When designing learning technologies, specifically k-12 educational technology, what learning strategies derived from cognitive psychology can be used to improve our product’s efficacy?
As far back as we have a record of human thought and the thought of our own existence, think the philosophy of Plato and Aristotle, we have evidence of the exploration of the nature of knowledge and how it is acquired. Behaviorism dominated the zeitgeist of learning theory in the early 20th century. Behaviorists considered learning to take place through a series of stimuli and positive reinforcement (Edgar, 2012). The breadth of stimuli and the form with which it can take multiplies with every generation of learners and the ever-increasing tools at their disposal.
In the late 1950s, the learning theory approach shifted from behavioral to cognitive models. Psychologists and educators focused less on observable behavior and more on cognitive processes like thinking, problem-solving, language, concept formation, and information processing (Ertmer & Newby, 2013). From empiricism to rationalism to behavioralism to cognitivism and constructivism, our understanding of how we learn has grown to align with what we know of the physiology of the brain and the contemporary way, at each historical period, humans take in information. The goal of this literature review is to understand the science behind how we learn and then align that to the construction of educational technology.
So, if constructivism means that we create meaning by attaching it to our own schema, thus creating a reality specific to each individual learner, how do we curate that learning so that the outcome matches the intended learning goal no matter the schema? In “Managing Cognitive Load in Technology-Based Learning Environments,” Kalyuga and Liu break down Cognitive Load Theory as an instructional theory that addresses the two main components of cognitive architecture: working memory and long-term memory (Kalyuga & Liu, 2015). In order to improve educational technology efficacy, we can start by employing the implications of cognitive load theory. Curum & Khedo, 2020, categorize cognitive load into intrinsic, germane, and extraneous loads. The goal of the instructional designer is to simplify the intrinsic load by decreasing the complexity of new information, reduce the extrinsic cognitive load by eliminating distractions from processing new information, and then maximize the germane cognitive load by allowing for the deep processing of new information from short-term to long-term memory. We can scaffold learning by controlling and minimizing the cognitive load (Clark & Harrelson, 2002) in a digital learning platform.
The principles of cognitive load are a list of guiding pedagogical techniques used in cognitive learning theory. In my pursuit of research articles addressing psychological best practices of learning design, these principles were researched and written about at length. I decided to do a deep dive into understanding these principles through the study of the research articles referenced below.
I made the choice to focus on Cognitive Learning Theory rather than the Cognitive Theory of Multimedia Learning because I felt it important to first understand these guiding principles in the general application before studying the more focused and synthesized theory of multimedia learning.
Principles to Incorporate
The first principle, “Worked Examples,” is described by J. Sweller as when learners are asked to start their knowledge acquisition by first studying the completed examples (Sweller, 2019). This process allows the learner to get a global view of what is expected in the learning task by seeing a completed and correct version first. This reduces the primary cognitive load of introducing a new topic by introducing the strategies they can use to solve a similar task in the future.
“Worked Examples” also references the Information Store principle that says the information we store in our long-term memory may not necessarily improve our ability to problem-solve but rather give us more options with which to solve a problem (Sweller, 2019)- meaning the more correct examples learners are exposed to, the greater their repository for solutions. The learning technology that first comes to mind to masterfully use this principle is Khan Academy. Their primary use of simple videos working through problems rocked the instructional world. The crude handwritten, voice-over videos allowed the learner to reduce their cognitive load and access a low-stakes learning environment where problems were solved in the most simple methods.
Kalyuga describes the next layer of the gradual increase of cognitive demand principles, “Completion Tasks.” Completion tasks, give learners partially completed tasks and challenge the learner to fill in the gaps to completion (Kalyuga, 2012). This principle both decreases the cognitive burden and exposes learners to solutions for future reference. The simplification of intrinsic load primes the learner for a smoother transition of knowledge to long-term memory. This is the next logical step in scaffolding learning however, it is often skipped, especially in digital learning platforms. We see, more often, the direct transfer from “sit and get” (watch this amazing video or click through this learning management system ) and then immediately move to an unscaffolded demonstration of knowledge.
Qualities of gamification can play a huge role in the use of “Completion Tasks” as it allows learners to interact with content in a more creative and often collaborative way. Often, as learning technology designers, we focus on putting what we can do on display rather than giving learners opportunities to move through the process with us.
The “Imagination Effect” requires learners to form mental representations by imagining a solution or way to come to a solution for a problem before actually acting on those plans. An example might be when a student would begin by mentally visualizing how they would read and extract essential information from a graph before proceeding to answer the question.
The process of developing knowledge on a topic results in the formation of schemas, which are knowledge structures that aid in interpreting the surrounding world. By imagining instructions or plans for solving, students can automate these schemas, leading to reduced cognitive load in their working memory and a higher likelihood of transfer to long-term memory (Kalyuga & Liu, 2015). As learning technology designers, we can incite reflection and planning by forcing interactions around planning or withholding questioning before the learner has the opportunity to visualize. The ability to pause progress and incite this type of cognitive strategy can be immensely easier on a digital platform versus on a physical platform where the learner has more control.
