Problem Solving and Decision Making

Learning Objectives

By the end of this page, you should be able to:

• Distinguish well-defined from ill-defined problems and algorithms from heuristics
• Explain functional fixedness and mental set with classic experimental evidence
• Describe key heuristics and biases (availability, representativeness, anchoring) and how they cause predictable errors
• Explain Kahneman's dual-process theory (System 1 and System 2)
• Compare rational choice theory, prospect theory, and dual-process theory as models of decision-making
• Apply these concepts to evaluate real decisions and identify likely biases

Quick Answer

Problem solving is the cognitive process of moving from a current state to a desired goal state when the path isn't obvious, while decision-making is choosing among alternatives, usually under uncertainty. Cognitive psychologists study both because human reasoning is not simply logical calculation — it's shaped by mental shortcuts (heuristics) that are usually efficient but sometimes systematically wrong (biases). This matters practically because understanding these patterns explains real-world errors: why doctors sometimes misdiagnose based on how a case "resembles" a typical one, why investors panic-sell during market dips, and why juries can be swayed by information they were told to ignore. Daniel Kahneman and Amos Tversky's decades of research on this topic won a Nobel Prize in Economics and reshaped how psychology, economics, and medicine understand human choice.

Core Concepts

Well-Defined vs. Ill-Defined Problems, and Algorithms vs. Heuristics

Definition: A well-defined problem has a clear starting state, goal state, and set of legal steps to solve it; an ill-defined problem lacks at least one of these. An algorithm is a step-by-step procedure guaranteed to produce a correct solution; a heuristic is a mental shortcut or rule of thumb that usually works but isn't guaranteed.

Explanation: Well-defined problems (like a chess puzzle or an arithmetic equation) can, in principle, be solved with an algorithm — an exhaustive, guaranteed method — but algorithms are often too slow or resource-intensive for real-world use, especially for complex or ill-defined problems (like "how should I plan my career?"). Heuristics trade guaranteed accuracy for speed, which usually pays off but produces predictable, specific errors.

Example: Solving 4,382 × 17 by long multiplication is algorithmic (guaranteed correct if followed properly); estimating that the answer is "somewhere around 70,000-80,000" by rounding is heuristic (fast, usually close, sometimes wrong).

Real-World Example: Chess computers historically relied on brute-force algorithms checking millions of positions, while human grandmasters rely heavily on heuristics — pattern recognition of familiar board configurations — allowing rapid, "intuitive" strong moves without exhaustive calculation, though this occasionally leads them to overlook a rare, non-typical winning move that doesn't match a familiar pattern.

Why It Matters: Recognizing when a problem is well-defined versus ill-defined tells you whether an algorithmic approach is even possible — many important real-world decisions (career choice, ethical dilemmas, medical diagnosis with ambiguous symptoms) are ill-defined and can't be "solved" the way a math problem can, which is why they require different strategies (like breaking the problem down or generating multiple hypotheses).

Common Misunderstanding: Students think heuristics are simply "wrong" or a lesser form of thinking compared to algorithms. In reality, heuristics are usually adaptive and necessary — no one has time to run an exhaustive algorithm for every daily decision, and heuristics only become "biases" when their shortcuts systematically misfire in specific, predictable situations.

Functional Fixedness and Mental Set

Definition: Functional fixedness is the tendency to see an object only in terms of its typical use, blocking creative solutions; mental set is the tendency to keep using a problem-solving strategy that worked before, even when a better approach is available.

Explanation: Both are forms of cognitive rigidity — the mind defaults to familiar patterns because they're usually efficient, but this becomes a liability when the familiar pattern doesn't fit the new problem.

Example / Real-World Example: Karl Duncker's classic 1945 "candle problem" asked participants to attach a lit candle to a wall using only a box of tacks, a candle, and matches, so wax wouldn't drip on the table below. Most people struggled because they perceived the box only as a container for the tacks (functional fixedness) rather than as a potential platform — the solution is to tack the (now empty) box to the wall and set the candle on top of it. When the tacks were presented outside the box instead of inside it, far more participants solved the problem quickly, because the box was no longer pre-categorized as "a container."

