature; the premortem is a practical technique described in Harvard Business Review, not proof that the entire package improves every strategic decision.

The Mediating Assessments Protocol is another process proposal: assess defined dimensions separately, discuss each before forming an overall impression, and delay holistic intuition. It is best described as a structured design for strategic judgment, with outcome evidence evaluated separately.

1. Define the decision and success criteria. Specify what is being judged and which information is relevant before reviewing a favored option.
2. Use an outside view. Compare the case with an appropriate reference class or base rate when one exists.
3. Collect independent judgments first. Prevent early comments, status, or anchors from making views artificially similar.
4. Decompose the judgment. Assess distinct dimensions before forming an overall impression; avoid double-counting correlated criteria.
5. Use structured questions or scoring where appropriate. Standardization can improve comparability, but only when the criteria are relevant and the scale is used consistently.
6. Consider failure before commitment. A premortem can surface plausible failure modes; it is a prompt, not a prediction.
7. Audit disagreement and outcomes. Measure how much judgments vary, then test whether process changes improve calibration, consistency, fairness, or the outcome that actually matters.

The checklist is adaptable rather than canonical. A practice that helps repeated hiring judgments may be unnecessary or harmful in a time-critical emergency.

It can make criteria explicit, reduce premature social influence, reveal disagreement, improve comparability, and create a process that can be audited and revised.

It cannot supply missing expertise, validate a poor model, identify the correct values, or eliminate uncertainty. A consistent process can be consistently wrong, and lower variance is not automatically fairer or more accurate.

The outcome measure matters. A process may reduce disagreement without improving calibration, or improve predictive accuracy while creating unacceptable costs elsewhere. Hygiene should be evaluated against the goal and constraints of the actual decision.

“Decision hygiene targets noise, never bias.” Noise is the central framing, but practices such as blinding, base-rate prompts, and pre-specified criteria may also address directional bias.

“The package is proven because some components work.” Evidence for a structured interview, a mechanical combination rule, or one training intervention does not establish a universal effect for every practice used together.

“More structure is always better.” Structure can suppress useful context, add friction, create gaming, or lock in the wrong criteria. The design must fit the decision.

“Decision hygiene is a personal productivity cure.” The concept primarily addresses judgment processes, especially repeated organizational decisions. Personal adaptations can be useful prompts, but they are not clinical treatment or a validated cure for indecision.

Consider a hiring committee evaluating four candidates for a senior role. The unstructured default is a meeting where each member shares overall impressions, the most senior or vocal member shapes the conversation, and the group converges on a candidate. This process can introduce noise: a different speaker order or first comment could influence how the committee discusses the same evidence.

The decision-hygiene version restructures the process. The committee defines five or six relevant assessments (technical skill, leadership track record, cultural addition, stakeholder feedback, and so on) before reviewing any candidate. Each member scores each candidate on each assessment independently, before any discussion, with anchored scoring scales. Only after the independent component scoring does the committee discuss — and the discussion focuses on disagreements, not on overall impressions.

The structured version is not free. It takes more setup time, requires committee discipline, and feels less satisfying than the natural conversational flow of the unstructured version. The structure is intended to reduce unwanted variability across committees, days, and speaker orders, and to make component judgments easier to compare. Whether it improves consistency or outcomes should be tested in the actual hiring process.

• Decision support system — software that may implement criteria, aggregation, models, or audit trails.
• Cognitive bias — systematic deviation relative to a specified benchmark.
• Bounded rationality — the constraints that make workable procedures necessary.
• Heuristic — a simplified strategy that may be useful or error-prone depending on the environment.
• Analysis paralysis — prolonged indecision; a different construct that may call for stopping rules rather than more process.

Decision-hygiene principles are woven through the design of all LifeByLogic decision-support tools. The Should I Quit framework embeds a 72-hour compression delay before action and decomposes the decision into structured components rather than asking for a holistic verdict. The Cognitive Bias Test surfaces per-bias scores rather than a single composite to discourage premature closure. The full design philosophy is documented on the editorial policy page.

Use the Should I Quit framework →

Decision hygiene is an umbrella label rather than one standardized intervention. The sources distinguish the book-level framework from evidence about individual techniques.

• Kahneman, Sibony, and Sunstein (2021), Noise — the source that popularized the hygiene metaphor and noise-audit framework.
• Kahneman, Rosenfield, Gandhi, and Blaser (2016), Noise — organizational overview of unwanted judgment variability.
• Klein (2007), Performing a Project Premortem — practical description of the premortem technique, not a broad outcome trial of decision hygiene.
• Morewedge and colleagues (2015) — experimental evidence for one debiasing-training intervention; it does not validate every hygiene practice.
• Kuncel, Klieger, Connelly, and Ones (2013) — meta-analysis comparing mechanical and holistic combination in selection contexts.
• Kahneman, Lovallo, and Sibony (2019) — description of the Mediating Assessments Protocol for strategic decisions.

This entry is intended as a citable scholarly reference. Choose the format that matches your context. The retrieval date should reflect when you accessed the page, which may differ from the entry's last-reviewed date shown above.

LifeByLogic. (2026). Decision Hygiene: Definition, Examples, and Checklist. https://lifebylogic.com/glossary/decision-hygiene/

LifeByLogic. "Decision Hygiene: Definition, Examples, and Checklist." LifeByLogic, 12 July 2026, https://lifebylogic.com/glossary/decision-hygiene/.

LifeByLogic. 2026. "Decision Hygiene: Definition, Examples, and Checklist." July 12. https://lifebylogic.com/glossary/decision-hygiene/.

@misc{lbldecisionhygiene2026,
  author = {{LifeByLogic}},
  title = {Decision Hygiene: Definition, Examples, and Checklist},
  year = {2026},
  month = {jul},
  publisher = {LifeByLogic},
  url = {https://lifebylogic.com/glossary/decision-hygiene/},
  note = {Accessed: 2026-07-12}
}