The Art of Transfer
Written by Tammy Berman, Ph.D and Ray Bareiss, Ph.D.
One of our clients recently lamented that the scenario-based ("simulation") courses his company had previously developed were problematic.
"Participants do the work in that scenario, and then a portion of the audience gets back on the job and can't see how to apply what they learned to the specific type of work they do."
His issue is not unique. This is a common challenge: Learning professionals and subject-matter experts invest the time to design courses featuring authentic scenarios to help learners develop new skills-yet some learners continue to struggle to see how what they learned relates to their own work. Learning Sciences and Cognitive Science research offers ample evidence that learning by doing in relevant contexts is highly effective, so we know the culprit here isn't scenario-based learning. What is the problem?
Failure to Transfer
The fundamental goal of instruction is to help learners retrieve and apply - or "transfer" - what they learned when it's relevant. In instances like our client's, the transfer process fails for reasons that can be frustratingly difficult to pinpoint.
In this article, we discuss how some commonly-used training techniques can inadvertently cause problems with transfer and offer research-backed strategies to overcome these obstacles.
Common training approaches can impede transfer
Learning professionals sometimes develop solutions aimed at addressing the broadest audience with the most streamlined approach possible, in the interest of increasing efficiency and minimizing time in training. Unfortunately, these solutions can inadvertently result in failure to transfer.
There are a variety of obstacles to transfer, many of which are the result of underestimating the role of context in the learning process.
Let's start by looking at two common training approaches that can contribute to the problem.
Training approach 1: lecturing
...And any other learning strategy that teaches isolated concepts and skills without contextualizing when and how to apply them.
Example
A client wanted to teach general strategies for creative problem solving to a broad audience. They didn't want to use any specific context to illustrate how to do it because it might not feel relevant to everyone.
An example strategy: "Don't quit brainstorming just because the group thinks it has exhausted their ideas."
Outcome: Without context and application, learners struggled to identify how to apply those principles to their particular work.
Problem
Teaching generalized principles and skills without allowing learners to apply them in a relevant context hampers recall. Deprived of the chance to apply new skills in a meaningful context most learners are unable to store new knowledge in such a way they will be able to recall it and apply it when they need it.1
Training approach 2: "one-size-fits-all" scenarios
AKA: Creating a single case study or scenario-based simulation to teach learners from different roles or specializations.
Example
A client aimed to teach staff from several disciplines a method of framing and addressing client problems using one discipline-specific scenario. However, participants from one of the disciplines believed they always did the same type of work no matter what client they served, and therefore did not need to consider how they framed the client problem.
Outcome: Those participants did not notice what was relevant to them about the skillset and did not apply what they learned on the job.
Problem
A scenario-based approach helps people create specific memory indices (cues) to aid in future retrieval, but when the scenario context doesn't match the learner's real-world context, they may not notice what's relevant to them; then the indices they form won't help them recall what they learned when it's relevant (without assistance).
Facilitating transfer is hard
To understand what contributes to transfer and what hinders it, it's important to consider how people think and solve novel problems.
Learning Sciences and Cognitive Science research suggests the following model of how the human mind processes, stores ("indexes"), and retrieves information. When designing and evaluating learning solutions, it's helpful to keep these key concepts front of mind.
We are case-based reasoners
Evidence suggests people are case-based reasoners; our reasoning is informed by our own experiences, and our successes and failures.2, 3
We match and compare
People seek to understand new information, situations, problems, etc., through (non-conscious) comparison. Our minds match information with our own previous experience that most closely resembles it.
We recall prior solutions
When people are seeking to solve a new problem, they apply what they can recall about solutions they have used previously (both what worked and what did not) to the new situation. Typically, this works very well. This is why we don't have to reason through every novel situation we encounter.
We use memory cues
When problem solving, people focus on a few of the most salient features of the new situation in seeking a memory match. Those features serve as memory indices to remind them of similar prior experiences.
Memory retrieval can go awry
Sometimes we fail to retrieve the memory that could help us. This happens most often when a relevant memory's surface details don't match the current context.
