How learning works: explained simply

(10 minute read)

Multi-store model of memory explained

Recreation of the Multi-Store Model of Memory. Adapted from Atkinson & Shiffrin (1968).

Numerous models of memory exist, but we’ll focus on the Multi-Store Model of Memory, proposed by Atkinson and Shiffrin (1968). I chose this model because it strikes a balance between simplicity and conceptual power. For clarity, I will only include details that are crucial to the understanding of learning.

Sensory memory

Beginning from the left, sensory memory is a brief, high capacity storage system that registers incoming sensory information—sight, sound, smell, taste, and touch. Sensory memory holds an immense amount of information, but only for a fleeting moment. Most information is lost in a fraction of a second unless attention is directed to it, in which case it is transferred to short-term memory.

Short-term memory

You can think of short-term memory (STM) as your “conscious window”, a sort of bubble containing information that you are actively aware of (Baddeley & Hitch, 1974). STM is where you keep the last sentence of the book you’re reading, the phone number you just heard, or your grocery list, if you like to wing it.

However, STM is limited in both duration (typically 15 to 30 seconds) and capacity—you can only hold about four chunks of information at once (Cowan, 2001). A "chunk" is a unit of meaning. For example, the 50 states in the USA can be thought of as 50 chunks of information, but the USA itself is one chunk of information. Due to the limited capacity of your short-term memory, you cannot think of all 50 states at the same time. That is, unless you “chunk” the states together into larger pieces of information, such as the United States as a whole which easily fits within one of your four "slots".

Long-term memory

Some information in STM is encoded into long-term memory (LTM), where it can last anywhere from a few minutes to the duration of ones existence. Its capacity is virtually unlimited, as each of 86 billion neurons can form thousands of connections with others, each with varying strength. There are more possible combinations than atoms in the known universe. This is validated by observational studies, as no hard limit has ever been approached.

The most effective way to solidify these memories is through retrieval—the act of bringing them back into consciousness. Retrieval varies in quality, and high-quality retrieval significantly boosts both the longevity and accessibility of your memories.

Think of your memory system as an hourglass. A wide base (sensory memory), a narrow neck (short-term memory), and a broad reservoir again (long-term memory). The bottleneck—short-term memory—limits what can be retained.

Considering this severe bottleneck in short-term memory, our ability to encode information depends primarily on optimizing the potential of this cognitive window. Let’s now explore how to do that.

Understanding encoding: how it affects learning and recall

Encoding is the transfer of information to a form that can be preserved in long-term memory. This transfer allows the information—which would otherwise be quickly forgotten—to be stored for an appreciable amount of time.

To orient ourselves, lets understand our goal with encoding. As we sit in lecture or read our textbooks, our sensory memory will be bombarded by stimuli. Selective attention will allow us to capture the relevant details into our short-term memory. From there, we will deeply process the information to facilitate transfer to our long-term memory, where it can be stored strongly until we need it again.

Recreation of the Multi-Store Model of Memory. Adapted from Atkinson & Shiffrin (1968).

Consider the sea turtle. For every 1000 eggs laid, only one will become an adult turtle. Its perilous journey can be compared to the perilous journey of our sensory information, of which only a very small fraction of the will survive to LTM. While considerably less tragic, the casualties of the latter journey can make or break your learning outcomes. There are two stages of encoding, and only a fraction of the information will survive each one.

The initial stage–sensory memory to STM—is the most deleterious. Just as the hatchling sea turtles face immense danger on their initial crawl from the beach to the shore, few of the sensory details arrive at short-term memory.

Stage 1 of 2: sensory memory to short-term memory

Recall that sensory information is forgotten extremely quickly. The world around you moves rapidly, and so too does your sensory memory. It can’t be concerned with the sights, sounds, smells, tastes, and textures of one second ago, those are expelled to make room for the sights, sounds, smells, tastes and textures of now. If the information does not make it to short-term memory in this tiny window of time, it is lost forever.

The transfer of information from sensory memory to short-term memory is determined primarily by attention (Broadbent, 1958; Chun, Golomb, & Turk-Browne, 2011). If sensory information is not attended to, it is rapidly lost. The information that receives focused attention is transferred to your short-term memory where it can be further processed. This restricted transfer reflects the strict limit of your short-term memory, which holds very little information (if you recall, about four chunks).

