In this chapter, we review both well-established effects and emerging trends in disciplines such as cognitive and instructional psychology, as well as neuroscience, and discuss the implications of these findings for the design of effective learning environments. As such, we draw from the science of learning to inform training researchers and practitioners and thus affect training practice. For convenience, in this chapter we will use the term cognitive science to include diverse disciplines outside of industrial/organizational psychology such as experimental psychology, cognitive psychology, instructional design and instructional psychology, and cognitive neuroscience.
We thus differentiate the science of learning from the science of training, a term coined by Salas and colleagues (Salas & Cannon-Bowers, Reference Salas and Cannon-Bowers2001; Salas et al., Reference Salas, Tannenbaum, Kraiger and Smith-Jentsch2012). The science of training, which is rooted in I/O psychology, refers to best practices and guidelines for designing, implementing, and embedding training (in organizations) based on strong empirical support (Salas & Cannon-Bowers, Reference Salas and Cannon-Bowers2001). The science of training builds on sound training theories and research and advocates guidelines such as the importance of a priori needs assessment, boosting trainee motivation to learn, providing opportunities for practice and feedback, removing obstacles from transfer, and conducting purpose-based training evaluation (Salas et al., Reference Salas, Tannenbaum, Kraiger and Smith-Jentsch2012).
How does the science of learning (which is rooted in cognitive science) differ from the science of training? It is important to note that there is overlap. As one example, both recommend the use of practice, and both recognize the importance of trainee motivation and engagement in the instructional process. However, the science of training is built from the study of what works in training, for example, do training programs that promote self-regulation or encourage learners to make errors result in better outcomes than programs that don’t? In contrast, the science of learning is built more on a learner-centric approach, for instance, what is the difference between trainees who learn (effectively) and ones who don’t, and how can instruction be designed to foster more effective learning? The science of training embeds the training program in a broader organizational system, and considers the impact on learning of factors before, during, and after training.
Because of our broader systems perspective, which we used when researching this chapter, we ended up focusing more on training principles and less on learner characteristics. Subtopics of learner characteristics are well-covered within this volume, for example, cognitive ability (Beier, Villado, & Randall, Chapter 6), demographic differences (McCausland & King, Chapter 8), and goal orientation (Donovan & Nicklin, Chapter 7), as well as more generally by Wilson, Huang, and Kraiger (Reference Wilson, Huang, Kraiger, Christiansen and Tett2013: 543–564). We do not overlook learner characteristics because they are unimportant, but rather because we believe training researchers and practitioners are best served by understanding the fundamental ways in which most learners are alike, rather than how they may differ based on observable or measurable characteristics. Thus, we primarily focus on cognitive processes involved in learning and, primarily, the translation of research from applied cognitive and instructional psychology theory to principles for the effective design and delivery of training.
Specifically, in this chapter we draw on a broad array of cognitive science research to understand how to better the conditions under which trainees learn, and thus guide practice and research toward the design of more effective instructional environments. To narrow this endeavor, we address three primary questions: how to make training more engaging, how to make training more meaningful, and how to training more effortful (with the understanding that each general strategy results in more effective training). Finally, we discuss emerging trends in cognitive neuroscience that we believe have interesting and important implications for training design and delivery.
Training the Human Information Processor
The Human Information Processor refers to an influential 1989 chapter of the same name by Bill Howell and Nancy Cooke. Howell and Cooke (Reference Howell, Cooke and Goldstein1989: 121–182) ushered in the modern era of training theories by introducing training scholars to basic research on both human cognition and artificial intelligence and their relationship to learning and transfer. Howell and Cooke’s chapter served as an inspiration to Kraiger, Ford, and Salas’s (Reference Kraiger, Ford and Salas1993) transformative monograph, which emphasized the importance of cognitive and affective outcomes beyond merely behavioral ones as the intended results of training. This in turn has given rise to more than two decades of research on topics such as enhancing trainee self-efficacy, mental models, and self-regulation during instruction.
Due largely to the influence of behaviorists such as B. F. Skinner, instruction and training was dominated for more than four decades by a conceptualization of learning as changes in overt behavior. For example, Gagné (Reference Gagné1965) explicitly claimed that there must be a change in performance to conclude that learning occurred. While behavioral change is a pillar of modern training models (e.g., Kraiger et al., Reference Kraiger, Ford and Salas1993), an overreliance on observable behavior is deleterious for several reasons. First, adhering to behavioral instructional objectives sometimes creates disconnects between the stated learning outcomes and what is the true goal of instruction. For example, the Sunday school teacher who states that “Given a list of the 10 commandments, the student will correctly identify the 7th one” has created a measurable, behavioral objective, but what he or she really wants is for her charges to know that stealing is morally wrong. Second, the goal of many training programs is for trainees to do (or know) the right thing at the right time whether or not they can behaviorally reproduce skills immediately after training (e.g., preflight instruction on an airliner). Third, and most importantly for present purposes, a focus on behavioral reproduction as a training criterion ignores many important mental events that we know are necessary for learning (of any form) to occur.
Learning as a Cognitive Process
Given what we know from cognitive science about learning processes, what then is learning? Most definitions within I/O psychology frame learning in terms of the attainment of learning outcomes. For example, Salas et al. (Reference Salas, Tannenbaum, Kraiger and Smith-Jentsch2012: 77) stated that “Learning is a desired outcome of training when one acquires new knowledge or behavior through practice, study, or experience.” Similarly, Kraiger et al. (Reference Kraiger, Ford and Salas1993) defined learning as a relatively permanent change in affect, behavior, or cognition. Such definitions describe the outcomes of learning but not learning itself. Further, they are not particularly helpful for defining training or instruction; those would be simply anything that facilitates progress toward defined outcomes.
We prefer a definition of learning that more clearly articulates what is happening as the learner transitions from “not knowing something” and “knowing something.” (Here, we focus on “knowing” while acknowledging a broad array of possible learning outcomes including noncognitive ones.) Alexander, Schallert, and Reynolds (Reference Alexander, Schallert and Reynolds2009) defined learning as a multidimensional, dynamic process that produces an enduring change as a result of the relationship “between the nature of the learner and the object of the learning as ecologically situated in a given time and place as well as over time” (186). Another definition that incorporates both end states and transitional processes is offered by Mayer (Reference Mayer2008), who defined learning as: “a change in the learner’s knowledge that is attributable to experience ... [and] depends on ... cognitive processing during learning and includes (a) selecting – attending to the relevant material; (b) organizing – organizing the material into a coherent mental representation; and (c) integrating – relating the incoming material with existing knowledge from long-term memory” (761).
