If the ultimate goal of training and educational interventions is to increase what learners know so they can make more informed decisions and perform better, then it stands to reason that the ultimate value of any training intervention is dependent on the effective storage (and retrieval) of new information within the learners’ memories.

This might be one of those blinding-flashes-of-the-obvious moments: Training and learning are fundamentally dependent on human memory systems. Importantly, understanding the implications of human memory systems on training and learning is not simply a theoretical pursuit. In the past few years, more and more has been published and presented about the connections among adult learning theory, cognitive science and neurophysiology – or what some have simply coined “brain-based learning” or the “science of learning.”

While it may appear to some that the complexity of this evidence is overwhelming, there are examples of lower-hanging fruit that trainers can immediately leverage to evolve and improve their training programs. The simplest, most practical and perhaps most valuable way of understanding and leveraging the science of learning is to acknowledge the critical bottleneck created by our memory system, or systems, as it were, and then ground one’s planning and design intelligently.

Memory Systems

The sub-system model of memory was first proposed in the 1960s as a framework for understanding how and when learning can be accelerated and training can be enhanced.

As described by Atkinson and Shiffrin, information is received through the sensory memory system. This first sub-system can process a huge amount of information but can retain it only for a short period of time (milliseconds). As the information in sensory memory is instinctively filtered, the information that is deemed to have some relevance or value is raised to awareness and enters the second sub-system: working memory.

Working memory (re-)organizes the information so that it may be efficiently stored as discrete neural maps within the third sub-system: long-term memory. Absent of disease, long-term memory has a theoretically limitless capacity in terms of storage duration and volume, but storage and retrieval are not same thing.

For learning to take place, working memory must effectively encode the information by finding relevance to prior knowledge and attaching critical retrieval hooks that support long-term memory storage and are critical for down-the-road retrieval when the information is needed in the future.

Herein is the key problem: Sensory memory has evolved to be nearly limitless, allowing humans to see, hear, smell, feel and taste huge amounts of data per second. As such, sensory memory may best be understood as the gatekeeper to the fight-or-flight response: The more information we can take in and instinctively process and filter, the better our chances are in the wild (or the workplace).

Likewise, long-term memory similarly evolved to store a nearly limitless amount of information, allowing humans to retrieve, relate and compare prior experiences to new ones and determine how they may impact safety and security.

However, a variety of research has infamously suggested that working memory has a much, much smaller capacity. In 1956, Miller postulated that working memory cannot typically process more than about seven independent pieces of information at once. While the number and composition of this limited sub-system have been debated, one point holds true: Limitations in working memory create a learning “bottleneck” and limit the capacity for information to be processed into long-term memory.

Overcoming the Bottleneck

In subsequent research over the past four decades, three forces, or loads, have been identified that impact the working memory bottleneck:

• Intrinsic load depends on several factors: the experience of the individual, the complexity of information being presented and the extent to which the elements interrelate with each other. In other words, intrinsic load is determined by the relative complexity of content for a given learner.
• Germane load refers to the level of concentration required for learning and appears to be determined by the motivation or lack of motivation to learn.
• Extraneous load refers to the general effort necessary for consuming content but is not directly related processing information within working memory. In other words, extraneous load refers to the environment and distractions.

When these three forms of cognitive load exceed the learner’s working memory capacity, they impair performance and learning.

Both literature and experience suggest that more often than not, trainers have far less control over intrinsic and germane load than over extraneous load. For example, the content to be taught (or objectives to be met) is not always prescribed by a trainer. As a result, intrinsic load may be pre-set.

Likewise, the motivation or concentration of a group of learners is often variable, and a carrot-and-stick approach to learning may be counterproductive. As a result, germane load is beyond a trainer’s control. Therefore, for trainers to optimize learning, we must work tirelessly to minimize extraneous load.

It might help to restate this essential obligation once more: There are real, scientifically valid, critical limitations to learning that trainers must accept and design educational interventions to overcome. These limitations are why large, multi-day congresses, three-hour symposia, passive e-learning (i.e., hour-long play-and-stay video lectures) and disjointed educational experiences are largely ineffective for learning. Trainers cannot force more information into the heads of learners whose working memory has reached capacity, because they are uninterested, unengaged and incapable of processing more information. Simply put, the limitations of one’s memory systems prohibit it!

Identifying the Sources of Extraneous Load

By addressing the parts of the learning experience that are more readily controlled, trainers can optimize learning. The difficulty is that while working memory is limited, the sources of extraneous load are not – and nearly every decision a trainer makes will either add to or mitigate extraneous load. The duration of interventions is a source of load. The quality of the content and faculty is a source of load. The physical environment of learning is a source of load. The usability of technology is a source of load. This list goes on and on and on.

To highlight the challenge of these barriers, recent research found a critical source of extraneous load in a rather unexpected place: the actions learners take as they engage with content itself. Learners generally acknowledge employing a set of “learning actions” that they have long believed support their learning. These actions include note-taking (to document/prioritize key insights), setting reminders (to drive reflection and action), real-time searching (to fill holes in the content), and cueing off faculty and others learners (to identify critical topics and concepts).

In data collected from thousands of learners, it appears that these learning actions are unevolved. While learners may have taken notes and set reminders the same way for 30 years or more in their own careers, evidence now suggests that these approaches are largely a matter of habit and convenience and rarely a matter of trial and error.

Survey 50 learners, and you will uncover far more than 50 often chaotic approaches to leveraging these learning actions. This problem is the dual-edged sword of any natural or long-existing habit: The actions themselves have become entwined with what it means to learn, but they are consistently overlooked by learners and trainers alike.

Consider the impact on extraneous load when learners depend on a series of actions that they deem to be critical to learning but that fragment attention, feed a fluency illusion and disconnect their personal archive of the learning experience from the content itself.

First Step to a Fix

Trainers bear a critical responsibility to support lifelong learning. In doing so, countless hours, effort and dollars are spent creating training experiences that will increase what learners know so that they can make more informed decisions and perform better. Yet training, learning and performance are fully dependent upon a human memory system that is far from perfect.

The obligation for trainers, therefore, is to understand enough about these cognitive limitations to focus on how we can minimize the load placed on learners. With enough commitment to this new obligation, we can minimize extraneous load, maximize learning capacity and optimize the value of training itself. And it all begins with designing around the working memory bottleneck.