Thinking and Decision Making | IB

The study of thinking and decision making examines how people reason, judge, and choose. Research in This area has revealed that human thinking is subject to systematic biases and heuristics — mental Shortcuts that are often useful but can lead to predictable errors. This research has had Far-reaching implications for economics, law, medicine, and public policy, and was recognised by the Award of the Nobel Prize in Economics to Daniel Kahneman in 2002.

The heuristics and biases programme, developed by Amos Tversky and Daniel Kahneman in the 1970s, Demonstrated that people rely on simple mental shortcuts (heuristics) to make judgments under Uncertainty. These heuristics reduce the cognitive effort required for complex judgments but Systematically bias the resulting decisions.

The availability heuristic is the tendency to judge the frequency or probability of an event by the Ease with which instances come to mind. Events that are more vivid, recent, emotionally salient, or Extensively covered in the media are more “available” in memory and are therefore judged to be more Common.

Example: After extensive media coverage of plane crashes, people may overestimate the Probability of dying in a plane crash relative to a car accident, even though the statistical Probability of dying in a car accident is far higher (approximately 1 in 100 for a lifetime of Driving versus approximately 1 in 11 million for a single flight).

Tversky and Kahneman (1973) demonstrated the availability heuristic by asking participants whether More words in English begin with the letter “K” or have “K” as the third letter. Most participants Judged that more words begin with “K,” but in fact approximately twice as many words have “K” as the Third letter. Words beginning with “K” are more retrieved from memory (because we search for Words by their first letter), making them more “available.”

Strengths of the availability heuristic: It is generally adaptive. Events that are Recalled tend to be those that occur frequently or are important, so the heuristic often produces Reasonable estimates with minimal cognitive effort.

Limitations: The heuristic is systematically biased by factors unrelated to actual frequency, Such as media coverage, emotional salience, and personal experience.

The representativeness heuristic is the tendency to judge the probability that an object or event Belongs to a category based on how similar (representative) it is to the typical member of that Category. People tend to ignore base rates (the actual statistical frequencies) and focus instead on The similarity between the specific case and the category prototype.

The Linda problem (Tversky and Kahneman, 1983):

Participants were told: “Linda is 31 years old, single, outspoken, and very bright. She majored in Philosophy. As a student, she was deeply concerned with issues of discrimination and social justice, And also participated in anti-nuclear demonstrations.”

Participants were then asked which was more probable:

(A) Linda is a bank teller. (B) Linda is a bank teller and is active in the feminist movement.

Approximately 85% of participants chose option (B). This is a conjunction fallacy: the probability Of a conjunction (A and B) can never be greater than the probability of either component alone (A). Linda”s description is more representative of a feminist bank teller than of a bank teller in General, leading participants to violate the laws of probability.

The anchoring heuristic is the tendency for numerical estimates to be biased toward an initial value (the anchor), even when the anchor is irrelevant. Once an anchor is established, adjustments From the anchor are insufficient.

Example: Tversky and Kahneman (1974) asked participants to estimate the percentage of African Countries in the United Nations. Before answering, participants watched the experimenter spin a Rigged roulette wheel that stopped at either 10 or 65. Participants who saw the anchor of 10 Estimated approximately 25%, while those who saw the anchor of 65 estimated approximately 45%. A random and irrelevant number significantly influenced participants’ numerical estimates.

Mechanism: Anchoring occurs through at least two processes: (1) insufficient adjustment from the Anchor, and (2) selective activation of anchor-consistent information. The effect is robust across Many domains and is difficult to eliminate, even when participants are explicitly warned about it.

Confirmation bias is the tendency to search for, interpret, and remember information in a way that Confirms one’s pre-existing beliefs or hypotheses, while discounting or ignoring disconfirming Evidence.

Wason (1960): The 2-4-6 task. Participants were told that the number sequence 2-4-6 follows a Rule and were asked to discover the rule by generating their own three-number sequences and Receiving feedback on whether each sequence conformed to the rule. Most participants tested Sequences consistent with their initial hypothesis (e.g., “increasing even numbers”) and announced The rule after receiving only confirming evidence. Very few participants tested sequences designed To disconfirm their hypothesis (e.g., 10-7-4 or 3-1-5). The actual rule was “any three Increasing numbers.”

Implications: Confirmation bias contributes to the persistence of false beliefs, stereotyping, And poor decision making in scientific, medical, and legal contexts. It is particularly problematic Because people are generally unaware of the bias and believe that their reasoning is objective.

Hindsight bias (“the I-knew-it-all-along effect”) is the tendency, after an event has occurred, to Overestimate the extent to which the outcome could have been predicted beforehand. Hindsight bias Distorts memory for past judgments, making outcomes seem more inevitable than they actually were.

Fischhoff (1975): Participants were given descriptions of a historical event (the 1814 Battle of Waterloo) and asked to estimate the probability of several possible outcomes. Some participants were Told the actual outcome before making their estimates; these participants assigned significantly Higher probabilities to the actual outcome than participants who were not told the outcome. When Participants who had not been told the outcome were later informed of it and asked to recall their Original estimates, they recalled estimates closer to the actual outcome than their original Estimates had actually been.