What does Mayer’s (Reference Mayer2008) definition gain us? Foremost, we have insight into the mental processes that are necessary for change (in knowledge, affect, or skill) to occur. Learners must first be aware of or attend to training stimuli. We want them to pay attention to the same content we deem important, and not be distracted by less important information (e.g., incoming texts, thoughts about tomorrow’s workload). Next, they must manipulate that information to facilitate transfer to more permanent memory. This involves at a minimum simply holding information in working memory long enough for consolidation to occur (as opposed to “in one ear ... ”) but ideally they actively process the information so that it becomes meaningful to them and more easily stored and later retrieved. Finally, they must integrate new information with what they already know, thereby facilitating transition to long-term memory. This can be done both cognitively (e.g., relate one new fact to an existing knowledge structure) and behaviorally through practice in situations like those to be encountered on the job.
In addition, this definition of learning specifies more precisely what is effective training. Effective training is anything that facilitates learning by helping trainees select, organize, and integrate new knowledge, skills, and affect to improve later job performance. In subsequent sections, we review cognitive science research relating to making training engaging, meaningful, and effortful. These three goals roughly map onto the instructional intent of helping trainees select, organize, and integrate new information.
Also, this definition of learning helps us identify what can go wrong in a training environment. Training fails when the conditions are such that learners fail to select, organize, and integrate the correct information. For example, research summarized by Dunlosky et al. (Reference Dunlosky, Rawson, Marsh, Nathan and Willingham2013) made the case that learners generally are very bad at managing learning environments. For example, they tend to pay attention to the wrong cues and overestimate their ability to focus on instructional content while multitasking (selecting). Advanced organizers are one strategy by which training can help learners select and organize critical information (Mayer, Reference Mayer1979). In a classic study by Bransford and Johnson (Reference Bransford and Johnson1972), learners who received information about the relevant context (for the training content) before reading the content learned better than did a control group or who received no pretraining guidance at all. Finally, conditions of practice are often such that information is learned in a way that it is difficult to recall appropriate skills at a later time. For example, Schmidt and Bjork (Reference Schmidt and Bjork1992) reviewed multiple studies in which the absence of stimulus variability during practice (which would most likely occur on the job) accelerates initial skill acquisition, but retards performance in transfer environments.
While a full review of ineffective learner (and trainer) practices is beyond the scope of this chapter, the preceding paragraph makes the point that if training is not properly designed and managed, learning is likely to be suboptimal. In subsequent sections, we discuss empirically supported practices for optimizing learning during training. Specifically, how can we make training more engaging, meaningful, and effortful to maximize learning outcomes? Table 2.1 provides a summary of the points to follow.
Table 2.1 Designing effective learning environments by applying cognitive and neural science to training design and delivery
| Category | Actions | Outcome |
|---|---|---|
| Make training engaging | Ensure that learners “select” or attend to relevant learning information | Learners are cognitively, physically, and emotionally immersed in both training content and learning processes |
| Replicate the transfer domain | Incorporate the physical and psychological context into the training environment | Transfer is more likely to occur because the content was learned in a similar context to the working environment |
| Capitalize on the spacing effect | Distribute the training program over multiple sessions; segment training content | Training content will be better encoded and retrieved if presented across multiple sessions compared to one session |
| Provide feedback | Offer ongoing, elaborate, and future-oriented feedback | Learners are provided with a clear direction to improve later performance, which in turn enhances learner motivation |
| Make training meaningful | Organize incoming information to build upon appropriate cognitive structures | Training content is more likely to transfer to long-term memory because it builds upon previously known information |
| Reduce cognitive load | Lessen the amount of resources required by learners to consume training content | Learners can focus more on the most relevant training content as opposed to being distracted by unimportant stimuli |
| Provide meaningful examples | Provide relevant examples to demonstrate the necessary steps to execute a task or problem solve | Learners will be more likely to compare examples and understand commonalities across multiple contexts |
| Enhance coherence | Limit the amount of unnecessary material included within training content (e.g., cartoons) | Extraneous processing will be reduced to enhance immediate learning by enhancing short-term memory retention |
| Offer signaling cues | Highlight essential material to make it distinct from less relevant material | Learners are more likely to deeply process the most important training content |
| Utilize temporal contiguity | Present corresponding words and graphics simultaneously, not in succession | Less effort is required for learners to integrate the text with the graphics, so more processing can be devoted to the content |
| Provide pretraining guidance | Provide the names, locations, and functions of key elements in advance of more complex training | Learners will use more cognitive energy on understanding systems and procedures |
| Personalize content | Use collaborative learning platforms, first- and second-person narratives and personable language | Learners will relate to personalized content in a more meaningful way compared to how they would process lists of facts and procedures |
| Incorporate visuals | Integrate simple yet informative graphics to convey relevant information | Reduce germane load and provides the learner with relevance of the material to tasks to be performed |
| Make training effortful | Encourage learners to make effortful cognitive and behavioral connections between training content to prior knowledge and future applications | Ensure that long-term learning occurs by helping the learners retain and access information in future applied settings |
| Provide opportunity for practice | Provide opportunities for learners to apply training content to well-designed practice activities | Promote both recall and application of the learned knowledge and skills to real-world tasks, thus enhancing long-term retention |
| Interleave training content | Intersperse practice activities from previously learned topics throughout training modules | Learners will have to retrieve prior information from long-term memory, thus increasing the likelihood of deeply encoding content |
| Use practice variability | Require learners to engage in a variety of iterations of a practice activity, varying conditions across trials | Learners will be more likely to transfer training content to novel contexts |
| Test learners’ knowledge of content | Have trainees answer test and quiz questions to assess what they learned in training | Testing learners will increase their retention and retrieval of the training content, thus improving overall learning |
| Apply established neuroscientific findings | Although in its infancy, some neural underpinnings of learning have been identified (e.g., sleep) | Learning can be enhanced by integrating best practices from neuroscience to capitalize on how our brains best function |
| Allow for sleep consolidation | Allow consolidation effects using sleep to occur by spreading out training over multiple days | Newly learned information will be better stored and retrieved after being reactivated and reorganized into new representations |
| Encourage trainees to get adequate sleep | Structure the work and training environment to support good sleeping habits | Getting sufficient sleep can enhance motor skill performance, language acquisition, and learning declarative knowledge |
| Use breaks to increase attention of trainees | Strategically implement breaks or low cognitive load activities to allow for a resetting of attention | Voluntary attention reflects direct control of cognitive resources to attend to stimuli long enough to aid in transfer to long-term memory |
| Reduce the occurrence of distracting stimuli | Minimize irrelevant stimuli unless it is to provide cognitive breaks for learners’ retention and retrieval of the content | The brain can focus its energy on learning and remembering the most relevant content to the training, thus reducing unnecessary cognitive load |
Make Training Engaging
Mayer (Reference Clark and Mayer2008) argued that the first step in the learning process is for learners to “select” or attend to relevant learning information. From a pedagogical perspective, manipulating the instructional environment so that the learner engages in the relevant content is thus critical to facilitating learning. Both theory and research suggest that learners should be engaged in the learning process to maximize their learning (Bell & Kozlowski, Reference Bell and Kozlowski2008). Borrowing from the organizational psychology literature, engagement can be defined as a personal investment or absorption of the self into work-related tasks (Kahn, Reference Kahn1990). Extending this to the training domain, we define learner engagement as the extent to which learners are cognitively, physically, and emotionally immersed in both training content and learning processes.
