Learning Transfer: A Comprehensive Literature Review of Theoretical Foundations and Practical Interventions

May 26, 2026

The question of whether and how learning transfers across contexts has occupied educational researchers for over a century, producing one of psychology’s most contentious debates. Transfer—defined as the application of knowledge or skills learned in one context to new situations—represents the fundamental justification for formal education and workplace training. This review synthesizes theoretical foundations, empirical evidence, and practical interventions spanning from Thorndike’s seminal experiments to contemporary meta-analyses, revealing a field that has substantially advanced our understanding while leaving critical questions unresolved.

PART I: THEORETICAL FOUNDATIONS

Historical roots: Thorndike versus Judd

Modern transfer research began with Edward Thorndike and Robert Woodworth’s (1901) systematic challenge to the prevailing “formal discipline” doctrine—the belief that studying rigorous subjects like Latin and geometry strengthened general mental faculties applicable across all domains. Through careful experiments on estimating areas, lengths, and weights, Thorndike established his identical elements theory, arguing that transfer occurs “only where identical elements are concerned in the influencing and influenced function” (Thorndike & Woodworth, 1901, p. 250). Training on rectangular areas improved performance on similar rectangles but showed markedly less improvement for circles and triangles, suggesting the mind functions as “a machine for making particular reactions to particular situations.”

Charles Hubbard Judd’s (1908) generalization theory offered an alternative perspective emphasizing abstract principles over surface similarity. In his famous underwater target experiment, students who received instruction on light refraction adapted significantly faster when target depth changed compared to students who practiced without theoretical understanding. Judd argued that transfer occurs not merely through shared stimulus-response connections but through generalization of underlying principles—a position that presaged cognitive approaches emphasizing mental representations over behavioral associations.

These competing frameworks established a tension that persists today: does transfer require similarity of specific elements, or can principled understanding enable flexible adaptation to novel contexts? Singley and Anderson (1989) offered a compelling synthesis in The Transfer of Cognitive Skill, reviving Thorndike’s theory within modern cognitive architecture. They reconceptualized “elements” as production rules within Anderson’s ACT* framework, demonstrating that transfer is predictable based on shared procedural knowledge while remaining “abstract and mentalistic.” This formalization has been cited over 1,200 times and provided foundations for subsequent theorizing about cognitive skill transfer.

Near and far transfer: A multidimensional distinction

The near-far transfer distinction emerged gradually from classical debates, but Barnett and Ceci (2002) provided the most systematic taxonomy in their influential Psychological Bulletin article. Recognizing that “defining the terms near and far is no simple matter,” they developed a nine-dimensional framework crossing content dimensions (learned skill specificity, performance change, memory demands) with context dimensions (knowledge domain, physical context, temporal context, functional context, social context, modality).

This taxonomy revealed that apparent contradictions in transfer research often stemmed from comparing studies that differed on multiple dimensions simultaneously. A study demonstrating transfer between physics and chemistry represents near transfer on the knowledge domain dimension but may be far on temporal or functional dimensions. The practical implication is significant: estimating a single effect size for “far transfer” is misguided given this complexity. Transfer success depends on specific dimensional combinations, and no study has tested transfer that is simultaneously “far” on all dimensions.

Key theoretical frameworks explaining transfer mechanisms

Perkins and Salomon’s dual-pathway model distinguishes two fundamentally different transfer mechanisms. Low-road transfer occurs automatically when stimulus conditions resemble the learning context, triggered by well-developed responses from extensive practice. High-road transfer requires deliberate “mindful abstraction,” involving conscious search for connections and explicit principle extraction—what they termed “forward-reaching” or “backward-reaching” transfer depending on whether learners anticipate future applications or search memory for relevant prior learning (Perkins & Salomon, 1988; Salomon & Perkins, 1989).

Their instructional recommendations follow directly: “hugging” (making learning experiences closely resemble target contexts) promotes low-road transfer through automaticity, while “bridging” (encouraging explicit abstraction) promotes high-road transfer through principled understanding. This framework elegantly reconciles Thorndike’s emphasis on similarity with Judd’s emphasis on generalization by recognizing both as valid mechanisms operating under different conditions.