If you have ever read a book while distracted, you may have caught yourself “reading without reading”. Your eyes gloss over the words, but you cannot comprehend the text until you reorient yourself back into focused attention. In this case, your short-term memory is full of distracting thoughts, and there is little room for any of the sensory information to arrive at short-term memory. Instead of being transferred to short-term memory, the words in your sensory memory are overwritten by the newest words you just read second-by-second.

Stage 2 of 2: short-term memory to long-term memory

Among the information that you are consciously aware of (in short-term memory), some of it is remembered (transferred to long-term memory) and some of it is forgotten.

Once in STM, information must be processed in order to be encoded into LTM. But not all encoding is created equal.

Shallow processing (e.g., rote memorization) leads to fragile memories.
Deep processing (e.g., meaningful engagement) creates strong, lasting memories (Craik & Lockhart, 1972)

Later, we will learn exactly how to facilitate deep processing with as little extra effort as possible. But in general, it involves two processes: organizing the material into a meaningful schema (this happens within STM) and then integrating that schema with your prior knowledge (connecting what's in STM to something in LTM).

From a practical standpoint, encoding as a whole can be considered little more than a setup to the true golden child of learning: retrieval. Since retrieval is the recall of information from long-term memory, encoding to long-term memory just happens to be a prerequisite for retrieval.

How retrieval works in learning and memory

Sea turtles are unlikely to survive until adulthood without food. Similarly, your memories are unlikely to survive until your exam without retrieval. We are wrapping up with the sea turtle analogy, but first I want to point out that I am deliberate in equating retrieval to food. Retrieval is sustenance; its vital importance to the longevity and accessibility of our memories cannot be overstated. After all, getting information into your head is easy. The real challenge is getting the information out of your head when you need it: which is precisely your task during an exam!

Recreation of the Multi-Store Model of Memory. Adapted from Atkinson & Shiffrin (1968).

Recall our definition of retrieval: the transfer of information from your long-term memory to your short-term memory—or in other words—remembering.

When you remember something, the neurons representing that memory re-fire, strengthening the neural pathway (Karpicke & Roediger, 2008; Roediger & Butler, 2011). This results in three powerful benefits:

1. Accessibility

Accessibility refers to the ease of bringing a memory to consciousness. With each retrieval, the memory becomes easier to access; the activation threshold lowers, and less cueing is required to recall it.

Suppose you are encounter "homeostasis" (the body's tendency to maintain internal balance) in a biology class.

At first, the concept doesn’t stick easily. You need heavy cueing—like seeing diagrams of sweating and shivering, hearing the term repeatedly, or having a teacher walk through the steps. Without those cues, it’s hard to recall or explain.

You begin actively retrieving the idea—quizzing yourself, explaining it aloud, imagining examples (like body temperature, blood sugar, hydration). Over time, you need fewer cues. Just hearing “internal balance” or “regulation” makes the concept fire in your mind. Retrieval strengthens the neural pathways, making them more excitable and therefore making homeostasis more accessible during an exam.

2. Durability

As the concept of homeostasis is strengthened through repeated retrieval, it not only becomes easier to access but also remains accessible for a long time, especially compared to passive learning methods like re-reading (Karpicke & Roediger, 2008). The neural pathways become more deeply and comprehensively embedded in long-term memory. It is more resistant to both passive decay (forgetting over time) and interference (forgetting due to conflation).

3. Transfer

This strengthening also improves transfer and flexibility, allowing the knowledge being recalled to be used in a wider range of contexts and applications (Butler, 2010; Carpenter, 2012). The act of remembering—especially in different formats such as explaining, applying, and analyzing—helps your brain distill the general principles to form representations.

You may start seeing homeostasis beyond biology. In psychology, it explains emotional balance. In economics, market equilibrium mirrors it. In your own life, you realize sleep, stress, and diet follow similar balancing principles. The concept now applies flexibly across domains because retrieval helped you abstract the core idea and apply it in new ways.

Transfer of knowledge is crucial for strong exam performance. While many exam questions will ask you to simply recall knowledge, professors often test deeper understanding through a sneaky trick: transfer questions. These ask you to apply what you know in unfamiliar or novel scenarios. A rigid, surface-level understanding isn’t enough. Exams call for flexible knowledge, and retrieval is the way to get it. By and large, flexible knowledge is useful knowledge, whether you're in school or not.

Proper encoding and retrieval are essential for effective learning and studying. While knowing these processes is helpful, the real challenge lies in applying them consistently and effectively. In the next section, you'll discover not only the proven techniques that amplify encoding and retrieval, but also a field-tested, hard-earned system for actually integrating them into a study schedule. This is where theory meets practice and where real progress begins.