Why It Matters: Understanding functional fixedness has directly shaped creativity-training methods and design thinking approaches, which deliberately push people to list unconventional uses for familiar objects before problem-solving, breaking the default categorization.

Common Misunderstanding: Students think functional fixedness only affects "less intelligent" people. It's a universal cognitive tendency that even experts fall into — in fact, expertise can sometimes worsen functional fixedness within that expert's own domain, because strong habitual associations are harder to override.

Heuristics and Biases (Kahneman and Tversky)

Definition: Systematic patterns of deviation from purely rational judgment, arising from the mental shortcuts (heuristics) people use to make fast decisions under uncertainty.

Explanation: Kahneman and Tversky identified several specific heuristics, each producing a specific, predictable bias:

• Availability heuristic: judging the likelihood of an event by how easily examples come to mind. Vivid or recent events (like a plane crash reported extensively on the news) feel more probable than they statistically are, while less memorable risks (like heart disease) are underestimated.
• Representativeness heuristic: judging probability by how closely something resembles a typical case, while ignoring actual statistical base rates. Tversky and Kahneman's famous "Linda problem" (1983) described a hypothetical woman, Linda, as outspoken and concerned with social justice, then asked whether it's more probable Linda is "a bank teller" or "a bank teller active in the feminist movement." A majority of participants — including many with statistics training — chose the second option, even though it is logically impossible for a conjunction (two conditions) to be more probable than one of its parts alone. This is called the conjunction fallacy.
• Anchoring and adjustment: relying too heavily on an initial piece of information (the "anchor") when making subsequent judgments, even when the anchor is arbitrary. In one classic study, Tversky and Kahneman spun a rigged wheel of fortune that landed on either 10 or 65, then asked participants to estimate the percentage of African countries in the United Nations. Participants who saw the number 65 gave significantly higher estimates than those who saw 10 — despite the wheel being obviously random and irrelevant to the question.

Example: After watching news coverage of a shark attack, a swimmer overestimates the danger of sharks (availability), even though car accidents are statistically far more dangerous.

Real-World Example: In real estate, initial listing prices act as anchors — research shows that even professional real estate agents' valuations of a house are influenced by the listing price they're shown first, despite claiming the listing price didn't affect their independent judgment.

Why It Matters: These biases are not rare quirks — they operate in medical diagnosis (representativeness leading a doctor to overweight a "textbook" symptom pattern), financial decision-making (anchoring on a stock's previous high price), and legal judgments (anchoring on a prosecutor's opening sentencing suggestion). Recognizing them is the first step to designing "choice architecture" or decision checklists that counteract them.

Common Misunderstanding: Students think knowing about a bias automatically protects you from it. Research shows that biases like anchoring and availability persist even in experts who are explicitly aware of them and warned in advance — they are largely automatic and require deliberate structural counters (like checklists or blind review processes), not just willpower.

Dual-Process Theory: System 1 and System 2

Definition: Kahneman's model proposing that human thinking operates through two systems: System 1 is fast, automatic, intuitive, and effortless; System 2 is slow, deliberate, analytical, and effortful.

Explanation: Most everyday judgments are made by System 1, which relies on heuristics and produces fast, "good enough" answers with minimal mental effort. System 2 is engaged for effortful, novel, or high-stakes reasoning, but it's also lazy — it tends to accept System 1's automatic answer unless something triggers closer scrutiny (surprise, conflict, or explicit instruction to slow down).

Example: Answering "2 + 2" is System 1 (instant, automatic); answering "24 × 17" typically requires System 2 (deliberate calculation).

Real-World Example: The famous "bat and ball" problem — a bat and ball together cost $1.10, the bat costs $1.00 more than the ball, how much does the ball cost? — reliably produces the intuitive but wrong System 1 answer of "$0.10" (the correct answer is $0.05) in a majority of respondents, including at elite universities, because System 1 generates a fast, plausible-sounding answer and System 2 often fails to double-check it.