Implications: Hindsight bias impairs learning from experience (because past judgments are Misremembered), undermines accountability (because outcomes seem inevitable in retrospect), and Contributes to overconfidence in judgment.

Daniel Kahneman, in his book “Thinking, Fast and Slow,” proposed that human cognition is governed by Two distinct systems:

System 1 (Fast Thinking):

• Automatic, unconscious, and effortless.
• Operates quickly, producing intuitive judgments with minimal cognitive effort.
• Relies on heuristics and associations.
• Is always active and cannot be turned off.
• Is responsible for impressions, intuitions, and feelings.
• Examples: recognising a face, reading words on a billboard, answering 2 + 2 = 4.

System 2 (Slow Thinking):

• Controlled, conscious, and effortful.
• Operates slowly, requiring deliberate attention and cognitive resources.
• Is capable of logical reasoning, complex computation, and careful evaluation.
• Is lazy: it tends to accept System 1’s suggestions without scrutiny unless motivated or prompted to do otherwise.
• Is capacity-limited: it can only focus on one demanding task at a time.
• Examples: solving 17 x 24, comparing the value of two complex options, monitoring one’s behaviour in a social situation.

Interaction between the systems: In most situations, System 1 generates automatic impressions And intuitions, which System 2 may endorse with minimal scrutiny (leading to heuristic-based biases) Or may override with more deliberate reasoning. Cognitive biases arise when System 2 fails to Monitor and correct System 1’s output. This failure is more likely when:

• Cognitive resources are depleted (fatigue, stress, time pressure).
• The task is not personally motivating.
• System 1’s output feels subjectively confident.
• The individual lacks relevant expertise in the domain.

Evaluation:

• The dual-process framework provides an integrative explanation for a wide range of cognitive biases and errors.
• It has been influential across multiple disciplines, including economics, law, medicine, and public policy.
• Critics argue that the framework is oversimplified: some researchers have identified more than two types of processing (e.g., Epstein’s cognitive-experiential self-theory, Stanovich’s tripartite model). The distinction between System 1 and System 2 may be more of a useful heuristic than a precise description of cognitive architecture.
• The neural correlates of System 1 and System 2 processing have been identified in neuroimaging studies, lending some biological support to the distinction (e.g., increased prefrontal cortex activation during System 2 tasks).

Prospect theory is a descriptive model of decision making under risk that challenges the expected Utility theory of classical economics. Expected utility theory assumes that people are rational Agents who evaluate decisions based on the expected value (probability multiplied by outcome) and Are risk-neutral. Prospect theory demonstrates that people systematically violate these assumptions.

1. Reference dependence: People evaluate outcomes relative to a reference point ( the Status quo), not in absolute terms. A gain of USD 100 feels very different depending on whether the Reference point is the expectation of receiving USD 50 (in which case it is a gain of USD 50 above Expectations) or the expectation of receiving USD 200 (in which case it is a loss of USD 100 below Expectations).

2. Loss aversion: Losses loom larger than equivalent gains. The psychological impact of losing USD 100 is approximately twice as great as the psychological impact of gaining USD 100. This Asymmetry is reflected in the value function, which is steeper for losses than for gains.

3. Diminishing sensitivity: The marginal psychological impact of each additional unit of gain (or loss) decreases as the magnitude of the gain (or loss) increases. The difference between USD 0 And USD 100 feels larger than the difference between USD 900 and USD 1000.

4. Probability weighting: People do not evaluate probabilities linearly. Small probabilities are Overweighted (leading to risk-seeking behaviour for gains and risk-averse behaviour for losses — Explaining the purchase of lottery tickets and insurance), while moderate to high probabilities are Underweighted.

The Asian disease problem (Tversky and Kahneman, 1981):

Participants were asked to choose between two programmes to combat a disease expected to kill 600 People.

Framing 1 (gain frame):

• Programme A: 200 people will be saved.
• Programme B: 1/3 probability that 600 people will be saved, 2/3 probability that no one will be saved.
• 72% chose Programme A (risk-averse).

Framing 2 (loss frame):

• Programme C: 400 people will die.
• Programme D: 1/3 probability that nobody will die, 2/3 probability that 600 people will die.
• 78% chose Programme D (risk-seeking).

The two framings describe identical outcomes (Programme A = Programme C; Programme B = Programme D), But the framing in terms of lives saved versus lives lost produced a complete reversal of risk Preferences. This demonstrates that people are risk-averse when options are framed as gains and Risk-seeking when options are framed as losses, consistent with the S-shaped value function of Prospect theory.

