Cognitive Load Theory and Learning Medicine

T here has been a huge increase in the amount and quality of brain science over the past five decades. Study of how the brain functions has been abetted by increasingly sophisticated and powerful imaging technologies. Experts in psychology, linguistics, neuroscience, computer science, anthropology, and philosophy have joined forces to help elucidate how the brain processes and stores information, and, among other tasks, how we learn. ^1,2

One of the most widely accepted concepts concerning how we learn is that of cognitive load theory. This theory of brain function simultaneously examines the structure of information to be learned and the cognitive brain architecture that allows learners to process (learn) that information. ³

Cognitive load theory puts forward the notion that human memory has two main components: (1) working or short-term memory and (2) long-term memory. ^{3 –7} In this construct, working memory is a temporary system, which actively holds information in the mind to perform verbal and nonverbal tasks (reasoning and comprehension) and to make it available for further information processing and complex cognitive tasks. It is the area in which all conscious cognitive processing takes place. ^3,8

Working memory, thought to be located in the prefrontal cortex and parietal areas, deals with the amount of information the brain can hold and manipulate at one time. ^3,9 Working memory is further described as being of short duration – probably only a few minutes, but up to 24 h – and of limited capacity. ³ It is the brain system that actively holds information to perform verbal and nonverbal tasks for further information processing (learning).

Working memory is postulated to be composed of a central executive and a subcomponent for processing visual-spatial information and another subcomponent for processing auditory (speech) information. It has been suggested that these subcomponents process information in relative independence of one another. ⁶

Initial study of working memory suggested that young adults could hold up to seven discrete pieces of information – plus or minus two – and, that the duration of working memory was short, lasting, perhaps, up to several minutes. ⁴ There is newer research indicating that working memory can handle only a small number of novel interacting elements; probably no more than three or four. Because human intellectual activity typically involves the interaction of multiple interacting elements simultaneously, working memory alone would not have sufficient capacity to permit more than trivial human cognitive activity. ³

Long-term memory, on the other hand, allows for the brain to store an almost unlimited amount of information (knowledge) and is not constrained by capacity. The content (or units) of long-term memory has been theorized to be sophisticated, complex structures called schema, also known as cognitive maps, that permit us to process, think, and solve problems. ¹⁰ Schemas are cognitive maps that incorporate multiple elements of information into a single unit with a specific function. ³

Schemas may be thought of as a “units of representation” of a person's world. A schema is what is learned about some aspect of the world, combining knowledge with the processes for applying it. Schema theory seeks to show that all human mental phenomena reduce to (complex) patterns of schema activation and that schemas are instantiated as dynamic processes in the brain. ¹¹

The elements of information stored in long-term memory schema are different in both structure and function from the elements in short-term (working) memory.

Long-term memory is felt to be composed of two types of memory: declarative and procedural. ¹²

Declarative (explicit) memory is information that can be spoken; that can be recalled as facts, knowledge, and understanding. This portion of long-term memory is further divided into two categories: (1) episodic memory, which stores specific experiences, and (2) semantic memory, which stores factual information. The hippocampus and surrounding structures are thought to be important in declarative memory. ¹³

Procedural memory, that is, memory for overtly doing things, may be, but is not necessarily, below the level of consciousness. This expression of memory is typically retrieved as integrated performances involving cognition and motor skills, such as riding a bicycle. Procedural learning is enhanced and accelerated by careful observation, reflection, and repetitive practice. The cerebellar cortex is presumed to encode and store these memories. ¹⁴

Cognitive load theory, which builds on these studies of how our brains function, concerns itself with the amount of information and number of interactions that must be processed in working memory during complex learning activities. Working memory, or short-term memory, is finite, and is limited in the amount of discrete information – “chunks” or elements of information – that can be retained in working memory. All information in working memory must be encoded – transferred to long-term memory – before meaningful learning can continue. ¹⁵

Cognitive load theory deals with the interaction between information structures and cognitive architecture. Information or elements that learners must process varies greatly in complexity. Low-element interactivity can be understood and learned without consideration of other elements. High-element interactivity material, on the other hand, can be learned in individual pieces, but cannot be understood until all elements and their interactions are processed simultaneously. High-element interactivity material is difficult to understand and carries a high cognitive load. ³ When considering the above, it is not difficult to understand that cognitive load theory has broad implications for present-day medical instruction and educational design activities.

Cognitive load theory posits the presence of three types of cognitive load in the process of learning: intrinsic cognitive load, extraneous cognitive load, and germane cognitive load. ¹⁵

Intrinsic cognitive load is the inherent difficulty (load) of the instructional materials themselves. Extraneous cognitive load is that load generated by the manner in which information is provided to learners. Finally, germane cognitive load is the load devoted to processing, constructing, and automation of schemas.

The teacher of medical knowledge can make little reduction in intrinsic cognitive load, the inherent difficulty of instructional materials. In a large sense, intrinsic cognitive load has been thought to be immutable.

It should be noted, however, that there is a huge variation in a person's subjective experience of difficulty in learning specific material. Factors important in determining difficulty include prior experiences, attitudes toward the category of learning material, the extent of perceived need for learning that material, and more. It may be that levels of intrinsic difficulty may not be fixed or truly inherent in the material.

Extrinsic cognitive load, on the other hand, is load caused by the manner in which information is presented, and is under the control of those who design medical educational materials. Thinking of teachers as “presenters” of learning material is now probably an archaic concept. Learners need to be given situations, tasks, challenges, and context that make clear the need for the knowledge and understanding that they must acquire. The task of learning is the learners' responsibility.

When learners are convinced of the need to learn specific material, they do what works for them to learn. When learners are empowered and motivated to learn, they pursue learning with all the effort that they are capable of producing.

For learners to learn and teachers to teach, educational design for medical learning must focus on decreasing the load caused by the manner in which the educational materials are presented. There is broad agreement that individuals learn better when they build on what they understand (their existing schema).

Germane cognitive load, or load devoted to the processing, constructing, and automation of schema needs to be a focus of the learner's attention and promoted.

Major goals of medical education should, therefore, include a reduction in processing of task-irrelevant information by better design of learning materials and their presentation, and/or, by better design of instructional activities. At the same time, medical educators should seek ways to increase the amount of mental effort involved in the processing of task-relevant materials and construction of related schemas. ^16,17

Taken as a whole, these new insights into how we learn and what loads are imposed while we learn provide potential new ways of thinking about medical education. More importantly, these insights provide new ways to evaluate medical educational materials and improve the efficiency of learning.