Situated cognition represents a fundamental challenge to cognitive transfer theories. Jean Lave’s (1988) Cognition in Practice and subsequent work with Wenger (1991) on legitimate peripheral participation argued that knowledge is not abstract and portable but inherently bound to the activity, context, and culture in which it develops. Lave’s grocery shopping studies demonstrated that adults who successfully performed comparison shopping mathematics in stores performed poorly on structurally identical formal test problems, suggesting that “school knowledge” may be fundamentally inert—learnable but not usable outside academic contexts. Brown, Collins, and Duguid (1989) extended this critique, arguing that the gap between knowing and doing prevents learners from appropriately using knowledge.

This position sparked significant debate. Anderson, Reder, and Simon (1996) defended cognitive approaches, arguing that context-dependency reflects learning conditions rather than inherent knowledge properties, and that substantial evidence supports successful transfer between tasks. Contemporary researchers increasingly recognize that both individual cognitive processes and social-contextual factors matter, with the nature of knowledge varying by domain and learning conditions.

Cognitive load theory (Sweller, 1988, 1994) explains transfer through schema acquisition and automation. Working memory’s severe limitations (2-4 elements for active processing) mean that novel information must be processed carefully, but automated schemas retrieved from long-term memory impose no such constraints. Transfer success depends on whether learners have constructed appropriate schemas that can be triggered by new situations. The variable practice effect—whereby practice with high variability produces better transfer than low variability practice—reflects the importance of building robust schemas through varied exemplars.

Analogical reasoning research by Gick and Holyoak (1980, 1983) and Gentner (1983) revealed both the power and limitations of analogy for transfer. Using Duncker’s radiation problem paradigm, Gick and Holyoak found that only 10% of participants solved the problem spontaneously, rising to 30% after reading an analogous military story and 75% when explicitly told to use the prior story. The critical finding is that the main obstacle to analogical transfer is noticing and retrieval, not mapping or application—people can use analogies effectively when prompted but rarely spontaneously recognize relevant prior knowledge.

Gentner’s structure-mapping theory explains that analogies map relational structure rather than object attributes, with higher-order relations preferentially mapped according to the systematicity principle. This explains why analogies can bridge remarkably dissimilar domains (arteries and electrical networks, chess strategies and politics) when structural similarity exists despite surface dissimilarity.

Contemporary perspectives and reconceptualizations

The most significant reconceptualization comes from Schwartz, Bransford, and Sears (2005), who reframed transfer as preparation for future learning (PFL) rather than direct application. They criticized traditional “sequestered problem solving” paradigms for systematically underestimating transfer by making people “look dumb rather than look smart” (Norman, 1993). Their distinction between “transfer in” (bringing prior experiences to new learning) and “transfer out” (applying acquired knowledge) highlighted the field’s overemphasis on the latter.

Empirical support came from “double transfer” designs demonstrating that students who engaged in invention activities performed poorly on immediate tests but dramatically outperformed tell-and-practice students when given a learning resource during assessment. The invention group was twice as likely to transfer successfully when given opportunity to learn during testing, revealing “hidden” transfer that traditional paradigms missed.

Transfer-appropriate processing (Morris, Bransford, & Franks, 1977) established that memory depends on the match between encoding and retrieval processes. Semantic encoding was superior for standard recognition tests, but rhyme encoding was superior for rhyming recognition tests—contradicting the claim that “deeper” processing always produces stronger memory. This principle explains why contextual similarity matters for transfer: learning encoded in one mode may not transfer to retrieval situations requiring different processing.

Lobato’s (2003, 2012) actor-oriented transfer shifts from expert-centric to learner-centric analysis, examining “the personal construction of relations of similarity across activities” regardless of whether experts judge transfer as “correct.” This perspective reveals transfer that traditional paradigms conceal and informs instructional design by attending to what students actually perceive as similar across situations.

The transfer problem: Critiques and debates

Douglas Detterman (1993) articulated the strongest skeptical position, arguing that “transfer has been one of the most actively studied phenomena in psychology… reviewers are in almost total agreement that little transfer occurs.” He contended that what appears to be transfer often reflects simply following directions rather than spontaneous knowledge application, with significant transfer “probably rare and accounts for very little human behavior.”

