The structure and function of the multi-store model of memory

The multi-store model (MSM) of memory, developed by Atkinson and Shiffrin in 1968, provides a framework for understanding how information is processed through a series of storage systems. According to this model, memory involves a linear flow of information from initial sensory input to more permanent storage, passing through distinct stages.

Core components of the multi-store model

• Sensory memory (SM) - The first stage where information from the environment is briefly captured by the senses in its raw form.
• Short-term memory (STM) - A temporary storage system where information is held for a short period if attention is paid to it.
• Long-term memory (LTM) - A more permanent storage system where information is retained for extended periods after sufficient processing.
• Information flow - Sensory input enters SM, and if attended to, moves to STM. Through rehearsal or deeper processing, information can then transfer to LTM for lasting storage.

This model highlights the differences between these stores in terms of how information is encoded, their capacity to hold data, and the duration for which information is retained.

Characteristics of sensory memory, short-term memory, and long-term memory

Each memory store in the MSM has unique features that define how it processes and retains information.

Features of sensory memory (SM)

• Encoding - Information is stored in its original sensory form, such as visual images or auditory sounds, without conscious control.
• Capacity - Extremely large, capable of holding vast amounts of sensory data momentarily.
• Duration - Very brief, lasting only a few milliseconds for visual data and up to 2-3 seconds for auditory data.

Features of short-term memory (STM)

• Encoding - Primarily acoustic, meaning information is often stored as sounds, though other sensory forms can also be used.
• Capacity - Limited to approximately 5-9 items at a time.
• Duration - Lasts up to around 30 seconds, though this can be extended through active rehearsal of the information.

Features of long-term memory (LTM)

• Encoding - Mainly semantic, where information is stored based on meaning rather than sensory input.
• Capacity - Potentially unlimited, with the ability to store vast amounts of information.
• Duration - Can last for a lifetime, retaining information for extended periods.

Key research supporting the multi-store model

Numerous studies have provided evidence for the distinct nature of the memory stores described in the MSM. These investigations highlight differences in encoding, capacity, and duration across SM, STM, and LTM.

Studies on sensory memory

• Crowder (1993) - Discovered that visual information in SM lasts only a few milliseconds, while auditory information persists for 2-3 seconds, indicating separate sensory stores with varying durations.

Studies on short-term memory

• Peterson & Peterson (1959) - Participants recalled 85% of nonsense trigrams after 4 seconds, but only 10% after 22 seconds, suggesting STM's duration is limited to a short window unless rehearsed.
• Jacobs (1887) - Found that the capacity for recalling lists of numbers was around 10 items and for letters around 8 items, demonstrating the limited capacity of STM.

Studies on long-term memory

• Bahrick et al. (1975) - Showed that individuals who left school within the last 10 years recognised 85% of former classmates' faces and names from photos, while those who left 50 years ago recognised 75% of names and 65% of faces, indicating LTM's long-lasting duration.

Baddeley (1966) study on encoding in STM and LTM

• Method - Involved 80 participants who were given one of four word lists, repeated five times: acoustically similar (similar-sounding), acoustically dissimilar (different-sounding), semantically similar (similar meaning), and semantically dissimilar (different meaning). STM was tested by immediate recall of the correct order, while LTM was assessed after a 25-minute delay with an interfering task.
• Results - In STM, recall for acoustically similar words was poorest at 10%, compared to 60-80% for other lists, indicating acoustic confusion. In LTM, recall for semantically similar words was lowest at 55%, compared to 70-85% for other lists, showing semantic confusion.
• Conclusions - STM primarily relies on acoustic encoding, as similar-sounding words are harder to recall. LTM depends more on semantic encoding, as words with similar meanings cause confusion.
• Evaluation - The study supports the MSM by showing distinct encoding methods for STM and LTM. However, small differences in STM recall between semantically similar (64%) and dissimilar (71%) lists suggest some semantic encoding occurs in STM as well.

Evaluations of the multi-store model's strengths and limitations

The MSM has been influential in shaping our understanding of memory, but it is not without criticism.

Strengths of the multi-store model

• Support from amnesia cases - Some patients lose either STM or LTM capabilities but not both, reinforcing the idea of separate memory stores.
• Influential framework - As one of the first cognitive models of memory, it sparked significant interest and laid the groundwork for later theories like the working memory model.
• Research backing - Extensive studies confirm the existence of distinct stores for SM, STM, and LTM, validating key aspects of the model.

Limitations of the multi-store model

• Oversimplification - The model assumes single STM and LTM stores, yet evidence suggests multiple types within each, such as verbal and non-verbal STM stores, and procedural, episodic, and semantic LTM stores.
• Focus on structure over process - It emphasises the structure of memory stores but pays less attention to the processes involved in memory formation and retrieval.
• Neglect of information nature - The model measures capacity purely by quantity, ignoring that the ease of recall can depend on the type or complexity of information rather than just the amount.

Practical applications and issues in memory research

The MSM offers practical insights into enhancing memory, alongside raising important considerations about how memory research is conducted and interpreted.

Practical ways to improve memory - chunking

Chunking, or grouping information into larger, meaningful units, can increase STM capacity. For example, recalling 19141918 as two units (1914 and 1918, representing the start and end of the First World War) is easier than remembering it as eight separate digits.

Limitations in memory research

• Nomothetic approach - Most studies on the MSM rely on laboratory experiments that assume memory functions similarly for everyone, overlooking individual differences in memory processing.
• Controlled settings - While lab studies provide valuable data, they may not fully reflect real-world memory use, where context and personal relevance play significant roles.