Transfer of learning refers to the degree to which knowledge or skill gained in one context carries over and applies to a new, different context. Educators care about transfer because the ultimate goal of instruction is usually not just performance on the immediate task or test, but the ability to apply what was learned to new problems and situations encountered later, whether in another class, a job, or everyday life.
Near transfer versus far transfer
Researchers distinguish between near transfer, where a skill is applied to a new situation that closely resembles the one in which it was learned, and far transfer, where a skill is applied to a situation that looks quite different on the surface even though the underlying principle is the same. Far transfer is generally much harder to achieve than near transfer, and a large body of research has found that students often fail to apply a strategy or concept learned in one class to a different subject, even when the underlying logic is identical.
Conditions that support transfer
Research on transfer suggests that it is more likely when instruction explicitly draws attention to the underlying principle behind a skill rather than leaving students to notice it on their own, and when students practice applying a concept across multiple varied contexts rather than repeating the same type of problem. Some researchers argue that the common complaint that students cannot apply classroom learning to real life reflects a transfer problem, suggesting that how material is taught matters as much as what material is taught for building skills students can actually use beyond the classroom where they first learned them.
Historical background
Early twentieth-century psychologist Edward Thorndike conducted some of the first experimental studies on transfer, famously concluding that transfer was quite limited and depended on genuinely identical elements between the original learning task and the new situation, a more pessimistic view than earlier educators had assumed. Later researchers have offered more nuanced views, arguing that transfer is possible but requires deliberate instructional design rather than happening automatically simply because a student once learned a related skill.
The concept has become especially important in discussions of twenty-first-century skills, such as critical thinking and problem-solving, since employers and educators often hope that skills developed in school will transfer to workplace settings quite different from the classroom, which research on far transfer suggests is harder to achieve than commonly assumed.
Low-road and high-road transfer
Psychologist Gavriel Salomon and colleague David Perkins proposed a distinction between low-road transfer, which happens automatically when a skill has been practiced so extensively that it becomes nearly reflexive in a new but similar situation, and high-road transfer, which requires a learner to deliberately and consciously recognize an abstract connection between what they already know and a new, dissimilar problem. This distinction has influenced how some teachers design instruction, encouraging explicit discussion of underlying principles and asking students to articulate connections across different problems rather than relying on skills to transfer on their own through repetition alone.
Transfer in assessment design
Some educators argue that the best way to measure whether real transfer has occurred is to test a skill in a context meaningfully different from how it was originally taught, rather than simply retesting the same type of problem used during instruction, since a student can appear to have mastered a skill on a near-identical retest while still being unable to apply it in a genuinely new situation. This insight has influenced some assessment designers to deliberately build novel, unfamiliar problem contexts into tests meant to measure durable understanding rather than short-term memorization of a specific procedure.
FAQs
Editorially reviewed by Specialty Digest Editorial TeamLast reviewed November 2, 2026Researched and drafted with AI assistanceReport an issue