Why It Matters: This model explains why simply "trying harder" doesn't fix most everyday errors — the fix is structural: building in deliberate checkpoints (like the "sleep on it" rule for big financial decisions, or a formal checklist in aviation and surgery) that force System 2 to engage rather than relying on System 1's fast but fallible judgment.

Common Misunderstanding: Students often think System 1 is simply "bad" and System 2 is "good." System 1 is essential and usually correct — it's how experts make fast, accurate judgments (like a chess grandmaster's intuitive move or an experienced doctor's rapid pattern recognition); the problem arises specifically when System 1's shortcuts are applied to situations where they systematically mislead.

Models of Decision-Making

Definition: Rational choice theory assumes decision-makers weigh all available information and choose the option with the highest expected value; prospect theory (Kahneman & Tversky, 1979) describes how people actually value gains and losses relative to a reference point, rather than in absolute terms.

Explanation: Prospect theory's key finding is loss aversion: losses loom psychologically larger than equivalent gains — losing $100 feels roughly twice as painful as gaining $100 feels good. It also shows people are risk-averse for gains (preferring a sure smaller gain over a risky larger one) but risk-seeking for losses (preferring a risky chance to avoid a loss over accepting a smaller certain loss).

Example: Most people prefer a guaranteed $50 over a 50% chance of $100 (risk-averse for gains), but prefer a 50% chance of losing $100 over a guaranteed loss of $50 (risk-seeking for losses) — even though the expected value is identical in both pairs.

Real-World Example: Investors often hold onto losing stocks far longer than is rational, hoping to "avoid locking in the loss," even when selling and reinvesting elsewhere would be the statistically better choice — a direct real-world consequence of loss aversion documented extensively in behavioral finance research.

Why It Matters: Prospect theory revolutionized economics by showing that classical rational choice theory's assumptions don't match real human behavior, leading to the field of behavioral economics and practical applications like "nudges" in public policy (e.g., framing retirement savings as an opt-out default rather than opt-in, which dramatically increases participation because of loss aversion and status-quo bias).

Common Misunderstanding: Students think prospect theory says people are simply "irrational." It more precisely shows people are consistently, predictably non-rational in specific, measurable ways relative to strict expected-value logic — which actually makes their behavior easier to model and even anticipate, not more chaotic.

Visual Learning

This map shows the exam-relevant chain: whether a problem is well- or ill-defined determines whether an algorithm is even an option, and most everyday ill-defined judgments run through System 1's heuristics first — which is exactly where the classic biases creep in.

Real-World Applications

• Medicine: Understanding representativeness and availability heuristics has led to "cognitive debiasing" training for doctors and structured diagnostic checklists to counter premature pattern-matching in diagnosis.
• Behavioral economics and public policy: Prospect theory and loss aversion underlie "nudge" policies, such as automatic enrollment in retirement savings plans, which dramatically increase participation compared to opt-in systems.
• Law: Anchoring effects influence jury damage awards and sentencing recommendations, prompting research into how initial numbers presented in court should be regulated or contextualized.
• Business and design: Functional fixedness research informs brainstorming and design-thinking techniques that deliberately force people to reconsider default categorizations of objects and processes.
• Aviation and surgery: Checklists are a direct structural application of dual-process theory — forcing System 2 engagement at critical decision points rather than relying on fast, potentially error-prone System 1 judgments under pressure.

Key Terms

Term	Definition
Well-defined problem	A problem with a clear start state, goal state, and set of legal solution steps
Ill-defined problem	A problem lacking a clear goal state or set of solution steps
Algorithm	A step-by-step procedure guaranteed to produce a correct solution
Heuristic	A mental shortcut that usually produces a good, fast solution but isn't guaranteed to be correct
Functional fixedness	The tendency to see an object only in terms of its typical use
Mental set	The tendency to keep applying a previously successful strategy even when it's no longer optimal
Availability heuristic	Judging probability by how easily examples come to mind
Representativeness heuristic	Judging probability by resemblance to a typical case, ignoring base rates
Anchoring	Relying too heavily on an initial reference point when making judgments
Conjunction fallacy	The error of judging a conjunction of two events as more probable than one event alone
System 1	Fast, automatic, intuitive, low-effort thinking
System 2	Slow, deliberate, analytical, high-effort thinking
Prospect theory	A model describing how people value gains and losses relative to a reference point, including loss aversion
Loss aversion	The tendency for losses to be felt more strongly than equivalent gains

