Chapter 15: Key Takeaways
Core Concepts
-
Some cognitive biases are robust. Confirmation bias, loss aversion, sunk cost fallacy, availability heuristic, framing effects, and overconfidence have survived the replication crisis with strong evidence.
-
Some famous bias findings have not survived. Social priming effects (elderly priming, warm cup, money priming) have largely failed to replicate. Ego depletion failed. IAT has poor reliability and weak behavioral prediction.
-
Many "biases" are adaptive heuristics. Gigerenzer's ecological rationality perspective shows that cognitive shortcuts often work well in real-world environments. They look like "errors" mainly in artificial laboratory conditions or when judged against inappropriate formal standards.
-
"Humans are fundamentally irrational" is an oversimplification. The more accurate claim: humans use imperfect but generally functional cognitive tools that can produce systematic errors, especially in unfamiliar or evolutionarily novel environments.
-
Knowing about biases has limited debiasing effect. Structural interventions (choice architecture, checklists, group deliberation) outperform awareness-based approaches.
-
The "bias bias" is real. Seeing biases everywhere, using bias identification as an argument weapon, and ignoring the adaptive function of heuristics are themselves forms of biased thinking.
-
The Dunning-Kruger effect is real but overused. The core finding (low performers overestimate) has support, though the magnitude is debated. Its popular use as a universal dismissal of disagreement is ironic and unhelpful.
Evidence Ratings in This Chapter
| Claim | Rating | Summary |
|---|---|---|
| "Humans are fundamentally irrational" | ⚠️ OVERSIMPLIFIED | Real biases exist, but many are adaptive; the blanket irrationality narrative overgeneralizes |
| "Social priming reliably changes behavior" | ❌ DEBUNKED | Multiple high-profile failures to replicate |
| "Confirmation bias is real" | ✅ SUPPORTED | One of the most robust findings in psychology |
| "Loss aversion is real" | ✅ SUPPORTED | Well-replicated, though exact magnitude varies |
| "Knowing about biases helps overcome them" | ⚠️ OVERSIMPLIFIED | Awareness has limited effect; structural interventions work better |
| "The Dunning-Kruger effect explains why people disagree with me" | ⚠️ OVERSIMPLIFIED | Core effect has support; popular application is overextended and partly artifactual |
Key Terms Introduced
- Heuristics: Cognitive shortcuts that simplify decision-making
- Ecological rationality: The perspective that heuristics are adapted to specific environments and are rational within those environments (Gigerenzer)
- Wanting vs. liking in biases: Some heuristics produce outcomes people pursue (wanting) but don't necessarily prefer upon reflection (liking) — paralleling Berridge's dopamine distinction
- The bias bias: The tendency to see biases everywhere, use bias identification as a weapon, and ignore the adaptive function of heuristics
- Debiasing: Interventions designed to reduce the impact of cognitive biases on decisions
One Sentence to Remember
Cognitive biases are real but the pop narrative overstates them — many are adaptive shortcuts, some famous ones haven't replicated, knowing about them doesn't fix them, and "you're biased" is not an argument.