Sala and Gobet’s (2019) second-order meta-analysis (meta-analysis of meta-analyses) provided systematic evidence that far transfer does not occur across cognitive training programs. While near transfer showed consistent positive effects (g ≈ 0.25-0.46), they found no significant far transfer effects across working memory training, music training, chess training, or brain training programs—concluding that “the lack of training-induced far transfer is an invariant of human cognition.”

Responses to transfer skeptics take several forms. Bransford and Schwartz argue that transfer appears rare because traditional measures miss interpretive knowledge and preparation for future learning. Barnett and Ceci contend that apparent contradictions reflect comparing different dimensional combinations rather than genuine disagreement. Lobato suggests that transfer skepticism stems from researcher-centric definitions of “correct” transfer that miss what learners actually do. And methodological critics note that many “failed transfer” studies had inadequate original learning—transfer requires sufficient initial learning as a prerequisite (Klahr & Carver, 1988).

PART II: PRACTICAL INTERVENTIONS

Evidence-based instructional strategies

Bjork and Bjork’s (2011, 2020) concept of “desirable difficulties” provides the foundational framework for understanding instructional strategies that promote transfer. These are conditions that make learning more challenging during acquisition but enhance long-term retention and transfer. The key insight is that current performance is an unreliable index of learning—strategies that slow initial acquisition often optimize durable, generalizable knowledge.

Spacing refers to distributing learning over time rather than massing it in immediate succession. Cepeda et al.’s (2006) review of 400+ studies found robust benefits across all materials, learner types, and time scales. Critically, spacing promotes not just retention but generalization and transfer: Kornell and Bjork (2008) found that spacing enhanced generalization of artist painting styles to novel stimuli. The optimal gap increases with desired retention interval—10-20% of retention interval for short intervals, 5-10% for longer intervals—suggesting “many educational practices are highly inefficient” (Cepeda et al., 2008).

Interleaving involves alternating between different topics during practice rather than completing all examples of one type before moving to another. Brunmair and Richter’s (2019) meta-analysis found an overall effect of Hedges’ g = 0.42, with particularly strong effects for perceptual categories like paintings (g = 0.67). However, interleaving benefits depend critically on category similarity: greater benefits when categories are similar and confusable, while blocking may outperform interleaving for rule-based learning with low within-category similarity (Carvalho & Goldstone, 2014).

The discriminative contrast hypothesis explains that interleaving highlights differences between categories, facilitating discrimination, while also introducing natural spacing between same-category items and forcing memory reloading that strengthens representations. Yu, Decker, and Bhatt (2021) demonstrated that interleaving benefits both retention and transfer to novel sequences, with blocked training leaving learners “unprepared” for varied testing conditions.

Variability of practice creates more robust motor schemata supporting transfer to novel movement parameters (Schmidt, 1975). In Kerr and Booth’s (1978) classic study, children who practiced throwing to varied distances outperformed those who practiced at the specific tested distance, demonstrating that variability’s benefits outweigh any cost of non-specific practice. This principle extends to concept learning: using examples that vary on surface features while maintaining deep structure produces better transfer than constant exemplars.

Metacognition and self-regulation

Metacognitive skills—planning, monitoring, and evaluating one’s own learning—are considered domain-general and thus particularly important for transfer. The Education Endowment Foundation rates metacognition and self-regulated learning as having “high” security evidence based on 355 studies, with effect sizes equivalent to approximately +7 months’ academic progress.

Schuster et al. (2020) experimentally demonstrated that hybrid training combining metacognitive skills with cognitive strategies supported both near and far transfer in fifth-grade students. The key finding is that metacognitive strategies, as “higher-order strategies applicable in every domain,” transfer more readily than domain-specific cognitive strategies—but spontaneous transfer rarely occurs without explicit training.

Worked examples, analogical comparison, and invention

Worked examples reduce extraneous cognitive load while highlighting correct solution paths, with Chi et al. (1989) demonstrating that “good” students who self-explained achieved better transfer than those who did not. Self-explanation promotes acquisition of schema-based knowledge necessary for flexible application.

