In our long-term memory, information is stored indefinitely from the working memory. Learning is about retrieving and storing information in this memory. Our long-term memory stores information that we receive. This information is stored in so-called schemas where new schemas are created or existing schemas are expanded: this is what we call learning. These schemas (knowledge and skills) are context-dependent. The long-term memory consists of the non-declarative memory and the declarative memory. The non-declarative memory is used for automated knowledge and skills, such as tying your shoes, swimming, reading, and simple calculations. This memory also includes association: automatic ready knowledge, such as thinking of the word 'butter' when you hear the word 'bread'. This can also be learned through classical conditioning, as with Pavlov. The declarative memory is used for non-automated knowledge, such as having a substantive conversation or learning concepts. It consists of semantic memory (facts, meanings, and concepts) and episodic memory (reconstructing events based on time, place, and emotional context). These two systems work together: learned concepts are linked to the context (for example, the story or the physical space). These are like anchor points that make it easier to remember and recall knowledge, but can also make it harder if the context changes. Students often remember episodic details of a lesson (lesson flow), without precisely knowing what was discussed (concepts). By telling stories with interesting facts, you can utilize the connection between episodic and semantic memory: this makes storytelling a powerful way to learn. By breaking down complex information into small chunks, that knowledge can be automated (non-declarative). In this way, students can retrieve it more easily. Additionally, it is effective to have students apply knowledge (and not just reproduce it), so that it becomes less context-dependent. Information (schemas) in long-term memory is not stored like a file on a computer: memories continuously change and can thus also lead to false memories. This can be a problem in learning because memories (prior knowledge) are difficult to modify. Long-term memory supports working memory by allowing existing information to be linked to new information. In this way, you reduce cognitive load (see Cognitive Load Theory). By using learning strategies such as retrieval practice, spaced practice, and interleaved practice, you strengthen the schemas around the knowledge you are working on. This makes it increasingly easier to retrieve that information from long-term memory. This is called learning. It is also beneficial to partially forget information first, so that you train the brain to forget it less quickly the next time (better retention).
Dideau, D. & Rose, N. (2019). Psychologie in de klas: Wat iedere leraar moet weten. Culemborg, Nederland: Phrones
Weinstein, Y., Sumeracki, M., & Caviglioli, O. (2018). Undertanding how we learn: A visual guide. New York, Verenigde Staten: Routledge.