5.3 Decision-making models and cognitive biases

Decision-Making Models

Decision-making models explain how people navigate choices. Normative models prescribe how decisions should be made, while descriptive models capture how people actually decide. The gap between these two tells us a lot about human cognition and why smart people still make poor choices.

Normative vs. Descriptive Decision Models

Normative models assume decision-makers are fully rational and aim to maximize expected utility. Two major examples:

• Expected Utility Theory says you should calculate the probability and value of each outcome, then pick the option with the highest expected payoff.
• Bayesian Decision Theory says you should update your beliefs systematically as new evidence comes in, using probability to weigh that evidence.

These models work well as benchmarks, but they don't reflect how people behave in practice.

Descriptive models account for the cognitive limitations and shortcuts people actually rely on:

• Bounded Rationality (Herbert Simon) argues that people don't optimize; they satisfice, meaning they pick the first option that's "good enough" because they have limited time, information, and mental resources.
• Prospect Theory (Kahneman & Tversky) shows that people evaluate outcomes relative to a reference point and are more sensitive to losses than to equivalent gains. Losing $50 feels roughly twice as bad as gaining $50 feels good.

Cognitive Biases in Decision-Making

Cognitive biases are systematic patterns of deviation from rational judgment. They often stem from heuristics (mental shortcuts) that are useful in everyday life but can distort decisions in predictable ways.

Common Cognitive Biases in Decisions

Confirmation bias is the tendency to seek out, interpret, and remember information that supports what you already believe. You overweight evidence that fits your view and underweight evidence that contradicts it. For example, if you think a certain study method works best, you'll notice the times it helped and forget the times it didn't.

Anchoring bias occurs when the first piece of information you encounter (the "anchor") disproportionately influences your judgment. In one classic study, participants who were asked whether the population of Turkey was more or less than 5 million gave much lower estimates than those anchored at 65 million. Even when you get new information, you tend to adjust insufficiently from that initial anchor.

Availability heuristic leads you to overestimate the likelihood of events that come to mind easily. Plane crashes and shark attacks feel more common than they are because they're vivid and heavily covered in media, while far more common risks (like car accidents) feel less threatening.

Framing effect means the way information is presented changes your decision. People respond differently to "90% survival rate" versus "10% mortality rate," even though these are logically identical.

Sunk cost fallacy is continuing to invest in something because of what you've already put in (time, money, effort), even when quitting would be the better choice. Sitting through a terrible movie because you paid for the ticket is a small-scale example; companies pouring millions into failing projects is a large-scale one.

Impact of Biases on Decision Processes

At the individual level, biases lead to suboptimal judgments in predictable ways. Overconfidence, for instance, causes people to overestimate their own knowledge or abilities, which can result in poor planning or risky choices. Biases can also cause you to ignore relevant information or give too much weight to irrelevant factors.

At the group level, biases interact with social dynamics and can get amplified:

1. Groupthink occurs when a group prioritizes consensus and harmony over critical evaluation. Members suppress dissent, and the group converges on a decision without adequately considering alternatives.
2. Group polarization happens when group discussion pushes members toward more extreme positions than they held individually. Depending on the group's initial leaning, this can produce a risky shift (toward riskier choices) or a cautious shift (toward more conservative ones).
3. Group settings can either amplify or reduce individual biases depending on the structure of the discussion and whether dissent is encouraged.

Strategies for Mitigating Cognitive Biases

No one can eliminate biases entirely, but structured approaches help reduce their influence:

• Seek diverse perspectives. Input from people with different backgrounds and expertise surfaces blind spots. Environments that welcome constructive dissent make it harder for confirmation bias and groupthink to take hold.
• Use structured decision-making frameworks:
  1. Apply tools like decision trees or multi-criteria decision analysis to organize your reasoning.
  2. Establish clear criteria and weightings for evaluating options before you start comparing them.
• Build self-awareness through metacognition. Reflecting on your own thought process helps you catch biases in action. Training on common biases makes them easier to recognize.
• Apply specific debiasing techniques:
  1. Actively consider alternatives and counterarguments to your initial judgment (this directly combats confirmation bias).
  2. Use reference class forecasting: instead of estimating from scratch, look at how similar past situations actually turned out.
  3. Conduct a premortem analysis: before committing to a decision, imagine it has already failed and work backward to identify what could go wrong.