Analogical comparison facilitates transfer by inviting structural alignment and yielding common representations less context-specific than initial encodings. Gick and Holyoak (1983) found that comparing two source analogues dramatically improved transfer by facilitating schema abstraction—supporting the principle that comparison rather than isolated study promotes generalizable learning.

Schwartz and Martin’s (2004) research on contrasting cases and invention activities demonstrated that inventing with contrasting cases led to superior transfer to semantically unrelated topics (d = 0.33-0.66). Critically, being told procedures before problem-solving “can inadvertently undermine the learning of deep structures.” Invention activities serve as preparation for future learning, readying students to appreciate expert solutions by first discovering deep structure independently.

Learning environment design

Context-dependent learning presents a paradox for transfer: environmental specificity is advantageous for initial learning but disadvantageous for application across contexts. Smith, Glenberg, and Bjork (1978) demonstrated that studying material in two different rooms led to better recall than twice in the same room, indicating that varying contexts prevents learning from becoming “welded” to specific environmental cues. The practical recommendation is clear: to promote transfer, instruction should incorporate multiple contexts rather than constant, predictable conditions.

PART III: WORKPLACE LEARNING TRANSFER

Baldwin and Ford’s foundational model

Baldwin and Ford’s (1988) model remains “the earliest, and arguably most influential, conceptual framework dealing with training transfer” (Brown & Sitzmann, 2011). Synthesizing findings from 70+ empirical studies, they organized transfer around three training input categories:

Trainee characteristics: ability, motivation, personality, self-efficacy
Training design: learning principles, conditions of practice, content relevance, feedback
Work environment: transfer climate, supervisory support, peer support, opportunity to perform

The model specifies that training design primarily influences transfer indirectly through learning and retention, while trainee characteristics and work environment directly affect both learning outcomes and transfer. This interactionist approach—recognizing both individual and situational factors—has guided workplace transfer research for nearly four decades.

Holton’s Learning Transfer System Inventory

Holton, Bates, and Ruona (2000) developed the Learning Transfer System Inventory (LTSI), a comprehensive measurement instrument operationalizing transfer climate through 16 factors organized into ability (learner readiness, content validity, transfer design, personal capacity, opportunity to use), motivation (motivation to transfer, expectancy beliefs), and work environment domains (supervisor support, peer support, resistance to change, organizational sanctions and rewards).

The LTSI has been validated across 17 countries and 14 languages (N = 6,120 in Version 3), with confirmatory factor analysis supporting an 11-factor program-specific and 5-factor training-general domain structure. This cross-cultural validation provides practitioners with a reliable diagnostic tool for identifying transfer barriers in organizational contexts.

Transfer climate and organizational factors

Transfer climate—trainees’ perceptions of support they will receive and consequences for skill application—emerges as one of the most consistent predictors of workplace transfer. Blume et al.’s (2010) meta-analysis of 89 empirical studies (N = 12,496) found:

Supervisor support: ρ = .31 (strongest work environment predictor)
Peer support: ρ = .14
Transfer climate (overall): ρ = .27 (corrected for same-source bias)
Cognitive ability: ρ = .37
Conscientiousness: ρ = .28
Motivation to transfer: ρ = .23

Critically, the meta-analysis identified same-source/same-measurement-context bias as consistently inflating transfer relationships by ρ = .20-.30, suggesting that much prior research overestimated predictor-transfer associations. Effect sizes also varied dramatically by skill type: work environment predicted transfer for open skills (leadership, interpersonal; ρ = .26) far more strongly than closed skills (technical; ρ = .04), indicating that soft skills training requires stronger environmental support.

Martin (2010) demonstrated that peer support can mitigate negative transfer climates—trainees with peer support in negative climates achieved the same degree of transfer as trainees in positive climates. This finding has significant practical implications: cultivating peer networks may partially compensate for inadequate supervisor or organizational support.

Recent developments in workplace transfer research

Ford, Baldwin, and Prasad’s (2018) Annual Review synthesized reliable findings while identifying significant gaps. They emphasized transfer as a dynamic process unfolding over time rather than a static outcome, distinguishing between “maximum transfer” (what trainees can do) and “typical transfer” (what they will do). Huang et al. (2017) demonstrated that transfer behavior varies within individuals over time, challenging traditional cross-sectional approaches.

