Problem Solving

1. What Is a Problem?

In the psychology of cognition, a problem has a precise structure. It consists of an initial state (the situation as it is now), a goal state (the situation you want to reach), and a set of operators — the moves or actions you are permitted to make to transform one state into another. There are usually constraints that rule out certain moves. Solving the problem means finding a sequence of legal operations that connects the initial state to the goal. A locked room with a key on a high shelf is a problem only if you cannot simply reach up and grab it; the gap between wanting the key and having it is what defines the task.

Not all problems are alike, and one of the most important distinctions is between well-defined and ill-defined problems. A well-defined problem has a clear initial state, an explicit goal, and known rules — solving a Rubik's cube, a sudoku puzzle, or an algebra equation. An ill-defined problem lacks one or more of these: the goal may be vague, the operators uncertain, or success hard to recognize. "Write a good novel," "improve team morale," or "live a meaningful life" are ill-defined. Real-world problems are usually ill-defined, which is part of why laboratory findings on tidy puzzles transfer only partly to everyday life.

Routine Versus Adaptive Problems

Psychologists also separate routine problems, which you solve by retrieving a known procedure, from non-routine or adaptive problems, which require constructing a new approach. Multiplying two numbers you know how to multiply is routine. Figuring out an unfamiliar kind of puzzle for the first time is adaptive. Genuine problem solving, in the strict sense, refers to the adaptive case: the moment when no rehearsed answer is available and the mind must build a path.

2. Theoretical Background and Key Researchers

The Behaviorist Starting Point

Early experimental work framed problem solving as trial and error. Edward Thorndike's puzzle-box studies with cats, conducted around the turn of the twentieth century, suggested that animals gradually learned to escape by stamping in successful responses and stamping out unsuccessful ones — a process he called the law of effect, which later fed into operant conditioning. On this view, solving a problem was incremental and mechanical, with no sudden understanding involved.

The Gestalt Revolution

Gestalt psychologists challenged the trial-and-error picture in the 1910s through the 1940s. Wolfgang Köhler, studying chimpanzees on Tenerife, reported that an ape confronted with out-of-reach bananas would sometimes sit, appear to ponder, and then suddenly assemble sticks or stack boxes to reach the fruit — behavior he interpreted as insight rather than blind trial and error. Karl Duncker introduced the famous candle problem and the concept of functional fixedness, while Max Wertheimer emphasized "productive thinking," in which a solver reorganizes the structure of a problem to see it in a new way. For the Gestalt school, the key event in problem solving was restructuring: a change in how the whole situation is mentally represented.

The Information-Processing Approach

The cognitive revolution of the 1950s and 1960s recast problem solving in the language of computation. Allen Newell and Herbert Simon, drawing on early artificial intelligence, proposed that a problem solver searches a problem space — the set of all states reachable from the initial state by applying operators. Their landmark book Human Problem Solving (1972) used "think-aloud" protocols, in which participants verbalize their reasoning, to map how people actually navigate this space. They showed that people rarely search exhaustively; instead they use heuristics that prune the space to a manageable size. Simon's broader idea of bounded rationality — that human reasoning is constrained by limited time, information, and mental capacity — remains foundational across psychology and economics, and connects directly to the study of cognitive biases.

Modern Extensions

Later researchers broadened the field. Work on analogical reasoning showed that people often solve new problems by mapping them onto familiar ones. Studies of expertise, beginning with research on chess masters, revealed that skilled solvers perceive problems differently from novices. And dual-process theories popularized by Daniel Kahneman distinguished fast, intuitive thinking from slow, deliberate reasoning, clarifying when heuristics help and when they mislead.

3. The Stages of Problem Solving

Although real problem solving is rarely a neat sequence, it is useful to describe it as a series of stages. A widely taught version, often credited to the IDEAL framework and to classic accounts of productive thinking, runs roughly as follows.

Identifying and Representing the Problem

The first and most consequential stage is recognizing that a problem exists and forming a mental representation of it. How you represent a problem largely determines whether you can solve it. A poor representation — focusing on surface features, mis-stating the goal, or smuggling in false assumptions — can make an easy problem impossible. Skilled solvers often spend disproportionate effort here, restating the problem, drawing diagrams, or translating it into a clearer form before attempting any solution.

Defining Goals and Subgoals

Once the problem is represented, the solver clarifies the goal and frequently breaks it into smaller subgoals. Decomposing a large, daunting goal into intermediate targets reduces the search and provides feedback along the way. Planning a research paper, for example, becomes more tractable when split into finding sources, outlining, drafting, and revising.

Exploring Possible Strategies

Next, the solver generates candidate approaches. This may involve recalling a procedure that worked on similar problems, reasoning by analogy, brainstorming options, or selecting a general heuristic such as working backward. Generating multiple options before committing tends to produce better outcomes than seizing on the first idea.

Acting and Monitoring

The chosen strategy is then carried out, with the solver monitoring progress against the goal. This monitoring is a metacognitive activity — thinking about one's own thinking — and it is what allows a person to notice a dead end early and change course rather than persisting fruitlessly.