Learner engagement matters because considerable research shows that deeper processing of training content facilitates later recall and/or application (Craik & Lockhart, Reference Craik and Lockhart1972). Dunlosky et al. (Reference Dunlosky, Rawson, Marsh, Nathan and Willingham2013) speculated that “intentionality” on the part of learners promotes deeper processing that in turn enables learners to extract more meaning from materials. Learner engagement also promotes self-regulation (Schunk & Zimmerman, Reference Schunk and Zimmerman2012). Self-regulated learners are able to monitor learning activity and adjust proximal goals, effort, and behavior to accomplish more distal learning goals.
The importance of capturing and maintaining the interest and the energy of learners has been long recognized. For example, one of the founders of modern instructional design theory, Robert Gagné (Reference Gagné1965), referred to the “first event of instruction” as gaining learner attention. That is, Gagné recognized that instruction cannot occur unless the trainer first has the attention of the learner, who then becomes engaged in the process. Additionally, Herb Simon (Reference Simon1998), one of the first psychologists to study learning from a cognitive perspective, reflected on nearly 50 years of research and concluded that making the learner an active partner is the single most important act of instruction.
Clark and Mayer (Reference Mayer2008) distinguished between two types of engagement – behavioral and cognitive engagement – and argued that learning is optimized when both are high (although they also argue for the relative importance of cognitive engagement over behavioral). Behavioral engagement simply refers to acting upon training content – reading text and practicing skills are two forms of behavioral engagement. Cognitive engagement refers to mental activities that promote understanding, organization, and integration. Including learner-focused activities (e.g., questioning, elaborating, and explaining the training content) can enhance behavioral and/or cognitive engagement (Clarke, Reference Clark2014). For example, learners who ask themselves questions such as “Why is this important for me to know?” or “How is this related to the training objectives?” are more likely to master training content than learners who don’t. Thus, trainers should build in time to encourage learner reflection.
The combination of behaviorally and cognitively active learning environments works best when engagement activities share four properties: (1) they do not impose undo cognitive load; (2) they require behavioral responses similar to those required in the transfer domain; (3) they include spaced practice (see following text); and (4) they include explanatory feedback (Clark & Mayer, Reference Clark and Mayer2008). We talk about the importance of reducing cognitive load within the subsequent section on making training meaningful, so we will instead focus on how the transfer domain, spaced practice, and feedback can be used to increase engagement in the following text.
Replicate the Transfer Domain
Behavioral engagement can be enhanced by incorporating the physical and psychological context of the transfer environment into the training experience (Clark & Mayer, Reference Clark and Mayer2008). In other words, the training context should replicate the transfer context (i.e., the working environment in which what was learned during training will be applied) as closely as possible.
Capitalize on the Spacing Effect
Spacing out training activities over time, as opposed to massing all training into one session, increases trainee engagement by enhancing long-term retention (Clark & Mayer, Reference Clark and Mayer2008). Also known as distributed practice, the spacing effect occurs when learning is spread out over multiple intervals, resulting in better memory and retrieval in comparison to learning in one massed session. Evidence demonstrating the benefit of segmenting studying over multiple sessions dates back to 1885 and has been since reliably demonstrated in hundreds of studies (see Cepeda et al., Reference Cepeda, Pashler, Vul, Wixted and Rohrer2006). The spacing effect has been found to be one of the most “dependable and replicable phenomena in experimental psychology” (Dempster, Reference Dempster1988: 67).
Spaced learning works because it capitalizes on the fundamental memory mechanisms of encoding and retrieval. Multiple exposures to the same learning material increases the number of retrieval pathways created (Glenberg, Reference Glenberg1979). The same material is encoded not only more often, but also can be linked with multiple cues due to the variations in temporal, physical, or mental contexts (Estes, Reference Estes1955). Sisti, Glass, and Shors (Reference Sisti, Glass and Shors2007) found that spaced-trial trainings resulted in newer cells being more likely to survive in the hippocampus, which resulted in more persistent memories of spaced versus massed training materials. The spaced training allowed for brain cells to regenerate between study sessions, which resulted in more permanent neural connections and stronger memories, resulting in increased learning. As resources do not often allow for training to be broken down into multiple sessions, trainers can capitalize on the spacing effect by integrating practice activities throughout a training session, as opposed to administering all practice activities at the end. Whenever possible, training should occur over multiple sessions to not only increase engagement but also overall performance, as research shows that when learning is spaced, performance is better when compared to learning in one massed session (Baddeley & Longman, Reference Baddeley and Longman1978).