The Dynamic Transfer Model (Blume et al., 2019) reconceptualizes transfer as iterative: transfer intentions lead to initial attempts, which generate feedback that is integrated into subsequent attempts. This formulation calls for longitudinal research designs with multiple measurement points to capture transfer trajectories.

Research also increasingly recognizes that only 10-30% of training content transfers to the workplace (Ford et al., 2018; Saks & Belcourt, 2006), that no “magic bullet” exists (multiple factors with moderate effects must be leveraged in combination), and that organizational culture may be the “missing link” in training transfer—a factor proposed but significantly under-researched.

PART IV: ASSESSMENT, META-ANALYTIC EVIDENCE, AND GAPS

Meta-analytic findings on transfer interventions

Pan and Rickard’s (2018) meta-analysis on test-enhanced learning found an overall transfer effect of d = 0.40 (95% CI [0.31, 0.50]) relative to non-testing controls. Transfer was strongest for application and inference questions, medical diagnosis problems, and mediator word cues, while weakest for rearranged stimulus-response items and worked example problems. Response congruency and elaborated retrieval practice emerged as key moderators.

Pai, Sears, and Maeda’s (2015) meta-analysis found small-group learning produced d = 0.30 for transfer performance, with high variability across effect sizes indicating the need for investigating specific group features. Brunmair and Richter (2019) found interleaving produced g = 0.42 overall, with mathematical tasks showing g = 0.34 and visual categorization showing effects up to g = 0.67.

However, Sala et al.’s (2019) second-order meta-analysis represents the most comprehensive evidence on far transfer limitations. Across all cognitive training programs examined—working memory training, music training, chess training, brain training—no significant far transfer effects emerged. Near transfer effects were consistent (g ≈ 0.25-0.46), with typically developing children showing greatest benefits, but the lack of far transfer “appears to be an invariant of human cognition.”

Assessment and measurement challenges

The “criterion problem” (Ford et al., 2018) refers to inconsistent measurement operationalizations, over-reliance on same-source/same-time assessments that inflate correlations, limited longitudinal designs, and focus on short-term outcomes. Gegenfurtner’s meta-analysis found that self-rated transfer produced upwardly biased estimates compared to supervisor or peer ratings. Only 35% of organizations have formal processes to measure transfer (Institute for Corporate Productivity), suggesting the field’s sophisticated theoretical models rarely translate to practice.

Kirkpatrick’s (1959/2006) four-level model (Reaction, Learning, Behavior, Results) remains widely used despite limited empirical support for the assumed causal chain. Thalheimer’s (2018) Learning Transfer Evaluation Model offers eight levels with increasing complexity, while the Success Case Method (Brinkerhoff) provides qualitative approaches comparing successful and unsuccessful transfer cases.

Literature gaps and future directions

Multiple systematic reviews identify consistent gaps requiring attention:

Longitudinal research: Most studies use cross-sectional designs unable to capture transfer dynamics over time
Far transfer mechanisms: Why far transfer rarely occurs remains theoretically unexplained
Boundary conditions: When do different transfer-enhancing strategies work or fail, and for whom?
Classroom-based studies: Laboratory findings need validation in authentic educational settings
Strategy combinations: Sequential effects of combining retrieval practice with generative activities remain unstudied
Organizational culture: Proposed as critical but significantly under-researched
Technology-mediated training: Gaps in understanding online and blended learning transfer
Cross-cultural validation: Limited research outside Western contexts

The ongoing debate about far transfer’s existence versus its measurement reflects deeper theoretical disagreements. Skeptics interpret null findings as evidence that far transfer is cognitively impossible; optimists argue that inadequate instructional designs and insensitive measures obscure real but fragile effects. Resolving this debate will require both theoretical advances specifying precise conditions under which far transfer should occur and methodological innovations capable of detecting subtle transfer effects.

