Should you study the solutions or solve problems?

Main Ideas

• Effort alone does not guarantee learning. Trying to solve a difficult problem can consume mental resources without helping learners notice the principles or patterns needed to solve new problems.
• Worked examples can be especially effective for beginners. A worked example is a complete solution that a learner studies, often followed by a similar practice problem. In studies described in the video, learners who studied solved examples did as well as problem-solvers on familiar tasks and better on unfamiliar ones.
• Problem-solving can overload working memory. When people do not know how to solve a problem, they often use means–ends analysis: they start with the goal, identify what would achieve it, and work backward through subgoals. This can help them make progress, but keeping all those goals and subgoals in mind uses working memory that might otherwise support learning.
• Cognitive load theory distinguishes two kinds of demand:
  • Internal load is the mental effort needed to learn the material.
  • External load is effort imposed by the task or presentation that is not itself necessary for learning—for example, tracking many subgoals or dealing with irrelevant information.
  • If total demand exceeds working-memory capacity, important information may be lost. Effective instruction minimizes unnecessary load and directs effort toward learning.
• The best approach changes with expertise. Beginners often benefit more from clear explanations and worked examples because they have not yet built the knowledge needed to solve problems efficiently. As learners gain experience, they develop schemas—organized knowledge that helps them recognize familiar problem types and proceed more directly.
• This is known as the expertise-reversal effect. As expertise grows, worked examples become less useful, while independent problem-solving becomes more valuable. Practice and retrieval are important for mastery, but they are most effective when introduced at the appropriate stage.

Research Examples Discussed

• Sweller’s number problems: Students could solve problems by multiplying by three and subtracting 29, but most did not discover the underlying alternating pattern. They could reach answers without learning the general rule.
• Geometry problems without a specific goal: Students asked to find as many angles and side lengths as possible—rather than work toward one target—performed better on a new kind of geometry problem than students given conventional goal-directed problems. This approach reduces means–ends analysis, though it may not suit problems with very large search spaces.
• Algebra worked examples: Students who studied solved examples performed about as well as problem-solvers on similar tasks, and substantially better on a new problem type. The result suggests that examples can support understanding and transfer, not just memorization.

Strategies for Moving from Examples to Independent Problem-Solving

1. Self-explanation

   • Begin with a fully worked solution.
   • Explain why each step is taken and how it contributes to the solution.
   • This encourages deeper processing than simply reading through the answer.

2. Fading

   • Start with a complete worked example.
   • Gradually remove parts of the solution and ask the learner to fill them in.
   • The video notes that fading works best when the final steps are omitted first, with more of the solution removed over time.

3. Step-by-step recall

   • Reveal a solution one step at a time.
   • Before each step is shown, try to recall what should come next.
   • This provides retrieval practice without requiring a complete solution from scratch. The approach improved learning in the cited experiment, though it took learners longer.

Practical Takeaway

Start with clear instruction and worked examples when a topic is new. Add self-explanation, partial solutions, and stepwise recall to bridge toward independent work. As learners develop relevant knowledge, shift more time toward solving problems themselves.

Speakers and Sources Featured

• Scott Young — narrator and presenter.
• John Sweller — psychologist whose problem-solving and cognitive-load research is discussed.
• Herbert Simon and Allen Newell — researchers whose work on human problem-solving and heuristics is described.
• Alexander Renkl and colleagues — researchers cited for work on self-explanation.
• Gijs van den Broek and colleagues — researchers cited for step-by-step retrieval practice.