Provide Feedback
Trainees should be provided with thorough and ongoing feedback as another means to enhance engagement with training content. The general effectiveness of feedback on performance has been confirmed through several meta-analyses both in terms of work performance (Kluger & DeNisi, Reference Kluger and DeNisi1996) and computer-based training (Azevedo & Bernard, Reference Azevedo and Bernard1995). Applied cognitive and instructional psychology research reveals specific ways in which feedback during practice can be optimized. Responses to practice should be provided in such a way that facilitates better performance in the future – it should be more feed “forward” rather than feed “back” (Clark, Reference Clark2014). Research also shows that elaborative feedback, as opposed to only being informed that one’s work is merely correct or incorrect, results in better learning outcomes (Azevedo & Bernard, Reference Azevedo and Bernard1995; Moreno, Reference Moreno2004). Feedback should also be provided after each step when learners are engaged in complex practice activities. Compared to providing feedback only at the end of the activity, step-by-step feedback can result in better learning and more trainee motivation (Corbalan, Paas, & Cuypers, Reference Corbalan, Paas and Cuypers2010). Therefore, learners should be provided with detailed feedback as they complete their practice activities.
Summary
Training programs should be designed to maximize trainee engagement with the content. Engagement results in deeper processing during training and, by consequence, facilitates later recall and greater likelihood of posttraining applications. There are several strategies that can be applied to enhance the cognitive, physical, and emotional immersion of trainees. First, the transfer domain should be replicated as closely as possible to the working environment where the training content will be applied. Second, training programs should be spaced out over time instead of being lumped into one singular session as a way to increase long-term retention of the training material. Lastly, trainees will be more engaged with the learning process if provided with clear and timely feedback throughout training. We will now discuss the significance of meaning on learning outcomes, as well as present explicit strategies that can be applied to make training more meaningful for participants.
Make Training Meaningful
The second phase of learning is organizing incoming information in way that builds appropriate cognitive structures to facilitate transfer to long-term memory (Mayer, Reference Mayer2008). The role of instruction then is to prime the appropriate cognitive processing in learners, helping them to form coherent cognitive representations and integrate new representations with what is already known, including how the new knowledge or skills will be applied back on the job. This is also referred to as generative processing, in which the learner imposes structure on part based on what is already known (Sweller, Reference Sweller1999). The more training content is meaningful to learners, the more likely they are to organize new information in meaningful ways.
Reduce Cognitive Load
Empirical and theoretical reasons for the importance of making training meaningful rest largely on reducing cognitive load during instruction. Cognitive load can be thought of as the sum of total press on a learner, given the cognitive resources of that learner (Chandler & Sweller, Reference Chandler and Sweller1991). There are three types of cognitive load: intrinsic, extrinsic, and germane. Intrinsic load is related to the difficulty of the material, and while difficult content can sometimes be broken down into smaller chunks, intrinsic load is primarily a function of the tasks to be trained, not instructional design.
In contrasts, extraneous and germane load are directly related to training design. Germane load is related to the level of (useful) mental effort devoted to processing, constructing, and automating cognitive structure or schema in working memory. As a simple example, germane load for a difficult learning task can be reduced by highlighting key words, the use of graphics, or providing learning heuristics (e.g., “Every Good Boy Does Fine”). Conversely, germane load for a simple task can be increased by reducing presented content and making learners generate their own definitions or examples. Extraneous load is created by the way in which material is presented to learners, requiring superfluous processing on the part of the learner. Training segments that run too long, or training slides that contain irrelevant or misleading information are examples of extraneous load. In general, instructional design that reduces extraneous load and optimizes germane load is more effective than training that does not (Merriënboer & Sweller, Reference Merriënboer and Sweller2005).
When the training in meaningful to learners, extraneous load is reduced. For example, if the equipment used to train a new skill is identical to that used on the job, then learners can focus on skill acquisition, and devote fewer cognitive resources to processing the differences between equipment in training and equipment used in the job. The use of unfamiliar jargon and acronyms can also create extraneous load, and when trainers cannot avoid the use of jargon, having an easily accessible glossary is one strategy for minimizing deleterious effects.
Provide Meaningful Examples
One well-supported strategy for reducing cognitive load is the worked example (Renkl, Reference Renkl and Mayer2005: 229–245), in which the trainer demonstrates solving a problem step by step to complete a task or solve a problem. Worked examples reduce extraneous load by the use of scaffolding earlier steps – learners are presented with only the information they need to solve the next step. Germane load can be managed by prompting learners for self-explanations of actions at key steps, focusing their attention (through prompts) of what they did right and what led to errors in execution.
As suggested in the preceding text, examples are an effective way to make training meaningful. Examples should not only be relevant to the tasks performed on the job, but should demonstrate the steps necessary to execute a task or solve a work-related problem. Clark (Reference Clark2014) listed other characteristics of effective examples: (1) for strategic tasks, the examples should be set in a variety of contexts; (2) use more worked examples for novice learners and practice assignments for experience learners; and (3) increase learner engagement related to the examples by using faded examples (gradually reducing the number of steps described), adding self-explanation questions required of learners, and require learners to compare examples to understand commonalities and differences across contexts.
Enhance Coherence, Offer Signaling, and Use Temporal Contiguity
Mayer (Reference Clark and Mayer2008) summarized several evidence-based and theoretically grounded principles for reducing extraneous processing and increasing essential processing by learners during training. While several of these apply more to the design of computer-based learning environments, others apply to any form of instruction. Three key principles for reducing extraneous processing are coherence, signaling, and temporal contiguity. The coherence principle states that extraneous material should be reduced in training content; a common example of extraneous material is a cartoon or humorous graphic on a presentation slide. While humor can help with long-term learner engagement, in the short term, it can create extraneous processing and reduce immediate learning. The signaling principle states that essential material should be highlighted to separate it visually or orally from less essential material. A common signaling technique used by effective speakers (and parents!) is to lower one’s voice in advance of the most important information. Finally, the temporal contiguity principle states that corresponding words and graphics should be presented at the same time, not in succession. For example, in a web-based statistics course, the formula for the slope of a line-of-best-fit should be given on the same slide that illustrates that line.
Provide Pretraining Guidance
Two key empirically supported principles for encouraging essential processing are segmenting and pretraining guidance. Segmenting is simply breaking up larger presentations into shorter, learner-paced segments. While segmenting can be difficult in classroom training, it is more readily accomplished in either online, one-on-one, or on-the-job training. Segmenting is one form of guidance. Another form, pretraining guidance, presents learners the names, locations, and functions of key elements in advance of more complex training. We would suggest that pretraining guidance can take the form of a training prerequisite, training prework, refresher training within training, or sequencing content within training so that fundamental declarative knowledge precedes advanced procedural knowledge.