Sweller et al. describe the principle of the “Isolated Interacting Elements Effect” as the act of breaking down larger, more complex concepts into individual elements first and then integrating the elements into the more complex learning goal (Sweller et al., 2019). This principle states that once the individual elements are stored in the learner’s long-term memory, the learner only needs to now learn how to integrate the individual elements, which because of the nodes connected to elements learned, the new integrated skill will more easily attach to schema and transfer to long-term memory. For a learning designer to be successful here, in most cases, it will be imperative that they are also the content expert or are closely working with the content expert. Accuracy in breaking down content into relevant chunks takes a deep understanding of the content, and incoherent attempts to do so could prove not only futile but will negatively add to the cognitive load of the learner.
In the “Guidance Fading Effect,” the gradual release of scaffolding is given to the learner as they become more proficient on a topic (Sweller et al., 2019). This principle ensures that the level of support maintains the appropriate level of cognitive load, maintaining the zone of proximal development (Vygotsky, 1978). Employing the aforementioned principles, the “Guidance Fading Effect” may be achieved by starting with a “Worked Examples” activity, then moving to a “Completion Tasks” activity, and then to a complete problem to solve independently. Frequent assessments of knowledge would need to be integrated into the learning platform to understand where the learner is on their learning trajectory. Then the design must authentically gradually release scaffolding by adapting content based on the assessment data. I will cover this more when exploring “Retrieval Practice” below.
Perhaps the most abstract instructional activity would utilize the “Goal-Free Effect”. Much like an open-ended question, the “Goal-Free Effect” lessens the scaffolding of the pursuit of one specific answer and opens the opportunities for demonstrating proficiency to an indeterminate number of possibilities. “Goal-Free” problem-solving is an effective method for reducing cognitive load and facilitating knowledge construction by forcing learners to focus on the knowledge they have to facilitate problem-solving, providing an optimal combination of low load and solution-focused thinking (Sweller et al., 2019). To employ the “Goal-Free Effect” learning technology, designers have to avoid the constraint of designing specific goals around a task and instead design a platform for learners to develop creative and innovative ideas around exploring the problem and understanding its various elements to create multiple outcomes or solutions.
Principles to Avoid
The “Split Attention” and “Modality Effect” involve the problem of increasing cognitive overload due to the presence of information in different places, like various pages in a book or intertwined together, like text over images. Another example is when information is displayed in a written format on one side of a slide and a diagram on the other, necessitating constant switching between the two. This can result in cognitive overload, leading to lower learning outcomes for students who learn in a split-source format, according to research. Comprehension can be hindered by both of these by the disruption of the learning pathway when the learner must switch their attention over and over (Dunlosky et al., 2013). A great example of this has been on the Texas state language arts assessment. For years, students have been asked to read a passage and then, pages later, answer specific content questions. In a natural scenario, we would lay our resources and activities side by side to reduce the cognitive load due to “Split Attention”. However, years of adherence to antiquated methods have separated resources for learning and assessment. The new State of Texas Assessments of Academic Readiness (STAAR) positions passages next to the assessment questions, drastically reducing the presumed cognitive load and allowing students to focus on the content and their demonstration of knowledge.
Schwonke reviews the detriment of incorporating elements in your learning environments that would be applicable to the “Redundancy Effect” (Schwonke, 2015). The redundancy effect unnecessarily increases the cognitive load by cluttering the learning environment (be it physical or digital) by including unimportant information. This inclusion of information makes the process of moving memory from short-term to long-term far more taxing by increasing the extrinsic cognitive load. As a longtime elementary educator, I have heard or seen this over and over; a well-meaning teacher believing that “more is better” or that a “cute butterfly” in the corner of a presentation having nothing to do with butterflies will increase engagement. Good design eliminates the clutter so that the learner can focus on the learning task at hand while still delivering an aesthetically pleasing interface.
Sweller cautions against the use of the “Isolated Interacting Elements Effect” on learners who are beyond the novice level; this is because of the “Expertise Reversal Effect”. This principle cautions against teaching below the learner’s level or providing scaffolds when they are not needed (Sweller et al., 2019). It is crucial to consider students' current level of understanding when designing lessons, and ensure that the material is appropriately tailored to their level. We see a backward slide from once high-performing students in classrooms all of the time. If learning is below the Zone of Proximal Development, those neurons are not being fired or challenged and assessments reveal a downward trend in performance. To understand a learner’s current level of understanding, periodic assessments of knowledge must continually take place.
Many, if not all, but especially the “Expertise Reversal Effect” requires both the student and the instructor to possess an acute knowledge of the learner’s current level of understanding so that correct principles for learning can be engaged. Constant assessment is necessary to gauge the current level of understanding. But, with the current zeitgeist concerned with over-assessment and the subsequent burnout of learners, the idea of assessment often is given a bad rap. This is where “Retrieval Practice” proves beneficial both in the process of learning and the continuation of learning.
