From Study Finding to Learning Design
The practical problem is how to turn the evidence about solving first, worked examples, retrieval, variation, and feedback into one coherent learning sequence. The useful action is not to copy a laboratory condition or force every learner through the same order. It is to diagnose what the learner can currently represent, choose an entry point that makes the next cognitive job possible, and then use explanation, feedback, and transfer evidence to adjust the sequence.
A workable design has a changing center of gravity. Early on, the learner may need to notice a pattern or see how a procedure is organized. Later, they may need to retrieve that rule without help. Still later, they may need to recognize it in a problem that does not resemble the practice item. The order of activities should follow those changes rather than treating “problem solving,” “examples,” or “practice” as permanent identities.
Start with a Useful First Encounter
Begin by defining the transfer target. “Learn fractions,” “understand a scientific principle,” or “study a programming method” is too broad to determine the first activity. A useful target says what the learner should recognize, explain, or do when the surface details change. For the distributive property, the target might be recognizing that an outside factor applies to each term and using that relationship in a new expression. That target tells the designer what evidence to look for: not only whether a learner can repeat one transformation, but whether they can explain and apply the relation elsewhere.
Next, ask what representation the learner already has. If they have enough relevant knowledge to make a meaningful attempt, a designed problem can reveal their current strategy. The problem should make the target relationship relevant without requiring a long chain of completely unknown operations. The learner might be asked to compare several expressions or predict what happens when a factor is applied to a sum. An incomplete answer can then expose whether they distribute the factor to both terms, attach it to only one, or focus on an irrelevant feature.
That attempt is valuable only if later instruction can use it. The teacher, text, or learning resource should be prepared to compare the learner’s approach with the underlying principle. The purpose is not to collect wrong answers as proof of engagement. It is to make the difference between a partial strategy and a transferable rule visible. If the learner’s work contains no interpretable idea because the prerequisites are missing, a first attempt may add difficulty without adding useful information.
In that situation, begin with a worked example or another form of guidance. A worked example can show how the relevant features of a problem connect to the steps of a solution while reducing the burden of inventing the entire procedure. This is not a lower standard. It is a way to protect attention for organization when independent generation would consume it. The evidence supports a conditional choice: designed attempts can prepare later instruction, while worked examples may be the safer entry when induction demands exceed the learner’s available resources.
The distinction can be made concrete with the distributive property. A learner with some grasp of multiplication might first examine several expressions and propose what changes when the outside factor is applied. Another learner who cannot yet track the relationship may first study a worked transformation such as 2(x + 3) = 2x + 6, with the explanation tied to both terms. The first route exposes a hypothesis for discussion; the second makes the procedure inspectable.