Confirmation Bias
Source: Wason, Quarterly Journal of Experimental Psychology, 1960; Nickerson, 1998 Institution: Multiple
Finding
Wason’s 2-4-6 task demonstrated that subjects seek confirming rather than disconfirming evidence. They test hypotheses by looking for examples that fit, not examples that would break their rule. This is honesty violated at the strategy level, not content level: the subject may believe each individual datum honestly, but the search strategy is biased toward confirmation. Nickerson (1998) provides comprehensive review.
Pattern Mapping
Honesty — Violated at the epistemic strategy level, not at the level of individual claims. The subject is not lying about any specific finding; the search itself is dishonest.
Non-fabrication — Confirmation bias fabricates support by omission. By not seeking disconfirming evidence, the subject constructs an artificially supportive evidence base.
Humility — The biased subject refuses to test hypothesis boundaries. They do not ask “under what conditions would I be wrong?” This is humility failure at the methodological level.
Connections
- System 1 and System 2 — confirmation bias is a System 1 strategy (→ Meta-Pattern 06 - Self-Reference and Instrument Trap)
- Dunning-Kruger Effect — both involve the inability to evaluate one’s own cognitive process
- Left-Right Asymmetry — L-R asymmetry traces honestly to its initial signal; confirmation bias dishonestly preserves its initial hypothesis
- Kidney and Nephron Filtration — the kidney sorts by actual molecular properties; confirmation bias sorts by hypothesis compatibility — honest vs dishonest filtering
- Selfie and Social Media — social media algorithms are confirmation bias made technological: showing users what confirms their existing views
Status
Robust finding. Wason (1960) is foundational. Nickerson (1998) is the definitive review. No controversy on core effect.
The mapping to the five properties is this project’s structural interpretation.