To truly make the best use of our memory, we must understand how it functions. Memory is broken down into three stages: encoding, storage, and retrieval. During encoding, information is acquired. In storage, information is preserved and retrieval is the process of accessing it when needed. One influential explanation of how memory functions is the Multi-Store Model (MSM), proposed by Atkinson and Shiffrin (1971). 

This model suggests that memory is stored in three different locations; the sensory register, short-term memory, and long-term memory. When a stimulus is experienced, it first enters the sensory register, where it remains for 0.5 to 4 seconds. Each sense has its own sensory register:

• Visual information is stored in the ionic register

• Auditory information in the echoic register

• Touch in the haptic register

• Taste in the gustatory register

• Smell in the olfactory register

After focusing on a stimulus, it moves into the Short-Term Store (STS). Short-term memory has a limited capacity of 7 ± 2 items and can hold information for approximately 15-20 seconds, or longer if rehearsed. Milner (1966) found that by ‘chunking’ data, the storage capacity and memory retention can increase. Information that is processed and rehearsed then moves into the Long-Term Store (LTS), a system that encodes, stores, and retrieves information and has unlimited capacity. There are several factors that influence encoding such as the depth of processing, the organization of information, and distinctiveness of items.

Support for the MSM has been found in the famous case study of patient HM by Milner et al. (1968). They studied the role of the hippocampus on memory formation and supported the idea that memory has more than one store. Patient HM had undergone brain surgery to treat severe epilepsy but developed anterograde amnesia, and could not form new episodic or semantic memories. This suggests the removed brain structures were needed for transferring information from short-term to long-term memory. Researchers conducted a longitudinal case study, spanning over 50 years. They used various methods, including psychometric tests, direct observations, interviews, cognitive testing, and MRI scans. Findings revealed that while HM’s short-term memory was intact, he could not transfer information to long-term memory. For example, he could remember a number like 584 for only a short period of time. However, his procedural memory remained unaffected; he could perform motor skills like mowing the lawn without remembering having learned them. MRI scans revealed damage primarily to the hippocampus and parts of the temporal lobe. Patient HM’s results indicate that the brain’s memory system is highly specialized and complex. HM’s ability to remember events that happened long before his surgery indicate that the medial temporal region is not the site of permanent storage but plays a role in the organization and storage of memories elsewhere in the brain. This further supports how the brain has separate stores for memory and highlights the role of the hippocampus in memory transfer from STM to LTM.

While the MSM has strong supporting evidence, it has also faced criticism. The Levels of Processing (LOP) framework was presented by Craik and Lockhart (1972) and proposed an alternative theory. It argued that memory retention depended more on the depth of processing rather than on distinct memory stores. Mueller & Oppenheimer (2014) explored this idea by conducting three experiments to investigate whether taking notes on a laptop or by hand affects academic performance, and to investigate how verbatim overlap might serve as an indicator of the depth at which information is processed. They found that students who took notes by hand performed better than students who typed their notes because handwritten notes required deeper cognitive engagement. Laptop users had verbatim notes, requiring less processing, and negatively affected comprehension and retention. This research contradicted the core concepts of the MSM as it emphasized that the way information is processed plays a greater role in moving it into the long-term memory rather than rehearsal. 

Despite its limitations, the Multi-Store Model of memory remains a foundational theory when understanding how memory functions. Supported by the studies such as Milner et al. (1968), it provides a framework for explaining memory storage. However, more recent research highlights that memory is a dynamic process, influenced by how deeply and meaningfully information is engaged with. Thus, while the MSM has its strengths and limitations, it continues to offer valuable insights and practical applications in the real-world.