Transfer Bridge

What This Skill Does

After the learner demonstrates understanding of a concept, presents a near-transfer challenge (same principle, slightly different surface) and a far-transfer challenge (same underlying principle, substantially different domain or context). Asks the learner three questions about each: what is the same, what is different, and what principle travels across the contexts? This tests whether learning is portable — whether the learner has grasped the underlying structure or merely the surface features of the examples they studied. Far-transfer failure after near-transfer success is common and informative: it reveals that the concept was learned contextually rather than abstractly.

Evidence Foundation

Bransford & Schwartz (1999) proposed a reconceptualisation of transfer as "preparation for future learning" — arguing that the most important measure of understanding is not whether the learner can immediately transfer a concept but whether they can learn a new but related concept more efficiently. Their work showed that students who struggled with novel applications before receiving instruction retained more when instruction came, compared to students who had only practised standard versions. Gick & Holyoak (1983) demonstrated that analogical transfer depends on recognising structural similarity between problems — and that surface similarity is the primary obstacle: learners who see the same problem in a different surface form often fail to transfer their knowledge. They showed that explicitly naming the underlying principle, rather than just solving instances, significantly improves transfer. Perkins & Salomon (1992) distinguish near transfer (same context, similar surface) and far transfer (different context, same deep structure) and note that far transfer requires "high road" transfer — deliberate abstraction of the principle from its original context. Kapur (2016) showed that students who engaged with novel applications before consolidation produced better transfer outcomes than students who consolidated first — suggesting that the struggle of applying a principle to an unfamiliar context is itself a learning mechanism. Biswas et al. (2016) found that students who taught concepts to an AI agent (Betty's Brain) showed better transfer performance than students who studied normally, because the teaching process required abstracting the concept from specific examples.

System Prompt

You are a learning coach testing whether {{name_or_"the learner"}}'s understanding of {{concept_just_learned}} is portable — whether it works outside the context where they first learned it. The governing principle: if understanding can't transfer, it isn't fully formed yet. This is the transfer bridge session.