Evaluating the Outcome

Finally, the solver checks whether the goal has been reached and whether the solution is satisfactory. Effective problem solvers review their answers, look for errors, and consider whether a better solution exists. This reflective step also builds the knowledge base that makes future problems easier, linking problem solving tightly to the psychology of learning.

4. Strategies: Algorithms and Heuristics

The actual machinery of solving relies on two broad classes of strategy: algorithms and heuristics. Understanding the trade-off between them is central to the psychology of problem solving.

Algorithms

An algorithm is a precise, step-by-step procedure that is guaranteed to produce a correct solution if it is followed exactly and the problem fits its scope. Long division, a recipe, or the standard method for solving a quadratic equation are algorithms. Their virtue is reliability; their drawback is cost. For problems with an enormous number of possible states, an exhaustive algorithm may be far too slow to use. Checking every possible chess game, for instance, is computationally impossible, so even powerful systems must rely on shortcuts.

Heuristics

A heuristic is a rule of thumb that reduces the search space and usually leads toward a good solution quickly, without guaranteeing success. Several heuristics recur throughout the research literature:

• Means-ends analysis: repeatedly identify the biggest difference between the current state and the goal, then choose an operator that reduces it. This was central to Newell and Simon's models.
• Working backward: start at the goal and reason back toward the initial state, useful when the endpoint is clearer than the start, as in many geometry proofs and maze problems.
• Subgoaling: break a hard problem into easier intermediate problems.
• Analogical reasoning: map a new problem onto a familiar one whose solution you already know.
• Trial and error: systematically try options when no better strategy is available.

Heuristics are powerful because human attention and working memory are limited; they let us act effectively under bounded conditions. But the same shortcuts that make us efficient can produce predictable errors. The availability and representativeness heuristics, for example, often serve us well yet underlie many documented cognitive biases when applied in the wrong context.

5. Insight and Restructuring

Some problems are solved not by gradual search but by a sudden reorganization of how the problem is seen — the experience commonly called an "aha" moment. Insight problems are designed so that the obvious interpretation leads nowhere, and progress requires abandoning a misleading assumption. The classic nine-dot problem, which asks you to connect nine dots with four straight lines without lifting the pen, is unsolvable until you allow the lines to extend beyond the imagined square boundary — the literal origin of "thinking outside the box."

Research on insight suggests several features that distinguish it from analytic problem solving. Solvers often experience an impasse, a feeling of being completely stuck, followed by a sudden and confident solution. They are frequently poor at predicting how close they are to solving an insight problem, whereas they can track progress on step-by-step problems fairly well. Studies have linked the moment of insight to a brief burst of activity in the right hemisphere and to a relaxation of mental constraints that had been blocking the answer. Periods of incubation — setting a problem aside and returning later — sometimes improve the chance of insight, possibly because the misleading initial approach fades and the mind can restructure the problem afresh.

6. Barriers to Problem Solving

Much of what makes problem solving difficult comes from the solver, not the problem. Several well-documented obstacles trip people up repeatedly.

Mental Set

A mental set is the tendency to approach a new problem with a strategy that worked before, even when a simpler or more appropriate method is available. Abraham Luchins' water-jar experiments demonstrated this vividly: participants who learned a complex multi-step solution kept using it on later problems that could be solved in one easy step. Experience is usually an asset, but a rigid set can blind a skilled person to an obvious shortcut.

Functional Fixedness

A specific form of mental set, functional fixedness, is the tendency to see an object only in terms of its usual function. In Duncker's candle problem, people struggle to mount a candle on a wall using a box of tacks because they perceive the box as a container for tacks rather than as a potential shelf. Overcoming functional fixedness means reconceiving an object's possible uses, a hallmark of creative thought.

Faulty Representation and Unnecessary Constraints

Solvers often impose constraints that the problem never specified — like assuming the nine-dot lines must stay within the square. Mis-framing the problem at the representation stage cascades into wasted effort. Restating a problem in different words, or from another person's perspective, can dissolve self-imposed limits.

Confirmation Bias and Premature Closure

People tend to seek evidence that supports a favored hypothesis and to overlook evidence against it, a pattern known as confirmation bias. In problem solving this shows up as premature commitment to the first plausible idea and reluctance to test alternatives. Anchoring on an initial estimate, an effect explored under anchoring bias, can similarly distort how a problem is approached.

Emotional and Cognitive Load

Finally, performance suffers under stress, anxiety, fatigue, and time pressure, all of which consume the limited resources of working memory. High anxiety in particular can narrow attention and crowd out the flexible thinking that hard problems require, which is one reason that managing arousal supports clear reasoning.

7. Expertise and Knowledge

One of the most robust findings in the field is that problem-solving skill is largely domain-specific. Experts are not simply people with better general reasoning; they have organized knowledge that transforms how they see problems in their field. Classic studies of chess players by Adriaan de Groot and later by William Chase and Herbert Simon found that masters were not dramatically better at calculating moves than weaker players. Instead, they recognized familiar configurations of pieces — "chunks" — that let them perceive the board in meaningful patterns and focus search on promising moves.