Common Mistakes

Misconception 1: Heuristics are simply "wrong" ways of thinking that should be eliminated. Why it's wrong: Heuristics are generally adaptive, efficient strategies that produce good outcomes most of the time; they only become biases in the specific situations where their assumptions fail. Correct understanding: The goal isn't to eliminate heuristic thinking (which is often impossible and undesirable, given time constraints) but to recognize the specific conditions under which they mislead and add structural checks in those situations.

Misconception 2: If Linda's description sounds like a feminist bank teller, then "bank teller and feminist" really is more likely than "just bank teller." Why it's wrong: This ignores basic probability logic — the probability of two things being true together (a conjunction) can never exceed the probability of either one alone, regardless of how well the description matches a stereotype. Correct understanding: The representativeness heuristic causes people to substitute "how well does this match my mental image" for the actual, correct question, "what is the statistical probability," which is the conjunction fallacy Tversky and Kahneman documented.

Misconception 3: Knowing about a cognitive bias is usually enough to avoid it. Why it's wrong: Studies repeatedly show that anchoring, availability, and other biases persist even among experts who are explicitly warned about them beforehand, because these processes are largely automatic (System 1) rather than under full conscious control. Correct understanding: Reducing bias generally requires structural interventions — checklists, blind review, forced consideration of alternative hypotheses — not just awareness or willpower alone.

Comparison and Connections

Feature	Rational Choice Theory	Prospect Theory	Dual-Process Theory
Core assumption	People choose the option with the highest expected value	People value gains/losses relative to a reference point, with loss aversion	People use two distinct systems, one fast and one slow
Explains real behavior well?	Poorly — doesn't predict risk-seeking for losses or loss aversion	Well — accounts for framing effects, loss aversion, risk-reversal	Well — accounts for why fast, biased judgments dominate everyday choices
Origin	Classical economics	Kahneman & Tversky (1979), psychology/behavioral economics	Kahneman's later synthesis (2011, Thinking, Fast and Slow)
Key phenomenon explained	"Ideal" rational decision-making (a normative benchmark)	Loss aversion, framing effects	Why biases like anchoring and availability happen automatically

Practice Questions

Recall

1. Define algorithm and heuristic, and give one example of each. Answer guidance: An algorithm is a guaranteed step-by-step procedure (e.g., long division); a heuristic is a fast mental shortcut without a guarantee of correctness (e.g., estimating an answer by rounding).
2. What is loss aversion, and who developed the theory that describes it? Answer guidance: Loss aversion is the tendency for losses to feel psychologically more painful than equivalent gains feel pleasurable; it's a core finding of prospect theory, developed by Daniel Kahneman and Amos Tversky (1979).

Understanding

3. Explain why Duncker's candle problem is considered strong evidence for functional fixedness, and how changing the setup affected results. Answer guidance: Participants struggled to see the tack box as a candle platform because they'd pre-categorized it as a container for tacks (its typical function). When the tacks were presented outside the box, more people solved the problem quickly, because the box was no longer perceptually "locked" into the container role — showing the fixation was about categorization, not the difficulty of the physical solution itself.
4. Explain the conjunction fallacy using the Linda problem, and why it's a logical, not just an intuitive, error. Answer guidance: Participants judged "bank teller and feminist" as more probable than "bank teller" alone because the description matched a feminist stereotype (representativeness). This is a logical error because the probability of two events both being true (a conjunction) can never be greater than the probability of either individually — P(A and B) ≤ P(A) is a basic law of probability.