Conclusion: What we know and what remains uncertain

A century of research yields several robust conclusions. Near transfer is achievable and occurs reliably when learning and application contexts share substantial similarity. Specific instructional strategies—spacing, interleaving, variability, retrieval practice, analogical comparison, invention activities—enhance transfer with effect sizes ranging from d = 0.30 to 0.67 under appropriate conditions. Work environment factors, particularly supervisor support and transfer climate, predict workplace training transfer with meaningful effect sizes (ρ = .27-.31). Schema construction through varied practice and explicit abstraction represents a convergent mechanism across multiple theoretical frameworks.

Far transfer remains contested. The most rigorous meta-analytic evidence suggests it does not occur across cognitive training programs, potentially representing a fundamental limitation of human cognition. Yet alternative perspectives argue that traditional paradigms systematically miss transfer by using sequestered problem-solving assessments that underestimate preparation for future learning. The resolution likely lies in recognizing that “far transfer” is not a unitary phenomenon but a complex multidimensional space where some dimensional combinations may support transfer while others do not.

For practitioners, the evidence suggests focusing on near transfer as the realistic goal while incorporating strategies (bridging, analogical comparison, varied contexts) that may extend transfer distance modestly. The design of transfer-supportive work environments—through supervisor involvement, peer networks, and opportunity to practice—emerges as equally important as training design itself. And the sobering estimate that only 10-30% of training transfers to workplace application should prompt systematic attention to post-training support rather than assuming that good training automatically produces lasting behavior change.

The field has progressed substantially from Thorndike’s identical elements through contemporary dynamic models, yet the fundamental tension between specificity and generality, between context-bound and transferable knowledge, remains unresolved. Future progress requires integrating educational and organizational psychology traditions that have developed largely in parallel, developing validated transfer outcome measures that distinguish maximum from typical performance, and conducting longitudinal research capturing how transfer unfolds over time rather than treating it as a static outcome. The stakes remain high: transfer represents the fundamental justification for educational investment, and understanding its mechanisms, limits, and facilitation remains among psychology’s most consequential challenges.

References

Anderson, J. R., Reder, L. M., & Simon, H. A. (1996). Situated learning and education. Educational Researcher, 25(4), 5-11.

Baldwin, T. T., & Ford, J. K. (1988). Transfer of training: A review and directions for future research. Personnel Psychology, 41(1), 63-105.

Baldwin, T. T., Ford, J. K., & Blume, B. D. (2017). The state of transfer of training research: Moving toward more consumer-centric inquiry. Human Resource Development Quarterly, 28(1), 17-28.

Barnett, S. M., & Ceci, S. J. (2002). When and where do we apply what we learn? A taxonomy for far transfer. Psychological Bulletin, 128(4), 612-637.

Bates, R., Holton, E. F., III, & Hatala, J. P. (2012). A revised learning transfer system inventory: Factorial replication and validation. Human Resource Development International, 15(5), 549-569.

Bjork, R. A., & Bjork, E. L. (2011). Making things hard on yourself, but in a good way: Creating desirable difficulties to enhance learning. In M. A. Gernsbacher et al. (Eds.), Psychology and the real world (pp. 56-64). Worth Publishers.

Blume, B. D., Ford, J. K., Baldwin, T. T., & Huang, J. L. (2010). Transfer of training: A meta-analytic review. Journal of Management, 36(4), 1065-1105.

Blume, B. D., Ford, J. K., Surface, E. A., & Olenick, J. (2019). A dynamic model of training transfer. Human Resource Management Review, 29(2), 270-283.

Bransford, J. D., & Schwartz, D. L. (1999). Rethinking transfer: A simple proposal with multiple implications. Review of Research in Education, 24, 61-100.

Brown, J. S., Collins, A., & Duguid, P. (1989). Situated cognition and the culture of learning. Educational Researcher, 18(1), 32-42.

Brunmair, M., & Richter, T. (2019). Similarity matters: A meta-analysis of interleaved learning and its moderators. Psychological Bulletin, 145(11), 1029-1052.

Carvalho, P. F., & Goldstone, R. L. (2014). Effects of interleaved and blocked study on delayed test of category learning generalization. Frontiers in Psychology, 5, 936.

Cepeda, N. J., Pashler, H., Vul, E., Wixted, J. T., & Rohrer, D. (2006). Distributed practice in verbal recall tasks: A review and quantitative synthesis. Psychological Bulletin, 132(3), 354-380.