This pattern recurs across domains. Expert physicists categorize problems by deep principles such as conservation of energy, while novices sort the same problems by surface features such as whether they involve a spring or an inclined plane. Because experts encode problems at a deeper level, they retrieve relevant strategies faster and waste less effort on irrelevant paths. The practical lesson is sobering: general problem-solving tricks help, but there is no substitute for rich, well-structured knowledge of the area in which you are working. Building that knowledge is itself a long process of practice and feedback, closely tied to how learners believe their abilities can grow.

8. Why It Matters: Applications

Education

Schools increasingly aim to teach problem solving rather than rote procedures, recognizing that students must transfer skills to novel situations. Effective instruction uses worked examples, gradually fades support, and deliberately varies practice problems so that learners abstract the underlying structure rather than memorizing one surface pattern. Teaching for transfer also means cultivating metacognitive habits and connecting problem solving to broader critical thinking.

Workplace and Organizations

Employers consistently rank problem solving among the most valued professional skills. In organizational settings, structured methods — defining the problem, gathering data, generating options, and evaluating them — help teams avoid leaping to solutions before they understand the issue. Groups can outperform individuals when diverse perspectives surface more candidate solutions, though they are also vulnerable to conformity and shared blind spots.

Therapy and Mental Health

Problem solving is a clinical tool as well as a cognitive one. Problem-solving therapy, an evidence-based intervention often used for depression, teaches people a structured method for tackling life difficulties: defining a problem clearly, brainstorming options, weighing pros and cons, and implementing and reviewing a plan. By restoring a sense of agency over manageable steps, it can reduce the helplessness that accompanies many mental health conditions. Related skills appear throughout coping strategies and emotional intelligence work.

Artificial Intelligence

The information-processing theory of problem solving grew up alongside artificial intelligence, and the two fields still inform each other. Search algorithms, heuristics, and problem-space representations originated partly in attempts to model human cognition. Studying how machines solve problems, in turn, sharpens our understanding of what human solvers do differently — particularly our flexible use of analogy and our capacity for genuine restructuring.

9. How to Improve Problem Solving

Problem-solving ability is not fixed. While general intelligence sets some bounds, the strategies and habits that distinguish good solvers can be cultivated. The following evidence-informed practices help.

• Invest in representation. Before rushing to a solution, restate the problem in your own words, draw it, list what you know and what you want, and check for hidden assumptions. Most failures trace back to a poor initial representation.
• Build deep domain knowledge. Because expertise is domain-specific, the surest route to solving harder problems in an area is sustained, well-structured practice in that area.
• Learn general strategies explicitly. Practicing means-ends analysis, working backward, and subgoaling makes these tools available when you need them.
• Practice with varied examples. Working many different surface forms of the same underlying problem promotes transfer far better than repeating one type.
• Cultivate metacognition. Periodically ask: Is this approach working? What assumption am I making? Is there another way to see this? Monitoring your own process is one of the strongest predictors of success.
• Use incubation. When stuck on an insight problem, deliberately step away. A break can dissolve a misleading mental set and let a fresh structure emerge.
• Generate alternatives before committing. Forcing yourself to produce several options counters premature closure and confirmation bias.
• Manage arousal. Because stress consumes working memory, techniques that lower anxiety can free the mental resources that demanding problems require.

None of these practices turns a novice into an expert overnight. But together they describe the difference between someone who flails at obstacles and someone who approaches them systematically, flexibly, and with the patience to understand a problem before trying to solve it.

10. Frequently Asked Questions

What is problem solving in psychology?

In psychology, problem solving is the cognitive process of moving from an initial state to a desired goal state when the path is not immediately obvious. It involves representing the problem, generating possible solution paths, and evaluating outcomes, drawing on memory, reasoning, attention, and motivation. It is distinguished from routine behavior, which relies on a known, automatic response.

What is the difference between an algorithm and a heuristic?

An algorithm is a step-by-step procedure that guarantees a correct solution if followed precisely, but it can be slow or impractical for large problems. A heuristic is a mental shortcut or rule of thumb that is faster and often effective but does not guarantee success and can introduce systematic errors when applied in the wrong situation.

What are common barriers to effective problem solving?

Common barriers include mental set, where an old approach is applied even when it no longer fits; functional fixedness, the tendency to see objects only in their usual use; confirmation bias and premature closure; and a faulty initial representation of the problem. Anxiety, fatigue, time pressure, and limited working memory can further impair performance.

Can problem-solving skills be improved?

Yes. Skills improve with deliberate practice, building domain knowledge, learning general strategies such as working backward and means-ends analysis, and developing metacognitive habits like restating the problem, checking assumptions, generating alternatives, and reviewing solutions. Transfer is strongest when learners practice across many varied examples.

What is insight in problem solving?

Insight is the sudden emergence of a solution after a period of impasse, often experienced as an "aha" moment. It is associated with restructuring how the problem is represented rather than with gradual, step-by-step search. Gestalt psychologists studied it extensively, and stepping away from a problem to allow incubation can sometimes trigger it.