Personalize Content
Finally, Clark (Reference Clark2014) advocated for two additional techniques to promote meaningful learning. The first is to personalize content. While we think of personalized learning as being primarily related to one’s job, personalized learning is also accomplished through the use of simple techniques such as use of first- and second-person language, collaborative learning platforms including discussion boards (see following text), an engaging or personable voice by the trainer, and training content that takes the form of a narrative or story more than a listing of facts and procedures (Ginns, Martin, & Marsh, Reference Ginns, Martin and Marsh2013). While many successful trainers intuitively understand the importance of being personable, it is useful to note that such a style is also empirically supported.
Incorporate Visuals
The second of Clark’s (Reference Clark2014) additional techniques to promote meaningful learning is the use of effective visuals. Visuals can reduce germane load and also provide the learner with relevance of the material to tasks to be performed (Butcher, Reference Butcher2006). The importance of visuals is reinforced by considering the number of home do-it-your-selfers who ignore text-based assembly instructions to instead watch narrations online. Consistent with the goal of reducing extraneous load, visuals should generally be as simple as possible to convey relevant information. The most obvious form of a useful visual or graphic is one that illustrates stimuli to be acted on, or actions to be taken. For example, in a word-processing class, a screenshot can be shown with key menu functions highlighted. A second useful form is an explanatory visual. Explanatory visuals are relatively simple graphics that show static or dynamic relationships among elements. One type of an explanatory visual is an organizational one, for example, including an organization chart during socialization to help newcomers understand their position within the organization. A relational visual summarizes quantitative information in a meaningful way; for example, a pie chart of the amount of the earth’s surface covered by land versus water creates a clear understanding of how vast are our oceans. Transitional visuals can either be animated or simply a series of steps, for example, a flow diagram showing the steps a researcher must follow to get internal review board approval.
Summary
In sum, once learners are engaged and attending to the relevant training content, the goal of training is to increase the processing of essential content, and minimize processing of extraneous content. Principles of cognitive load, increasing the perceived relevance of training, the personalization of training delivery, and the use of effective visuals all support these objectives.
Make Training Effortful
The goal of training is not to simply present engaging, meaningful material to a trainee, but to also employ methods that help the trainee retain and access information later on in an applied setting. This relates back to the integrative step of Mayer’s (Reference Mayer2008) definition of learning. Learners should make cognitive and behavioral connections between the training material to prior knowledge and future applications. In other words, training should be effortful to ensure that long-term learning occurs. Learning activities should consist of “desirable difficulties” that require the learner to actually work toward mastery over material (Bjork, Reference Bjork, Metcalfe and Shimamura1994: 185–205). By rigorously using the knowledge and skills presented in a training program, stronger encoding and retrieval pathways will be created when compared to engaging in more passive learning activities (e.g., listening and reading). One way to increase learner effort is through carefully designed practice conditions.
Provide Opportunity for Practice
To make training material stick, learners should be provided with ample opportunities for practice (Salas et al., Reference Salas, Tannenbaum, Kraiger and Smith-Jentsch2012). Practice is defined as a well-designed activity assigned to promote both recall of the training content and, more importantly, application of the learned knowledge and skills to real-world tasks (Clark, Reference Clark2014). Drawing from cognitive-neuroscience, practice aids in mastery over training material due to long-term potentiation. Long-term potentiation is the repetitive activation of synapses caused by activity- and experience-dependent learning. The increased synaptic strength that results restructures neural circuitry, which explains how memories are created and stored at the molecular level (Bear, Cooper, & Ebner, Reference Bear, Cooper, Ebner and Inkhauser1989: 156–160). In other words, the more often one engages with the training stimuli, the more likely it will be stored in long-term memory.
Practice activities should replicate how the training material will be used on the job (Baldwin & Ford, Reference Baldwin and Ford1988). According to transfer appropriate processing theory, content is most likely to be remembered when encoding and retrieval processes are as similar as possible (Morris, Bransford, & Franks, Reference Morris, Bransford and Franks1977). Greater similarity between training conditions and performance conditions is also referred to as near (rather than far) transfer (Clark & Voogel, Reference Clark and Voogel1985). Therefore, it is important to make practice activities as realistic as possible. Several recommendations for improving near transfer include increased specificity about where and how the training will be used on the job, overlearning of the task, and emphasizing procedural knowledge and skills to be utilized on the job throughout training (Clark & Voogel, Reference Clark and Voogel1985).
Beyond ensuring near transfer, the extent to which practice in training results in successful transfer can depend on training design factors (Hesketh, Reference Hesketh1997). The science of learning has informed us of several empirically supported methods we can apply to make practice activities as effective as possible. These techniques include interleaving learning, practice variability, and providing feedback.
Interleave Training Content
Practice activities covering all previously learned topics should be interspersed throughout training modules. As opposed to covering the material related to one topic in a blocked practice session before moving onto another topic (which we know is not effective, as discussed in the Capitalize on the Spacing Effect section), interleaving is a method that exposes learners to a variety of material by switching between topics throughout the session. Interleaving works because it requires information to be retrieved from long-term memory, whereas blocked practice allows information to remain in working memory with the risk of never being encoded into long-term memory in the first place (Rohrer & Taylor, Reference Rohrer and Taylor2007). Empirical support for interleaving was found when students in an interleaved-practice group had substantially better accuracy compared to those in a blocked-practice group (Taylor & Rohrer, Reference Taylor and Rohrer2010). It is important to note that accuracy was assessed by a criterion test administered one day after the practice took place (the blocked practice group performed better during the practice), demonstrating the positive effect of interleaving on long-term memory.
Moreover, interleaving can help promote transfer of the training material as it is more likely that the trainee will have to switch between a variety of tasks on the job as opposed to tackling a problem by applying only one skillset at a time. Therefore, when designing practice activities, exercises covering a variety of topics should be delivered throughout the training, as opposed to only providing module-specific practice activities.
Use Practice Variability
One way to create engagement and aid retention is through low stakes quizzes or testing. Trainees should be tested on training material, and tested often. Learners have been found to have better retention of learned material after being tested on that material as opposed to simply restudying it, otherwise known as the testing effect (Roediger & Karpicke, Reference Roediger and Karpicke2006). The testing effect works because it capitalizes on the retrieval process of memory (Rohrer, Taylor, & Sholar, Reference Rohrer, Taylor and Sholar2010). Retrieval plays a dynamic role in learning insofar that the more often information is retrieved from long-term memory, the easier that information will be remembered at a later time (Bjork, Dunlosky, & Kornell, Reference Bjork, Dunlosky and Kornell2013).
