A student aces a physics exam on momentum calculations. Three weeks later, faced with a real-world problem about car collisions, they stare blankly—unable to connect classroom formulas to the scenario before them. This frustrating disconnect between knowing and applying represents one of education's most persistent challenges.

Transfer of learning—the ability to use knowledge acquired in one context to solve problems in another—sits at the heart of what education should accomplish. Yet research consistently shows that students struggle to bridge this gap. They can reproduce information but fail to recognise when and how to deploy it elsewhere.

The problem isn't that students haven't learned. It's that their knowledge remains locked to specific contexts, encoded alongside surface features that prevent flexible retrieval. Understanding why this happens—and what instructional approaches can overcome it—offers educators powerful tools for developing genuinely applicable expertise.

Memory research reveals a fundamental principle: retrieval works best when conditions match encoding. Tulving's encoding specificity principle explains why students who learned statistical concepts in a psychology classroom may fail to apply them in a business context—even when the underlying mathematics remains identical.

This context dependency extends beyond physical environment. The mental state, the format of problems, the language used, even the examples presented during learning become woven into the memory trace itself. When students encounter a problem that looks different, they often fail to access relevant knowledge—not because it's forgotten, but because the retrieval cues don't match.

Educational research documents this phenomenon repeatedly. Students who master biology concepts through textbook diagrams struggle with photographs of actual organisms. Mathematics learners who excel at word problems fail when the same relationships appear in spreadsheets. The knowledge exists but remains inaccessible.

This explains a common classroom puzzle: students who performed brilliantly on assessments seem incapable of basic application weeks later. The test conditions matched their encoding context closely. Real-world problems don't offer such convenient alignment. Knowledge that seems robust proves surprisingly fragile outside its original learning environment.

Takeaway

Knowledge becomes bound to the conditions under which it was learned. To promote transfer, deliberately vary the contexts, formats, and examples during instruction so retrieval pathways remain flexible rather than context-locked.

When experts encounter new problems, they quickly identify underlying structure—the deep principles that determine what type of problem it is and what solution approaches apply. Novices, by contrast, fixate on surface features: the cover story, the specific numbers, the domain-specific vocabulary.

Classic research by Chi and colleagues demonstrated this vividly. Physics experts categorised problems by underlying principles (conservation of energy, Newton's second law). Physics students categorised the same problems by surface features (problems about inclined planes, problems about springs). This surface-level encoding creates transfer barriers.

The implications for education are significant. When students learn through limited examples, they inadvertently encode surface features as essential. A student who only encounters percentage calculations in retail contexts may believe percentages are fundamentally about shopping. The deep mathematical structure gets obscured by consistent surface wrapping.

This phenomenon, called negative transfer, can actively interfere with new learning. Students apply familiar procedures to problems that look similar but require different approaches. The surface pattern triggers the wrong knowledge, and confident errors result. Breaking free from surface-level encoding requires deliberate instructional attention to structural features.

Takeaway

Novices tend to encode surface features while missing deep structure. Use varied examples that share underlying principles but differ in surface appearance, and explicitly teach students to identify structural similarities across different-looking problems.

Evidence-based strategies for promoting transfer share a common thread: they help learners abstract underlying principles from specific examples. Comparison stands as one of the most powerful techniques. When students examine multiple examples side by side, noting similarities and differences, structural features become visible while surface features recede.

Hattie's meta-analyses highlight the importance of making thinking visible. Teachers who explicitly discuss problem structure, model their reasoning process, and have students articulate how new problems connect to previous learning show stronger transfer outcomes. This metacognitive scaffolding helps students develop the pattern recognition experts display naturally.

Interleaved practice—mixing different problem types rather than blocking them—forces students to practice recognising what type of problem they're facing, not just how to execute procedures. Though initially slower and more effortful than blocked practice, interleaving builds the discrimination skills essential for transfer.

Perhaps most importantly, instruction should incorporate bridging: explicit discussion of where and when knowledge applies. Abstract principles need concrete connection to multiple contexts. Students benefit from generating their own examples, predicting applications, and reflecting on structural similarities. Transfer doesn't happen automatically—it must be taught.

Takeaway

Promote transfer through comparison of varied examples, explicit attention to underlying structure, interleaved practice that develops problem recognition, and bridging discussions that connect abstract principles to multiple concrete applications.

The gap between knowing and applying isn't a student failure—it's a predictable consequence of how memory works. Knowledge encoded in narrow contexts, focused on surface features, will remain stubbornly immobile regardless of how thoroughly it was originally learned.

Effective instruction acknowledges this reality and builds transfer-promoting practices into curriculum design. Varied examples, explicit structural comparison, interleaved practice, and bridging conversations don't slow down learning—they ensure learning actually becomes useful.

The goal of education isn't to produce students who perform well on assessments that mirror instruction. It's to develop flexible expertise that travels beyond the classroom. Designing for transfer makes that goal achievable.