Chi, M. T. H., Bassok, M., Lewis, M. W., Reimann, P., & Glaser, R. (1989). Self-explanations: How students study and use examples in learning to solve problems. Cognitive Science, 13, 145-182.

Detterman, D. K. (1993). The case for the prosecution: Transfer as an epiphenomenon. In D. K. Detterman & R. J. Sternberg (Eds.), Transfer on trial: Intelligence, cognition, and instruction (pp. 1-24). Ablex Publishing.

Ford, J. K., Baldwin, T. T., & Prasad, J. (2018). Transfer of training: The known and the unknown. Annual Review of Organizational Psychology and Organizational Behavior, 5, 201-225.

Gentner, D. (1983). Structure-mapping: A theoretical framework for analogy. Cognitive Science, 7, 155-170.

Gick, M. L., & Holyoak, K. J. (1980). Analogical problem solving. Cognitive Psychology, 12, 306-355.

Gick, M. L., & Holyoak, K. J. (1983). Schema induction and analogical transfer. Cognitive Psychology, 15, 1-38.

Holton, E. F., III, Bates, R. A., & Ruona, W. E. A. (2000). Development of a generalized learning transfer system inventory. Human Resource Development Quarterly, 11(4), 333-360.

Judd, C. H. (1908). The relation of special training to general intelligence. Educational Review, 36, 28-42.

Kornell, N., & Bjork, R. A. (2008). Learning concepts and categories: Is spacing the “enemy of induction”? Psychological Science, 19, 585-592.

Lave, J. (1988). Cognition in practice: Mind, mathematics and culture in everyday life. Cambridge University Press.

Lave, J., & Wenger, E. (1991). Situated learning: Legitimate peripheral participation. Cambridge University Press.

Lobato, J. (2012). The actor-oriented transfer perspective and its contributions to educational research and practice. Educational Psychologist, 47(3), 232-247.

Morris, C. D., Bransford, J. D., & Franks, J. J. (1977). Levels of processing versus transfer appropriate processing. Journal of Verbal Learning and Verbal Behavior, 16(5), 519-533.

Pai, H. H., Sears, D. A., & Maeda, Y. (2015). Effects of small-group learning on transfer: A meta-analysis. Educational Psychology Review, 27(1), 79-102.

Pan, S. C., & Rickard, T. C. (2018). Transfer of test-enhanced learning: Meta-analytic review and synthesis. Psychological Bulletin, 144(7), 710-756.

Perkins, D. N., & Salomon, G. (1988). Teaching for transfer. Educational Leadership, 46(1), 22-32.

Salomon, G., & Perkins, D. N. (1989). Rocky roads to transfer: Rethinking mechanisms of a neglected phenomenon. Educational Psychologist, 24(2), 113-142.

Sala, G., Aksayli, N. D., Tatlidil, K. S., Tatsumi, T., Gondo, Y., & Gobet, F. (2019). Near and far transfer in cognitive training: A second-order meta-analysis. Collabra: Psychology, 5(1), 18.

Schmidt, R. A. (1975). A schema theory of discrete motor skill learning. Psychological Review, 82, 225-260.

Schuster, C., Stebner, F., Leutner, D., & Wirth, J. (2020). Transfer of metacognitive skills in self-regulated learning: An experimental training study. Metacognition and Learning, 16, 297-318.

Schwartz, D. L., Bransford, J. D., & Sears, D. (2005). Efficiency and innovation in transfer. In J. P. Mestre (Ed.), Transfer of learning from a modern multidisciplinary perspective (pp. 1-51). Information Age Publishing.

Schwartz, D. L., & Martin, T. (2004). Inventing to prepare for future learning: The hidden efficiency of encouraging original student production in statistics instruction. Cognition and Instruction, 22(2), 129-184.

Singley, M. K., & Anderson, J. R. (1989). The transfer of cognitive skill. Harvard University Press.

Sweller, J. (1988). Cognitive load during problem solving: Effects on learning. Cognitive Science, 12, 257-285.

Thorndike, E. L., & Woodworth, R. S. (1901). The influence of improvement in one mental function upon the efficiency of other functions (I). Psychological Review, 8(3), 247-261.