Dan Ariely’s research extends the heuristics and biases programme by demonstrating systematic Irrationalities in everyday decision making. Key findings include:

• The power of free: People are disproportionately attracted to “free” options, even when the free option is not the best value. For example, participants offered a choice between a USD 0.01 Lindt truffle and a USD 0.15 Hershey’s Kiss overwhelmingly chose the truffle. When the prices were reduced by one cent (truffle: USD 0.00; Kiss: USD 0.14), participants overwhelmingly chose the free Kiss, even though the relative value had not changed.
• Arbitrary coherence: Initial prices serve as anchors that influence subsequent valuations, even when the initial prices are arbitrary. Once an anchor is established, prices for related items are judged relative to that anchor, creating a coherent but arbitrary price structure.
• Decoy effect: The addition of an inferior option (a decoy) can shift preferences between two other options. If option A is superior to option B on one dimension but inferior on another, adding an option A’ (which is similar to A but slightly worse on the dimension where A is superior) makes A look more attractive relative to B.

For an overview of cognitive topics, see Cognitive Level of Analysis.

1. Writing pseudocode that is too language-specific rather than using standard algorithmic constructs.

2. Mixing up Big O, Big $\Omega$, and Big $\Theta$ notation. Big O is an upper bound, not necessarily tight.

3. Confusing authentication (who you are) with authorisation (what you can do) in security contexts.

4. Forgetting edge cases in algorithm design (e.g., empty input, single element, already sorted data).

flowchart TD
    A[Thinking And Decision Making] --> B[Key Concepts]
    A --> C[Core Principles]
    A --> D[Practical Applications]
    B --> E[Fundamental definitions]
    C --> F[Design patterns]
    D --> G[Real-world usage]

The key principles covered in this topic are linked in the sub-pages above. Focus on understanding the definitions, applying the formulas or frameworks, and evaluating strengths and limitations of each approach.

Worked examples demonstrating the application of key concepts are covered in the detailed sub-pages linked above.

• Research Methods provides the experimental methodology used to study cognitive biases and decision-making processes.
• Nervous System explains the neural mechanisms underlying perception, memory, and the cognitive processes studied in decision-making research.
• Cell Biology covers the cellular foundations of neural function that enable the cognitive processes examined in thinking and judgement.

Human thinking is not the perfectly rational, calculator-like process we often assume it to be. Instead, our brains rely on quick shortcuts—heuristics—that evolved to help us make fast decisions with limited information. These shortcuts work remarkably well in everyday life, but they come with predictable blind spots. When someone judges the frequency of a car accident by how efficiently they remember news stories about it, or when they cling to a first impression despite contradictory evidence, they are experiencing the natural side effects of a system designed for speed over accuracy. Understanding these biases is not about learning that humans are irrational; it is about recognising that the very mechanisms that make us effective thinkers also make us systematically vulnerable to certain kinds of errors.

The dual-process framework offers a useful way to think about this tension. System 1 handles the vast majority of our daily cognition—reading facial expressions, navigating familiar environments, making snap judgments—and it does so effortlessly and mostly well. System 2, the slower, more deliberate mode of thinking, is what we engage when we pause to double-check a calculation or consider whether our initial reaction was warranted. The key insight is that System 2 is lazy: it defaults to accepting System 1’s suggestions unless something specifically prompts it to intervene. This means that biases often persist not because people cannot think critically, but because they are not prompted or motivated to do so.

Prospect theory deepens this picture by showing that we do not evaluate outcomes in absolute terms but relative to where we start. Losses sting roughly twice as much as equivalent gains feel good, and the way a choice is framed—whether we hear about lives saved or lives lost—can flip our preferences entirely. Together, these findings reveal that our decisions are shaped not just by the facts of a situation but by the psychological lens through which we view those facts. Recognising these patterns is the first step toward making more informed choices, both individually and in the design of policies, organisations, and everyday systems.

1. Confusing System 1 and System 2 with “good” and “bad” thinking. System 1 (fast, automatic) is not inferior to System 2 (slow, deliberate) — it is optimised for speed and efficiency in familiar situations. Biases arise when System 1’s shortcuts are applied to problems that require deliberate analysis. The goal is not to eliminate System 1 but to recognise when System 2 needs to intervene.

2. Misunderstanding the framing effect. The framing effect shows that identical information presented differently (e.g., “90% survival rate” vs. “10% mortality rate”) produces different preferences. This is not irrationality — it reflects how the brain processes gains versus losses. When evaluating decisions, reframe the problem in both gain and loss terms to check whether your preference is stable.

3. Assuming heuristics are always biases. Heuristics (availability, representativeness, anchoring) are generally useful cognitive shortcuts that produce correct judgments most of the time. Biases occur in specific, predictable situations where the heuristic fails. Do not dismiss all heuristic reasoning as flawed — understand when it works and when it does not.

4. Confusing loss aversion with risk aversion. Loss aversion means losses loom larger than equivalent gains (the value function is steeper for losses). Risk aversion means preferring a sure outcome to a gamble with equal expected value. They are related but distinct: a loss-averse person may still take risks if the potential gain is large enough relative to the potential loss.

5. Applying prospect theory to all decision contexts. Prospect theory was developed for decisions under risk with known probabilities. Many real-world decisions involve ambiguity (unknown probabilities), social preferences, or long time horizons where other models (expected utility theory, hyperbolic discounting) may be more appropriate. Choose the framework that matches the decision context.