Another way in which trainees are exposed to different types of stimuli is through practice variability. Practice variability occurs when the conditions of practice vary substantively across trials. For example, a novice tennis player who is learning a backhand experiences practice variability if, instead of receiving 100 consecutive shots in the exact same place and requiring the exact same motion, instead must move in and out, side to side, and practice multiple versions of the same motion in response to different stimuli. It has been well-established that practice variability slows initial skill acquisition, but promotes effective transfer to novel tasks (Schmidt & Bjork, Reference Schmidt and Bjork1992), which is often the objective for training. It appears that with reduced variability, there are insufficient processing (effort) demands on learners to properly facilitate long-term learning (Wulf & Shea, Reference Wulf and Shea2002).
Test Learners’ Knowledge
Dunlosky et al. (Reference Dunlosky, Rawson, Marsh, Nathan and Willingham2013) reviewed several empirically supported techniques for increasing learner engagement, including quizzes, tests, and clickers. Dunlosky et al. distinguish practice tests from more formal periodic assessments of learner mastery (e.g., end of training exams). Practice tests and quizzes show relatively strong effects on subsequent learning, although the mechanisms by which they work are not completely clear. Practice tests seem to result in both direct and mediating effects. Direct effects suggest that the act of taking the test enhances subsequent retrieval processes, for example, triggering more elaborative retrieval. Mediating effects suggest that testing may trigger better organization and storage of tested material, which in turn facilitates later retrieval (see Dunlosky et al. for an overview). Regardless of the mechanism, practice tests increase active engagement by learners which improves retention and retrieval. Multiple studies have shown that learners who respond to questions during training using clickers demonstrate better learning outcomes than learners who do not respond, or do not have access to clicker technology (e.g., Mayer et al., 2009).
Summary
Training should be designed is such a way that learners should take an active approach to their learning. The more trainees work at applying the knowledge and skills taught during the training, the better they will remember this material and use it for better performance on the job. Trainers can incorporate these desirable difficulties which capitalize on encoding and retrieval processes by building practice and testing activities into training programs. Practice and testing allows learners to spend more time on the output side of learning, rather than the input side (Bjork & Bjork, Reference Bjork, Bjork, Gernsbacher, Pew, Hough and Pomerantz2011: 56–64), resulting in better memory and transfer of the training material.
Emerging Trends from Neuroscience
We title this section “emerging trends from neuroscience” because while cognitive neuroscientists are making great progress in identifying the ways in which brain activity (broadly defined) is related to learning, this research is not nearly as mature as that of cognitive science (Bruer, Reference Bruer2006). As such, the implications from cognitive neuroscience research for best practices in training are not as clear. Nevertheless, we believe it is important for training researchers to be aware of emerging neuroscience research for several reasons. First, understanding brain activity in response to learning stimuli may change the way we study learning, if not the questions we ask and the way we design training programs. Additionally, as neurological underpinnings of learning become better understood, greater collaboration between training researchers and neuroscientists will be necessary to interpret and apply these findings.
Implications for Measurement
First it is important to understand the predominant experimental paradigm and measurement techniques to understand the neuroscientific implications for the science of learning. Recognize though that this is an oversimplification of the research on both the stimulus and response side. Typically, neuroscientists use some form of functional technique, in which several variations of a learning or memory task are given and direct or indirect measures of brain activation is assessed (Bunge & Kahn, Reference Bunge and Kahn2009). The former refers to methods of directly measuring electrical activity resulting from neuronal firing; the most common such method is electroencephalography (EEG). The latter methods include functional magnetic resonance imaging (fMRI) in which MRI scanners detect the presence of elevated levels of oxygen in certain areas of the brain, symptomatic of greater neural activity.
Insights from neuroimaging can be seen by considering research on error management training. We know from training research that error-management training works – under the right conditions, encouraging trainees to make errors is associated with greater learning by trainees (Keith & Frese, Reference Keith and Frese2008). Further error-management research has begun to identify training conditions that moderate the effectiveness of specific training conditions related to the promotion of errors (e.g., Carter & Beier, Reference Carter and Beier2010). Less is known about individual differences as moderators, or precisely how exposure to errors promotes learning. Klein et al. (Reference Klein, Neumann, Reuter, Hennig, Von Cramon and Ullsperger2007) used fMRI data to study the role of dopamine in reactions to negative feedback on a learning task. The researchers found that learning from errors requires dopaminergic signaling; additionally, learners with reduced dopamine D2 receptor densities responded less to negative feedback than did “normal” learners. While applications of such research are still well in the future, we can speculate that neuroimaging screens one day could be used to assign learners to optimal training conditions.
In the end, training is effective when it is well-designed and trainees “learn.” One could argue that given evidence of increased recognition, recall, transfer, and so forth, that what is occurring in the brain is less critical to observe and quantify; that is, the neurological activity that underlies learning is a set of mediating processes that we don’t need to study directly to improve training. However, the preceding example highlights one benefit of understanding the neurological underpinnings of learning processes. A second benefit is that evidence of increased brain activation confirms that learning was, if not intentional, replicable. That is, we have increased confidence that the same training intervention would work in future applications. Thus, neural imaging has great potential as a dependent variable in training research. Finally, neuroimaging has potential for studying internal processes (or confirming a construct) like meta-cognition or self-regulation (Willingham, Reference Willingham2012). It is well-accepted in the science of learning that meta-cognition (thinking about thinking) not only occurs, but is important to learning over time. Yet, measuring meta-cognition in the moment is challenging – simply asking participants what they are monitoring disrupts the act of monitoring. However, interventions that encourage meta-cognitive activity can be evaluated using neuroimaging techniques. Thus, it is easy to understand why Poldrack (Reference Poldrack2000) concluded that the “neuroimaging of learning and development is one of the most exciting and quickly growing areas of cognitive neuroscience, and will no doubt continue to grow as new techniques ... are added to the quiver of neuroimaging methods” (p. 10).
How Applicable Is Neuroscience Research?
