This is a Lexicon entry: the mechanism, when it shows up, and the countermeasure. The full correction, with real-world cases and the audit prompt, hasn't been written yet.
Definition
The habit of favouring information that confirms our existing beliefs while ignoring contradictory evidence.
When it shows up
Researching a topic you already have a view on. Reading news sources you trust. Evaluating a new hire, investment or relationship after you have formed an initial impression. Any belief you would feel stupid for being wrong about. Modern context: A data analyst builds a dashboard to test a product hypothesis, then iterates the chart parameters until the visualisation supports the hypothesis - genuinely believing they are doing rigorous analysis, not realising the parameter choices were driven by the desired conclusion. High-stakes instance: Intelligence analysis - assessments have systematically filtered incoming data through existing threat models, missing signals that contradicted the prevailing picture until events forced revision. Failure mode: The search process is asymmetric: you generate queries, sources and interpretations that could verify the current view and unconsciously skip or deprioritise ones that could falsify it. The evidence gathered is filtered before it reaches the conclusion. Estimated cost: High - often irreversible where this bias dominates unchecked decisions.
Failure mode
The search process is asymmetric: you generate queries, sources and interpretations that could verify the current view and unconsciously skip or deprioritise ones that could falsify it. The evidence gathered is filtered before it reaches the conclusion.
Countermeasure
Ask: 'What specific evidence would change my mind?' Then go look for that evidence. If you cannot name the disconfirmer in advance, the belief is immune to evidence and you are confirming. When correction costs more than the bias: In domains where you have genuine expertise and your prior is well-calibrated, seeking confirming evidence is efficient - you are exploiting accumulated knowledge rather than exploring unnecessarily. The bias becomes a problem when priors are wrong and the stakes of being wrong are high. For low-stakes decisions with well-calibrated priors: confirming evidence search is rational.
Related traps
Also connected in the map24 more, locked
Credit & first seen
DiscoveredWason, P.C. (1960).On the failure to eliminate hypotheses in a conceptual task.Quarterly Journal of Experimental Psychology 12(3), 129-140 source ↗
Also revealed
Entry #14 of 631 in The Lexicon · see the full Lexicon