Learning Science · Learning in different knowledge and skill domains
Checking what transfers and what stays domain-specific
Test the boundaries of a learning gain instead of assuming that expertise spreads to every task.
Use the examples to connect this lesson with the earlier topics. No specialist equipment is needed.
Back to the chapterAfter reading, you can
- Identify which knowledge a new task shares with practice.
- Design a sequence of increasingly changed applications.
- Report domain-specific success without implying a general ability gain.
Ask what is shared
A musician reads notation fluently. That accomplishment depends on knowledge of symbols, timing and musical structure. It does not by itself predict fluent reading of chemical equations. Both tasks involve symbols, but the relationships and decisions differ.
Transfer becomes more plausible when useful knowledge, representations or strategies are shared and can be recognised in the new setting. Superficial resemblance is not enough. The new task may also require additional knowledge that was never taught.
Specify a boundary
Imagine practice on ratios in recipes. A nearby check changes ingredient names and quantities. A more demanding check asks the learner to choose whether a ratio applies. A still broader task involves interpreting scale in a map. These changes should be described individually rather than collapsed into a single transfer score.
A failure in the map task might reflect map conventions, units or representation rather than loss of ratio knowledge. Inspect the reasoning before deciding what the gain did or did not include.
| Check | Shared knowledge | Added demand |
|---|---|---|
| Double a familiar recipe | Multiplicative relationship | Familiar ingredient context |
| Choose whether scaling fits a new recipe task | Multiplicative relationship when appropriate | Select the model without a method cue |
| Interpret a 1:500 map scale | Corresponding length ratio | Map notation and consistent units |
Separate trained, related and broader outcomes
A training evaluation can include the exact task, a related untrained task and a more distant outcome chosen for a clear reason. Improvement on the trained task is meaningful but narrow. A broad outcome requires its own comparison and evidence.
If many possible outcomes are tested, preserve the full set and distinguish planned checks from exploratory additions. Selecting the one favourable result after the fact exaggerates the apparent generality of training.
Teach a bridge when it is needed
If learners do not notice a shared structure, compare the two situations explicitly. Name the common relationship and the features that do not carry over. Then test a fresh application without supplying every step.
This bridge is teaching, not proof that all transfer barriers vanish. Domain knowledge remains important. A useful conclusion can celebrate a specific capability while identifying the additional preparation required elsewhere.
Worked example
Recipe ratios and a map scale
Why might a learner who doubles a recipe still misunderstand a 1:500 map scale?
- A learner doubles a recipe correctly but treats a 1:500 map scale as if it described the total page area. The underlying ratio knowledge is not automatically enough to interpret the map notation.
- The tutor explains the scale as corresponding lengths, works through one example with units and asks for a different distance. The new evidence concerns that taught connection. It should not be described as a general increase in all quantitative ability.
The map adds a notation and corresponding-length convention. Teach that bridge explicitly, use units, and test a new distance; the result concerns the taught connection rather than all quantitative ability.
Try it yourself
Think through each question before opening the answer.
Why is improvement on a memory game insufficient to establish general reasoning improvement?
Show the answer and reasoning
The trained task and broad reasoning differ. The broader outcome needs an appropriate independent check and comparison.
Improvement may be specific to the game’s stimuli, rules or strategy. A separate broader-reasoning outcome and relevant comparison are needed before expanding the claim.
How can an apparent transfer failure guide teaching?
Show the answer and reasoning
Inspect which shared relationship was missed and which new knowledge was required, then teach and test a specific bridge.
A failure can localise the missing bridge: recognising the shared ratio, reading notation or supplying new domain knowledge. The diagnosis guides a specific taught connection and fresh check.
What to carry forward
- Describe shared knowledge and changed demands.
- Measure each claimed breadth of improvement.
- Teach bridges explicitly and keep conclusions bounded.
References
Sources supporting the explanations and learning checks in this lesson.