Tests are commonly used for assessment in educational settings, but they also have a lesser-known benefit: improving memory of the information being tested (Weinstein et al., 2018). Despite our memories being like libraries of information, retrieval (as happens during testing) actually strengthens our memory. Essentially, the act of trying to retrieve the information or retrieval-based learning in an assessment gives the learner practice moving the information from long-term to working memory. This theory makes the case for frequent formative assessments to benefit the learner and instructional author. We see this in digital learning platforms that boast the highest yield of learning, referred to as “adaptive” this means that the software strategically stops along the learning projection to assess and recalculate the instructional map. However, we often see a lack of reflection points for the learner on these assessments, which would disrupt the retrieval process as they are never able to confirm or reassess what is stored in their memory.
Conclusion
Continued research and reflection on the current state, as it applies to learning in the digital landscape, of cognitive load will always need to be reassessed as the human condition evolves. The ability to take in more information at one time from a digital platform varies radically from generation to generation. This is already evident when observing the number of platforms with which a young person can continually and authentically interact, versus the one-by-one nature in which older adults communicate digitally and shows strong signs of an adaptation towards the increased ability to process multiple streams of information in split-attention situations. Sit in a car with a 14 year old who is carrying on full conversations in multiple platforms cover any number of topics, while also hearing what you say in real-time (if they choose to) and it is easy to see that the capacity for input and output is rapidly changing (of course the depth of communication is actually being engaged it is questionable).
As learning technologies continue to evolve along with our ability to authentically engage and learn with technology-based learning tools, we have to recognize the pitfalls along with the benefits: personalized learning experiences, increased engagement, and flexible learning platforms. However, it has also led to a number of challenges, one of which is the issue of cognitive overload with which learning designers must be ever cognizant.
Cognitive Load Theory principles give learning designers a framework for understanding how learners process information and how our cognitive resources are utilized in learning. Each of these principles, while developed with a more antiquated platform for learning in mind, are applicable to the design of learning technologies because they are rooted in brain science and foundational truths about learning. By utilizing digital design tools thoughtfully in tandem with Cognitive Learning Theory principles, we can simplify the intrinsic load and reduce the extrinsic load, thus maximizine the germane load and facilitating better retention and recall. By knowing how to control the cognitive load and employing these strategies in learning technology, we can greatly improve the efficacy of our learning technology products and resources.
References
Clark, R., & Harrelson, G. (2002). Designing Instruction That Supports Cognitive Learning Processes. J Athl Train, 37(4 Suppl), S152–S159. PMID: 12937537; PMCID: PMC164417.
Curum, B., & Khedo, K. K. (2020). Cognitive load management in mobile learning systems: principles and theories. Journal of Computers in Education, 8(1), 109–136. https://doi.org/10.1007/s40692-020-00173-6
Dunlosky, J., Rawson, K. A., Marsh, E. J., Nathan, M. J., & Willingham, D. T. (2013). Improving Students’ Learning With Effective Learning Techniques: Promising Directions From Cognitive and Educational Psychology. Psychological Science in the Public Interest, 14(1), 4–58. https://doi.org/10.1177/1529100612453266
Edgar, D. W. (2012). Learning Theories and Historical Events Affecting Instructional Design in Education. SAGE Open, 2(4), 215824401246270. https://doi.org/10.1177/2158244012462707
Ertmer, P. A., & Newby, T. J. (2013). Behaviorism, Cognitivism, Constructivism: Comparing Critical Features From an Instructional Design Perspective. Performance Improvement Quarterly, 26(2), 43–71.
Kalyuga, S. (2012). Interactive distance education: a cognitive load perspective. Journal of Computing in Higher Education, 24(3), 182–208. https://doi.org/10.1007/s12528-012-9060-4
Kalyuga, S., & Liu, T. C. (2015). Managing Cognitive Load in Technology-Based Learning Environments. Educational Technology & Society, 18(4), 1–8.
Karpicke, J. D., Blunt, J. R., & Smith, M. A. (2016). Retrieval-Based Learning: Positive Effects of Retrieval Practice in Elementary School Children. Frontiers in Psychology, 7. https://doi.org/10.3389/fpsyg.2016.00350
Schwonke, R. (2015). Metacognitive Load ? Useful, or Extraneous Concept? Metacognitive and Self-Regulatory Demands in Computer-Based Learning. Journal of Educational Technology & Society, 18(4), 172–184.
Sweller, J. (2019). Cognitive load theory and educational technology. Educational Technology Research and Development, 68(1). https://doi.org/10.1007/s11423-019-09701-3
Sweller, J., van Merriënboer, J. J. G., & Paas, F. (2019). Cognitive Architecture and Instructional Design: 20 Years Later. Educational Psychology Review, 31(2), 261–292. https://doi.org/10.1007/s10648-019-09465-5
Vygotsky, L. (1978). Mind in Society: The Development of Higher Psychological Processes. Harvard University Press.
