Module I·Article III·~1 min read

Cognitive Biases: A Catalog of Systemic Errors

Cognitive Biases

Turn this article into a podcast

Pick voices, format, length — AI generates the audio

What Are Cognitive Biases

Cognitive biases are systematic errors in judgments that arise from simplified patterns of thinking (heuristics). They are predictable and universal: all people are subject to them regardless of intelligence.

Key Biases in Business

Anchoring: the first number heard becomes an “anchor” for all subsequent estimates. The first price mentioned in negotiations establishes the discussion range.

Confirmation Bias: we seek information that confirms an already formed opinion and ignore information that contradicts it. Managers in love with their own idea fail to notice red flags.

Availability Heuristic: the probability of an event is assessed by how easily its examples come to mind. After an air crash, people overestimate the risk of flying.

Sunk Cost Fallacy: “We have already invested so much money/time, we can’t stop now.” Sunk costs should not influence the decision to continue.

Overconfidence Bias: people systematically overestimate the accuracy of their forecasts. 90% of drivers consider themselves “better than average.”

Planning Fallacy: projects systematically take more time and money than planned. We ignore the “outside view” (base rate) in favor of an optimistic “inside view.”

Halo Effect: a positive impression in one area spreads to others. “Attractive and confident” → “smart and competent.”

How to Combat Biases

  • Pre-mortem: “The project failed. Why?” — forms oppositional thinking
  • Devil’s advocate: an intentionally appointed critic
  • Base rates: always ask “what is the average result in similar projects?”
  • Separation of roles: analyst and decision-maker — different people

Practical Assignment

Choose a current decision in your organization (project, hiring, investment). Go through the list of 7 biases above. (1) Which of them may affect this decision? (2) What data or perspectives might you be ignoring? (3) What procedures could protect against these biases?

§ Act · what next