That said, there is currently a strong debate on the extent to which neuroscience-based findings can be readily translated into the design and delivery of training and education (Bruer, Reference Bruer2006; De Bruyckere, Kirschner, & Hulshof, Reference De Bruyckere, Kirschner and Hulshof2015; Hruby, Reference Hruby2012; Tokuhama-Espinosa, Reference Tokuhama-Espinosa2014). On the one hand, Tokuhama-Espinosa calls for a completely new instructional paradigm based on an integration of mind (psychology), brain (neuroscience), and education science, assuming that research is sufficiently advanced in all three domains inform practice. On the other hand, De Bruyckere et al. cautioned that the state of neuroscience relevant to learning is as yet too immature to readily translate it into instructional principles, concluding, “For the time being, we do not really understand all that much about the brain. More importantly, it is difficult to generalize what we do know into a set of concrete precepts of behavior, never mind devise methods for influencing that behavior” (p. 94; see also Bruer, Reference Bruer2006 and Hruby, Reference Hruby2012).
As Willingham (Reference Willingham2009) noted, the application of neuroscience to instructional principles faces both a vertical and horizontal problem. The vertical problem is one of levels analysis. The neuroscientist studies phenomena somewhere between the level of the neuron and brain structures, and attempts to study cognitive functions (e.g., recall) in isolation. Thus, we cannot assume that findings at this level generalize to the learner in the training context, in which the lowest level of analysis is the learner and cognitive functioning occurs in a rich milieu. The horizontal problem refers to the mapping of stimuli to brain activity and back. Even if we are able to map human cognitive processes and representations in response to certain stimuli, that doesn’t necessarily inform how instructional methods on how to trigger that activity in a learning context. For a well-rounded perspective on the promise and shortcomings of applying neuroscience to learning contexts, see Devonshire and Dommett (Reference Devonshire and Dommett2010). For present purposes, we reiterate that the primary value to training researchers of neuroscience is its role as unique (and perhaps necessary) measures of mediating and dependent variables. For training practitioners, we encourage caution in the application of neuroscience findings to instructional practices unless there is a relatively mature set of findings that address both the horizontal and vertical problems. We illustrate two such applications in the following text.
Allow for Sleep Consolidation and Adequate Sleep
Research over the past several decades has accumulated evidence of the relationship between sleep and cognitive performance, including learning. It is well-established that workers worldwide suffer from a lack of sleep (Killgore, Reference Killgore2010). According to a 2008 poll by the National Sleep Foundation, 44% of survey respondents reported receiving less than seven hours of sleep during the work week, and 29% reported becoming very sleepy or falling asleep at work in the prior month.
The quantity and quality of sleep effects learning performance in two broad ways. First, there is considerable support for the proposition that sleep has been shown to enable consolidation of new information into memory (Diekelmann & Born, Reference Diekelmann and Born2010), with consolidation being an active process in which new knowledge is reactivated and reorganized into new representations (Born, Rasch, & Gais, Reference Born, Rasch and Gais2006). Born et al. provided an overview of the specific mechanisms by which sleep types (e.g., slow-wave v. REM sleep) facilitate consolidation depending on where in the brain memories are formed. The benefits of sleep on learning (Curcio, Ferrar, & De Gennaro, Reference Curcio, Ferrara and De Gennaro2006) has been documented for a number of outcomes including motor skill performance (Walker et al., Reference Walker, Brakefield, Morgan, Hobson and Stickgold2002), language learning (Fenn, Nusbaum, & Margoliash, Reference Fenn, Nusbaum and Margoliash2003), and declarative knowledge (Stickgold, Reference Stickgold2005). Consistent with our earlier discussion, note though that Stickgold pointed out that there is insufficient data as of yet to suggest sleep aids in more complex forms of human learning.
When coupled with earlier cited behavioral research, findings on the relationship between sleep and learning help explain the benefits of distributed practice or spacing over multiple days. Neuroscientific research indicates that these benefits accrue in part from the opportunity to sleep between learning trials. Stickgold (Reference Stickgold2005) quoted the first century AD, Roman rhetorician Quintilian who in turn noted that “what could not be repeated at first is readily put together on the following day; and the very time which is generally thought to cause forgetfulness is found to strengthen the memory” (1272). Though the benefits are documented more for declarative knowledge and simpler skills, our recommendation is that, when feasible, training be spread over multiple days. Sleep prepares the brain for learning.
The second impact of sleep on training performance is when learners have not slept enough. There is considerable research that documents the deleterious effects of lack of sleep on cognitive performance. Killgore (Reference Killgore2010) noted that there is a “broad consensus” that a lack of sleep results in slower response speeds and increased variability (primarily negative) in alertness, attention, and vigilance, though there is less agreement regarding the effects of sleep deprivation on certain higher level cognitive capacities including memory. Sleep deprivation has also been shown to have negative effects on integrating information in novel or creative ways (Payne, Reference Payne2011), divergent thinking, and self-regulation (Harrison & Horne, Reference Harrison and Horne2000), processes each important in training performance.
While research originally focused on sleep deprivation, more recent research has also focused on chronic sleep restriction. The former is defined as the result of at least one night with no sleep, the former refers to continually restricting sleep below one’s optimal time in bed (TIB) (Banks & Dinges, Reference Banks and Dinges2007). For example, Van Dongen et al. (Reference Van Dongen, Maislin, Mullington and Dinges2003) compared participants receiving four or six hours of TIB over 14 consecutive days to participants who went three nights without sleep. We suggest that empirical evidence of chronic sleep restriction is considerably more relevant to most of the workforce than are studies of sleep deprivation. Somewhat surprising then are the conclusions that the neurobehavioral effects and cognitive performance effects due to chronic sleep restriction are very similar to those found for multiple days of total sleep loss (Balkin et al., Reference Balkin, Rupp, Picchioni and Wesensten2008; Banks & Dinges, Reference Banks and Dinges2007; Curcioa, Ferraraa, & Gennaroa, Reference Curcio, Ferrara and De Gennaro2006). With respect to neuropsychological functioning, Banks and Dinges concluded: “Recent experiments reveal that following days of chronic restriction of sleep duration below 7 hours per night, significant daytime cognitive dysfunction accumulates to levels comparable to that found after severe acute total sleep deprivation” (519). Further, Curcioa et al. reported strong correlational evidence between sleep loss and academic performance among adolescents.