To study effectively, it’s essential to recognize that not all subjects call for the same types of learning. Broadly speaking, academic learning falls into two major categories: declarative and procedural. Each calls for different learning and study approaches.

Declarative vs procedural learning: differences and study strategies

Declarative learning is about acquiring facts, concepts, and explanations—things you can consciously recall and explain.

Common in: Biology, psychology, history, medicine
Involves: Terminology, definitions, cause-effect relationships, conceptual frameworks
Typical exam tasks: Define, explain, compare, recall, outline

Success in declarative classes depends heavily on your ability to remember and organize large amounts of information, and retrieve it accurately when needed.

Procedural learning is about developing skills: learning to apply knowledge through repeated, deliberate practice until it becomes more automatic.

Common in: Math, physics, programming, chemistry, statistics
Involves: Problem-solving, algorithms, formulas, multi-step processes
Typical exam tasks: Solve, apply, analyze, calculate, troubleshoot

These classes demand more than knowing a concept. You need to use it quickly and correctly in new situations.

Higher-order study strategies like analysis, synthesis, and evaluation are especially valuable for procedural or applied subjects, as they build flexible understanding and problem-solving ability. While these methods deepen learning across all domains, lower-order strategies (such as targeted memorization and retrieval practice) can be far more time-efficient when preparing for declarative-heavy exams. The goal isn’t just to learn deeply, but to maximize outcomes while minimizing wasted effort.

The most effective study method is the one that helps you earn the highest GPA in the least amount of time. Often, it’s the method that best matches the type of knowledge you’re learning and the kind of thinking your test demands.

Studying isn't yoga...

...it’s powerlifting. Your brain adapts to deliberate, effortful practice—not 10,000 Anki cards.

Most students waste entire days “studying” without feeling prepared. SAM distills decades of cognitive science into a brutally efficient system that multiplies your results 3–4x.

I made the first part free, and it's just a 10 minute read. Read it, save hours, no strings attached. I'm sharing it because it's genuinely very useful and I want you to see for yourself.

Read part 1 free

Why SAM?

Add back 10–15 hours a week—the equivalent of a whole day—without falling behind.
Two-hour read. Lifetime return.
Zero guesswork. Zero wasted time. Know exactly what to do for every course, every exam, every semester.
Unlock med school, law school, or any grad program without burning out your time, energy, or sanity.

References:

Atkinson, R. C., & Shiffrin, R. M. (1968). Human memory: A proposed system and its control processes. In K. W. Spence & J. T. Spence (Eds.), The psychology of learning and motivation (Vol. 2, pp. 89–195). Academic Press. Baddeley, A. D., & Hitch, G. J. (1974). Working memory. In G. H. Bower (Ed.), The psychology of learning and motivation (Vol. 8, pp. 47–89). Academic Press. Broadbent, D. E. (1958). Perception and communication. Pergamon Press. Butler, A. C. (2010). Repeated testing produces superior transfer of learning relative to repeated studying. Journal of Experimental Psychology: Learning, Memory, and Cognition, 36(5), 1118–1133. https://doi.org/10.1037/a0019902 Carpenter, S. K. (2012). Testing enhances the transfer of learning. Current Directions in Psychological Science, 21(5), 279–283. https://doi.org/10.1177/0963721412452728 Chun, M. M., Golomb, J. D., & Turk-Browne, N. B. (2011). A taxonomy of external and internal attention. Annual Review of Psychology, 62, 73–101. https://doi.org/10.1146/annurev.psych.093008.100427 Cowan, N. (2001). The magical number 4 in short-term memory: A reconsideration of mental storage capacity. Behavioral and Brain Sciences, 24(1), 87–185. https://doi.org/10.1017/S0140525X01003922 Craik, F. I. M., & Lockhart, R. S. (1972). Levels of processing: A framework for memory research. Journal of Verbal Learning and Verbal Behavior, 11(6), 671–684. https://doi.org/10.1016/S0022-5371(72)80001-X Karpicke, J. D., & Roediger, H. L. (2008). The critical importance of retrieval for learning. Science, 319(5865), 966–968. https://doi.org/10.1126/science.1152408 Roediger, H. L., & Butler, A. C. (2011). The critical role of retrieval practice in long-term retention. Trends in Cognitive Sciences, 15(1), 20–27. https://doi.org/10.1016/j.tics.2010.09.003