System noise is undesirable variability in judgments
System noise is “undesirable variability in the judgments of the same case by multiple individuals. We have identified its two major components, which can be separated when the same individuals evaluate multiple cases:
- Level noise is variability in the average level of judgments by different judges.
- Pattern noise is variability in judges’ responses to particular cases.”[1]
See also:
- Good decision-making does not mix values with facts
- Data must be interpreted in context and without noise
Noise – Kahneman, et. al. (2021), ch. 6, § “The Components of Noise.” ↩︎