Application

5. A hiring manager sees an extremely high starting-salary figure mentioned early in a negotiation and ends up offering more than budgeted, even after "adjusting" downward. Which heuristic explains this, and how would you counteract it? Answer guidance: Anchoring — the initial high figure serves as a reference point that skews the "adjustment" process, even when consciously judged to be too high. Counteracting it requires setting an independent budget range before hearing any figures, similar to how researchers control for anchors experimentally.
6. A student, worried about airplane crashes after seeing extensive news coverage, decides to drive a long distance instead of flying. Using heuristic research, explain the flaw in this reasoning. Answer guidance: This reflects the availability heuristic — vivid, heavily covered events (plane crashes) feel more probable than they statistically are, while common risks (car accidents, which are statistically far more dangerous per mile traveled) are underweighted because they're less memorable and less reported.

Analysis

7. Compare how System 1 and System 2 would each approach the "bat and ball" problem, and explain why System 2 often fails to intervene even when it's available. Answer guidance: System 1 quickly generates the plausible-sounding but wrong answer ($0.10) based on a simple subtraction pattern; System 2 could correctly work through the algebra ($0.05) but is "lazy" and tends to accept System 1's confident, fluent-feeling answer without double-checking unless something explicitly signals the need for closer scrutiny.
8. Analyze why prospect theory was considered a major challenge to classical economic theory (rational choice theory). Answer guidance: Classical economics assumed people are rational agents who maximize expected value consistently. Prospect theory showed people's choices flip depending on how options are framed (as gains vs. losses) and that losses are weighted roughly twice as heavily as equivalent gains — behavior that violates the consistency assumptions rational choice theory requires, which is why it helped found behavioral economics as a distinct field.

FAQ

Are heuristics always bad or wrong? No — heuristics are generally efficient and produce good results in most everyday situations; they become "biases" (systematic errors) only in specific circumstances where their built-in assumptions don't hold, such as when vivid media coverage distorts your sense of real statistical risk.

Can experts avoid these biases because they know more? Not automatically. Studies show experts (doctors, judges, financial analysts) remain susceptible to anchoring, availability, and representativeness biases in their own domains, partly because these processes are largely automatic (System 1). Structured tools like checklists and blind reviews are more effective than expertise alone.

What's the difference between a bias and a heuristic? A heuristic is the mental shortcut or strategy itself (e.g., judging by resemblance to a typical case); a bias is the specific, predictable error that results when that heuristic is applied in a situation where it misleads (e.g., the conjunction fallacy in the Linda problem).

Is System 1 thinking always less reliable than System 2? No — System 1 is often highly accurate, especially for experts operating within their trained domain (a chess grandmaster's rapid intuitive move, an experienced nurse noticing something "off" about a patient). The issue is that System 1 is applied to everything by default, including novel or statistical problems where it systematically fails.

How is prospect theory used outside of psychology? It's foundational to behavioral economics and is directly applied in public policy "nudges" (like opt-out retirement savings enrollment), marketing (framing prices as "avoiding a loss" rather than "gaining a discount"), and financial regulation aimed at protecting consumers from predictable decision-making errors.

Quick Revision

• Well-defined problems have a clear goal/path; ill-defined problems don't — algorithms only work reliably for well-defined problems.
• Functional fixedness (Duncker's candle problem): fixating on an object's typical use blocks creative solutions.
• Availability heuristic: judging probability by how easily examples come to mind (overestimating shark attacks after news coverage).
• Representativeness heuristic: judging probability by resemblance to a stereotype, ignoring base rates (Linda problem, conjunction fallacy).
• Anchoring: initial arbitrary numbers skew later judgments (the rigged wheel/UN estimate study).
• System 1 = fast, automatic, intuitive; System 2 = slow, deliberate, effortful (Kahneman's dual-process theory).
• The bat-and-ball problem shows System 2 often fails to override System 1's fast, wrong answer.
• Prospect theory (Kahneman & Tversky, 1979): people are risk-averse for gains, risk-seeking for losses, and losses feel worse than equivalent gains feel good (loss aversion).
• Rational choice theory is the classical "ideal" model; prospect theory better matches actual human behavior.
• Awareness of a bias alone rarely prevents it — structural fixes (checklists, blind procedures) work better.

Related Topics

Prerequisites: Introduction to Cognitive Psychology, Memory and Learning

Related Topics: Behavioral economics, judgment and decision-making research, social cognition and stereotyping

Next Topics: Language and Thought, Cognitive Development and Aging