In sum, emerging research suggests that chronic sleep restriction can negatively affect cognitive performance, which in turn negatively affects learning in training. There are many determinants of sleep restriction, but undoubtedly one is work, for example, shift work, stress, and high workloads (Âkerstedt, Reference Âkerstedt2006; Harrison & Horne, Reference Harrison and Horne2000). Training can exasperate the strain caused by work, either because workers are held responsible for fulfilling their work roles or, in the case of online training, workers are expected to complete training on their own time (which is often late at night after family responsibilities are met). Accordingly, we recommend that in preparing workers for training, organizations ensure they have adequate time to fulfill their work roles without the need for less sleep time.
Use Breaks and Reduce Distractions
Attention is critical to learning; attention is the primary means by which we hold sight and sound in sensory memory long enough for it to transfer to working memory. Attention also explains why two learners can attend the same training course and come out with very different memories of what was covered and what should be applied. Physiologically, attention consists of top-down signals throughout the cortex, with frontoparietal areas regulating activity in sensory cortex and the thalamus. Importantly, while neural activity is activating by perceptual processing (attending to external stimuli), it can also be increased simply by directing attention to a location (i.e., focusing).
Because attention requires focused effort, and because our brains were not designed for sustained attention, there are multiple implications both for learners and for instructional design. As should be intuitive, the brain needs rest and recovery to optimize focus and attention. In the short term, this means planned breaks in the learning process (e.g., Ariga & Lleras, Reference Ariga and Lleras2011). Trainers frequently (and smartly) use learning games or insert simpler material to enable recovery time. Research also suggests that physical activities such as walking or even looking out a window (Kaplan & Berman, Reference Kaplan and Berman2010) are restorative as is practicing mindfulness (Evans, Baer, & Segerstrom, Reference Evans, Baer and Segerstrom2009; Vago & David, Reference Vago and David2012).
Assisting the learner in returning to attention assumes we have it in the first place. Researchers distinguish between voluntary and involuntary attention (Posner & Rothbart, Reference Posner and Rothbart2007). Involuntary attention lasts several milliseconds and is the process involved, for example when we are startled by a sneeze behind us. Voluntary attention reflects direct control of cognitive resources to attend to stimuli sufficiently long enough to aid in transfer to long-term memory, and thus is a key to effective instruction.
To oversimplify, effective instruction requires capturing and maintaining learner attention, although, as we’ve seen, maintaining attention is largely taking short breaks and coming back to voluntary attention. As effective trainers know, one way to increase effortful attention is to signal that upcoming information is significant and/or relevant. The first author once observed a half-day training program that culminated in trainees breaking a board with their hand in front of their peers. Although the training content consisted largely of having goals and believing in oneself, the trainers held high levels of attention for four hours by continually assuring, “The next thing we tell you will be critical to breaking the board.”
Neurologically, evidence exists for partially separated areas of the brain that hold responsibility for attentional functions (Corbetta & Shulman, Reference Corbetta and Shulman2002) or networks of brain areas that carry out different attentional functions. These areas are triggered primarily by stimulus novelty, as well as by relevance of stimuli to individual goals and to the defensive and appetitive motivational systems (Bradley, Reference Bradley2009). In short, we notice external stimuli that are novel and relevant to our fundamental needs. While it is a stretch to suggest that trainers can make training more relevant by appealing to (on-the-job) survival, neuropsychological evidence supports traditional training practices such as providing advanced organizers that both signal what is coming up in training content and how training is relevant to work-related goals (Mayer, Reference Mayer1979).
Neuroscience research suggests that even as we try to attend to relevant stimuli, task and environmental characteristics distract us. Salient-signal suppression theories of attentional control (e.g., Sawaki & Luck, Reference Sawaki and Luck2010) propose that salient content in our visual field compete for control of our visual perception systems and that the visual system (not the mind) influences whether or not to focus on the most salient features. Further, attention can be influenced by the context, priming, or the individual’s attentional control settings. Neuropsychological evidence suggests that our attention is fleeting and that visual distractors are, well, distracting, and thus, instruction should be designed in a way that minimizes irrelevant stimuli unless it is to provide cognitive breaks for learners (Sitzmann & Johnson, Reference Sitzmann and Johnson2014). A challenge for future research is to determine in more detail what distinguishes a dysfunctional distraction from a functional one.
Finally, as too many trainers and college instructors know, focusing attention is becoming an increasingly difficult task, particularly for learners who choose to multitask by texting or checking social media from the classroom (Carrier et al., Reference Carrier, Cheever, Rosen, Benitez and Chang2009). While both our abilities to selectively attend to priority information and hold it in working memory declines with normal aging beginning in the early twenties (Gazzaley et al., Reference Gazzaley, Cooney, Rissman and D’Esposito2005), attention dysfunction seems more prevalent among young learners. It appears that it is not only the distraction of multitasking but the expectations formed by digital media that affect young learners (Richtel, Reference Richtel2010). These same learners may also place lower priority on paying attention, believing that the knowledge can be obtained later by other means. In several interesting recent studies, Sparrow, Liu, and Wegner (Reference Sparrow, Liu and Wegner2011) reported that learners are less likely to remember new information if they believed that they could access that information later online. Further, learners seemed to prioritize remembering where the information could be found over the information itself. While it is tempting to think that the Internet, social media, video gaming, and so forth are creating negative effects on our long-term learning capabilities, it is also the case that technology has positive effects on cognitive development and our ability to learn and to imagine (Greenfield, Reference Greenfield2009). And if useful information changes, such as with regularly updated software or company policies, then it may be appropriate for learners to know where to get that information rather than memorize it in the form first encountered.
Summary
In summary, research in cognitive neuroscience has increasingly identified both structures of the brain and neurological response patterns related to learning. We believe that by incorporating psycho-neurological responses into existing research on instructional processes, we can better examine the effectiveness of and mediating processes related to training interventions.
We also offer a cautioned against overinterpreting the results of cognitive neuroscience research as it relates to training, given the maturity and sophistication of the research to date.
That said, we reviewed research on cognitive-neurological responses related to both sleep and attention and offered a number of recommendations for training practice including, when possible, spreading training over multiple days to allow consolidation effects through sleep, structuring the work and training environment to encourage trainees to get sufficient sleep during the course of training, building in breaks to help learners maintain voluntary attentional control, and reducing the number of distracting stimuli in the learning environment.
Conclusion
The implications we have discussed from human cognition and emerging neuroscience should be integrated into the science of training as strategies to increase learning and transfer by preparing learners and developing training programs that are engaging, meaningful, and effortful. We encourage the direct application of our recommendations to improve training as well as more research on connecting basic research on human learning to training.