People aren't good at recalling past experiences that apply to new situations unless the surface features of the experience closely match the current context. When the surface details don't align, our minds often overlook the memory's deeper, structural similarities to the current context; we fail to make a match, so we don't retrieve or apply the relevant memory.4, 5 This is especially common in non-experts.6
In other cases, we recall memories that are not a perfect match, and attempt to use them to help us solve a new problem. Problems can occur when:
We apply an imperfect reminding as is
People are good at identifying superficial similarities, but it takes a greater degree of experience and expertise - or explicit hints - to recognize deeper similarities and differences.5, 6 If people use a match inappropriately, they can land on a solution that only appears to apply to the current situation.
Example: A woman had a persistent fever. She took leftover antibiotics because her past experiences suggested persistent fevers were caused by infections. It turned out to be mono - a virus. Antibiotics were useless and possibly counterproductive.
We poorly adapt an imperfect reminding
A person might realize a memory is an imperfect match and try to adapt it. Poor adaptations generate flawed solutions that can lead to failure.
Example: A novice cook prepared a butterflied (bone out) turkey using the same rule of thumb he learned for baking a bone-in turkey-approx. 13 mins per pound. He didn't know that taking the bone out reduces the cooking time per pound by more than half and ended up with a badly overcooked bird.
The take-away here: The more the surface features of a learning scenario reflect the learner's real-world context, the more likely they will be able to access and apply that key information when it's useful. This is the best way to facilitate transfer.7, 4, 5
Instruction by other means, such as lectures, or practice activities that lack essential, relatable contextual cues, result in memories that learners can't appropriately retrieve.
How to overcome common pitfalls
Here we cover strategies for overcoming some of the most common pitfalls that can impede transfer.
Pitfall 1: teaching isolated concepts and skills (e.g., via lectures, decontextualized practice exercises, etc.)
Obstacle to transfer: People don't remember isolated concepts divorced from their usage context.1
Solution: Embed learning in authentic, relevant usage contexts
Immersive scenarios are great for learning when designed appropriately. Ideally, they should be highly similar to the learners' usage context. There are additional strategies to employ to help make scenarios more effective. Read on for more about this.
Pitfall 2: using one-size-fits-all scenarios
Obstacle to transfer: Immersive scenarios provide contextual cues that help learners recall new skills on the job. However, choosing a single scenario to suit a wide audience can be a problem. If the chosen scenario isn't similar enough to a learner's real-world context, they are unlikely to transfer and apply what they learn.
Solution: Train people from similar roles and work contexts together, where possible. Contextualize instruction by using immersive scenarios that represent their work.
Alternatively, when the training audience must include a diversity of roles, allow learners to practice in a range of contexts where the same knowledge and skills apply. To help learners identify relevant similarities, it's important to explicitly compare contexts.
Use a range of reflection strategies, such as discussing "what-if" scenarios (i.e., contexts with importantly different features). Learners can think through and discuss each what-if scenario, rather than do all the work involved in working through a full solution. These scenarios allow learners to consider how they would apply the skills similarly or differently given a new set of features.8 For more reflection strategies, see Pitfall 5.
When possible, have participants "action plan" how they will apply what they learned on the job in their own context. Better yet, include an on-the-job mentor to provide discussion and feedback on real-world application.
Pitfall 3: deploying a one-and-done approach. AKA: Courses that introduce a new set of skills or ideas and only allow learners one opportunity to apply it.
Obstacle to transfer: Old habits die hard. To develop facility with a new skillset or shifts in mindset, learners need repeated practice.
Solution: See solution strategies for Pitfall 2; they're relevant here too.
Pitfall 4: failing to surface and correct learners' misconceptions
Obstacle to transfer: Good ideas alone don't change behavior. If learners don't have a chance to test their intuitions, and unearth and correct their misconceptions, they tend to hold on to those misconceptions and apply them in new situations, even when they've been taught a better way. 9, 2
Solution: Allow learners to test their intuitions and fail where possible, or at least receive feedback on their intended approach.
Share stories of the consequences of using common, flawed strategies. Stories can serve as surrogate experiences in memory. They are not as powerful as personal experience for memory retrieval, but they are far more memorable than lessons without context.
Pitfall 5: assuming learners will identify relevant skills to transfer
Obstacle to transfer: It can be difficult for learners to generalize and apply what they've learned when they don't get explicit help mapping how it applies across contexts.10, 4, 65 Providing exposure to multiple problem situations is beneficial, but you need to go further.
Solution: Use reflection strategies
Through reflection, learners can identify, map, and transfer the relevant knowledge and skills.11
Several reflection strategies can help:
"Reflection in action" and "reflection after action." Reflection in action: Ask learners questions as they work to make their problem-solving process conscious and evaluable. Reflection after action: Help learners evaluate strategies they used after they complete their work.12
"Perceptual learning"7: encourage learners to compare approaches with their peers or experts, to determine what contributes to stronger performance.
"Criteria-based reflection"13, 14: After learners do relevant, authentic work, give them the opportunity to evaluate it against best-practice criteria.
Key learning insights to carry forward
The most effective training strategies require time, which is often considered too precious a resource to devote.
In today's time-crunched world, many learning professionals worry that the addition of training strategies like the ones covered in this article will make courses too long. However, Cognitive Science and Learning Sciences research suggests they are crucial to facilitating transfer. Investing the time upfront to help learners practice, identify, and map relevant skills helps ensure that learning programs achieve their intended result.
Short, digital on-demand learning solutions have limitations and should be used within boundaries.
Today's organizational leaders and learning professionals are under increased pressure to develop quick, available on-demand learning assets that don't require the involvement of a mentor (where possible). However, it's important to keep in mind that while digital on-demand solutions can be effective in certain circumstances, they come with limitations. In a future article, we will discuss these limitations, and how to make the best use of digital-on-demand training.
Learning-by-doing in realistic and relevant contexts remains a powerful tool.
We remain strong advocates for scenario-based instruction, when the learning context is highly authentic and similar to the application context. These scenarios provide learning experiences that help people create rich memory indices that aid in retrieval, enabling them to apply skills at the time of need. As we've seen in this article, they aren't sufficient on their own, but when used in combination with some of the strategies discussed in this article, they remain one of the best ways to facilitate transfer.
REFERENCES
Brown, J. S., Collins, A., & Duguid, P. (1989). Situated cognition and the culture of learning. Educational Researcher, 18 (1), 32-42.
Schank, R. C. (1999). Dynamic Memory Revisited. New York: Cambridge University Press.
Kolodner, J. (1993). Case-Based Reasoning. San Mateo, CA: Morgan Kaufmann Publishers.
Gick, M. L., & Holyoak, K. J. (1980). Analogical problem solving. Cognitive Psychology, 12(3), 306-355.
Gentner, D., & Landers, R. (1985, November). Analogical reminding: A good match is hard to find. Paper presented at the International Conference on Systems, Man, and Cybernetics, Tucson, AZ.
Chi, M. T., Feltovich, P. J., & Glaser, R. (1981). Categorization and representation of physics problems by experts and novices. Cognitive Science, 5(2), 121-152.
Bransford, J. D., Franks, J. J., Vye, N. J., & Sherwood, R. D. (1989). New approaches to instruction: because wisdom can't be told. In S. Vosniadou & A. Ortony (Eds.), Similarity and Analogical Reasoning (pp. 470 - 497). New York: Cambridge University Press.
Williams, S. M., (1994) Anchored Simulations: Bridging the Gap between Formal and Informal Reasoning in Mathematics, unpublished doctoral dissertation, Vanderbilt University.
diSessa, A. A. (1988). Knowledge in pieces. In G. Forman & P. B. Pufall (Eds.), Constructivism in the Computer Age. Mahwah, NJ: Lawrence Erlbaum Associates, Inc, 49-70.
Perkins, D. N., & Salomon, G. (1992). Transfer of learning. International Encyclopedia of Education (2nd ed.). Oxford, UK: Pergamon Press.
VanLehn, K. (1991). Rule acquisition events in the discovery of problem-solving strategies. Cognitive Science, 15(1), pp. 1-47.
Schön, D. A. (1987). Educating the Reflective Practitioner: Toward a New Design for Teaching and Learning in the Professions. San Francisco, CA: Jossey-Bass.
Gardner, H. (1991). Assessment in context: The alternative to standardized testing. In B. Gifford & O. C. C. (Eds.), Future Assessments: Changing Views of Aptitude, Achievement, and Instruction. Boston: Kluwer, pp. 77-120.
White, B. V., & Frederiksen, J. R. (1998). Inquiry, modeling, and metacognition: Making science accessible to all students. Cognition and Instruction, 16(1), pp. 3-118.
